Senate Analysis 2004

2004-11-24T15:02:00.000-05:00

Sorry TCO and other assorted folks. I still am mucho busy and haven't time to compile my thoughts on the '06 races.

I'll do it sometime, but not likely any time soon...

Sorry.

2004-11-11T14:11:00.000-05:00

I'm still around, but barely. I've gotten way too busy lately to type up a 2006 summary, but it should be here sometime. Don't worry - we've got two years to pundate on those races!

Prediction Analysis Wrap-up

2004-11-04T23:09:00.000-05:00

Well, at last the election is over. My predictions (posted last Saturday here) held up very well against the final results – much better than many of the left-wing blogs and media pundits predicted a break even result of better for the Dems – and, had I decided to amend my results due to my suspicions on Monday, I would have scored every single race right. But, alas, the caution that I have been taught throughout doing this blog convinced me otherwise. So, in all, I was 32 for 34 on the night.

I predicted the final net gain precisely at GOP +4, which, of course, was the final result. The Senate going into the 109th Congress will be 55-44-1 for the Republicans. I did pretty well in picking several spreads, too. But enough gloating. Let's go through every single race and my predictions, and maybe I can learn something for 2006:

Each race has a nifty table that shows my prediction and the final result. The bottom row shows whether my predicted winner actually won, and then the difference between my predicted margin and the final margin (negative means my picked candidate underperformed the prediction, positive, of course, means the opposite).

Alabama
Predicted		Shelby by 65+%
Results		Shelby 68%-32%
Overall	Yes	Margin	-29%

My most horrifically wrong call. I severely overestimated Shelby's popularity in this state and severely underestimated the number of straight-ticket Democratic voters. In fact, this occurred in many of the safe-incumbent races that had not been polled all season. Guessing at the margin in those races was basically a shot in the dark, and I generally overstated their margin of victory, underestimating the number of straight-ticket opposition party voters. Ah, well, such are the follies of politics.

Alaska
Predicted		Knowles 48%-47%
Results		Murkowski 49%-45%
Overall	No	Margin	-5%

What can I say? This one surprised me. There were very, very few polls of this race, so there was little to gauge Murkowski's support from. Assuming the KTUU polls to have been correct at the time, I underestimated the strength of Murkowski's late charge and overestimated Knowles' support. Funny - I was calling this race for Murkowski throughout the whole election season, citing Alaska's extremely conservative roots as my basis and effectively ignoring the polls until the end, when I finally bowed to the polls, and guess what? The polls were wrong. Alas, live and learn.

Arizona
Predicted		McCain by 60+%
Results		McCain 76%-21%
Overall	Yes	Margin	+5%

I should say I did fairly well on this one...

Arkansas
Predicted		Lincoln by 20%
Results		Lincoln 56%-44%
Overall	Yes	Margin	-8%

Once again, overstating the incumbent's support, although many polls showed this race wider before the election.

California
Predicted		Boxer by 17%
Results		Boxer 58%-38%
Overall	Yes	Margin	+3%

Likewise, little to say here. Boxer outperformed the final polls slightly, but not by much. My prediction was close.

Colorado
Predicted		Coors 49%-48%
Results		Salazar 51%-47%
Overall	No	Margin	-5%

Whereas conservative Alaska carried Murkowski to victory, Colorado is apparently just not conservative enough to lift Pete Coors to a win. The problem here was not, like Alaska, a lack of polls, but too many polls that disagreed with each other. In the days leading up to the election, polls literally alternated showing Salazar and Coors ahead. In the end, I simply relied too much on Colorado's Republican base to turn-out.

Connecticut
Predicted		Dodd by 47%
Results		Dodd 66%-32%
Overall	Yes	Margin	-13%

See Alabama. Another clear case of overstating a safe incumbent's support.

Florida
Predicted		Martinez 50%-49%
Results		Martinez 50%-48%
Overall	Yes	Margin	+1%

I do wish I had gone to Vegas on this one. I had folks e-mailing yelling at me that Florida was going to elect Castor. Bologna, I say. Really, calling it for Martinez was all that was needed - everyone knew it would be close, so the margin of victory was basically a given. I still think I did pretty well picking the final totals, though.

Georgia
Predicted		Isakson by 21%
Results		Isakson 58%-40%
Overall	Yes	Margin	-3%

Isakson, while outperforming the latest polls of the race, underperformed my predictions. No matter, as I was fairly close.

Hawaii
Predicted		Inouye by 60+%
Results		Inouye 76%-21%
Overall	Yes	Margin	+5%

No comments. A fairly good prediction for a race with no polls, yes? Or at least a fairly lucky one...

Idaho
Predicted		Crapo by 95+%
Results		Crapo w/ 99%
Overall	Yes	Margin	-4%

It must be immensely satisfying for Crapo to know that he won with 99% of the vote. I thought it nearly impossible to overperform this prediction, but, when one's only opponent is a write-in candidate, anything's possible.

Illinois
Predicted		Obama by 48+%
Results		Obama 70%-27%
Overall	Yes	Margin	-5%

I am proud that I was able to call such a wide-open race so well, although surely I was helped by the fact that there were about seventeen polls more of this race than anyone wanted or needed. Obama performed very nearly at the average of all the recent polls of this race.

Indiana
Predicted		Bayh by 33+%
Results		Bayh 62%-37%
Overall	Yes	Margin	-8%

Off a bit. Eh...

Iowa
Predicted		Grassley by 40+%
Results		Grassley 70%-28
Overall	Yes	Margin	+2%

See Indiana.

Kansas
Predicted		Brownback by 65+%
Results		Brownback 69%-28%
Overall	Yes	Margin	-21%

Off only slightly more. See Alabama - there were no polls of this one either.

Kentucky
Predicted		Bunning 52%-43%
Results		Bunning 51%-49%
Overall	Yes	Margin	-7%

This one was scary. Kentucky, at 7PM EDT, was one of the first states to close its polls, and as the first results started coming in, Mongiardo led approximately 60-40. Could this be a harbinger of impending Democratic blowout? Nay. The first returns were from eastern, urban, heavily Democratic Kentucky. As western and northern Kentucky returns came in, Bunning closed the gap. Although the margin was certainly smaller than he'd like it to be and far smaller than my prediction, Bunning pulled it out in the end.

Louisiana
Predicted		Vitter w/ 52%
Results		Vitter w/ 51%
Overall	Yes	Margin	-1%

Did I not tell you that Vitter would win without a run-off? Ah yes, I believe I did. I mispredicted John's and Kennedy's totals, understating and overstating their support levels, respectively, but it really matters not, as the race is over for good now.

Maryland
Predicted		Mikulski by 30+%
Results		Mikulski 65%-34%
Overall	Yes	Margin	+1%

Very good prediction for a race with no polls. Sometimes you just get lucky...

Missouri
Predicted		Bond by 25+%
Results		Bond 56%-43%
Overall	Yes	Margin	-12%

This race turned out much closer than nearly any poll had shown it to be. I have no idea why and no excuses to make. Bond still won, and by 12 points. Moving on...

Nevada
Predicted		Reid by 31+%
Results		Reid 61%-35%
Overall	Yes	Margin	-5%

No comments.

New Hampshire
Predicted		Gregg by 40+%
Results		Gregg 66%-34%
Overall	Yes	Margin	-8%

No comments.

New York
Predicted		Schumer by 47+%
Results		Schumer 71%-25%
Overall	Yes	Margin	-1%

A very good prediction, Josh. Very well done. You deserve a prize. I was a bit off on my second prediction (of Schumer leading 35% over Mills and O'Grady) as O'Grady polled just 3% of the vote, but still - very well done. ;)

North Carolina
Predicted		Burr 50%-47%
Results		Burr 52%-47%
Overall	Yes	Margin	+2%

I seem to remember saying, in the very first article, that although Bowles led in early polls by 9 or 10 points, Burr would likely come back and win this seat. I was right. Not only was I right, but I very nearly got the margin, too. I feel so special.

North Dakota
Predicted		Dorgan by 60+%
Results		Dorgan 68%-32%
Overall	Yes	Margin	-24%

What can I say? See Alabama.

Ohio
Predicted		Voinovich by 26+%
Results		Voinovich 64%-36%
Overall	Yes	Margin	+2%

See what you can do with polls? The number of polls of this race was horribly disproportionate to its competitiveness, but did help to call it much better.

Oklahoma
Predicted		Coburn 49%-45%
Results		Coburn 53%-41%
Overall	Yes	Margin	+8%

Wow. I knew that if Coburn was a slightly better candidate, he'd take this running away, but I had no idea that he would win by such a margin.

Oregon
Predicted		Wyden by 25+%
Results		Wyden 63%-32%
Overall	Yes	Margin	+6%

No comments.

Pennsylvania
Predicted		Specter by 16+%
Results		Specter 53%-42%
Overall	Yes	Margin	-5%

This race was similar to Kentucky, with early returns heavily favoring Hoeffel. They were the eastern, urban Pennsylvania returns, though, and with later returns Specter evened the race and eventually led by a significant margin, although underperforming many polls and my predictions.

South Carolina
Predicted		DeMint 51%-41%
Results		DeMint 54%-44
Overall	Yes	Margin	0%

Right on the dot. Sure the numbers weren't right on but I called the margin perfectly. Am I just the coolest or what?

South Dakota
Predicted		Thune 51%-48%
Results		Thune 51%-49
Overall	Yes	Margin	-1%

I have to get some points for this. Like Florida, it was bound to be close no matter who won, so the margin was a given. Tom Daschle is deeply, deeply saddened by these results, I'm sure.

Utah
Predicted		Bennett by 40+%
Results		Bennett 68%-29%
Overall	Yes	Margin	-1%

No comment, except to note how closely the polls called this one.

Vermont
Predicted		Leahy by 55+%
Results		Leahy 71%-25%
Overall	Yes	Margin	-9%

Slightly over-predicted, but few polls to work with. What do you expect?

Washington
Predicted		Murray by 17%
Results		Murray 55%-43%
Overall	Yes	Margin	-5%

In the end, it looks like Murray effectively split the difference between SV and the other pollsters.

Wisconsin
Predicted		Feingold by 16%
Results		Feingold 56%-44%
Overall	Yes	Margin	-4%

Rasmussen pegged this one best, and Zogby was WAAAY off. A fairly close call, in my opinion.

Looking at the final results as a whole, I note two trends:

1.In races with safe incumbents and few polls, it is generally difficult to call the exact margin. In many of these races, I underestimated the number of straight-ticket opposition voters. Without many polls of the races, predicting margins is a crapshoot. Still, I did alright on some of them.
2.I tended to overestimate the effects of thrid-party and independent candidates. When doing the predictions, I researched on Politics1 to see if there were any other candidates, and many races had several, and I accounted between 2-3% for them. What I didn't know is that many weren't on the ballot, and so polled significantly fewer votes than expected. Bilyeu in OK did about as expected, though. This trend affected the races very little, and only threw off predictions by a couple of percent.

So, out of all the races, my predictions turned out to be:

Straight up: 32-2
Against the spread*: 26-8

In the 13 most competitive races (the ones given Race Pages) – the numbers in parentheses are for the nine listed as Toss-up or Leaning in the last update:

Straight up: 11-2
Against the spread*: 11-2

* - Using a self-concocted analysis mechanism - when the predicted margin is under 20%, the difference between predicted and final must be 5% (a normal MoE in polls) or less to be considered a good prediction. When the predicted margin is over 20%, the difference must be within one-quarter the predicted margin.

Alrighty, there you have it. Coming tomorrow (maybe): a look at the 2006 races, and then it's curtains for this blog.

Finally starting...

2004-11-04T20:10:00.000-05:00

...the prediction wrap, if there's anyone still out there. Will be up tonight or tomorrow.

Running...

2004-11-04T14:11:00.000-05:00

...between about seven different things right now, but I'll try and have my prediction wrap tied up tonight.

Also, a reader suggested perhaps a quick overview of the Senate races in 2006 and what can be expected in them before I close up shop here. Good idea, and if schedule allows, one I will pursue in earnest, before all the readers desert...

The Morning (slash Afternoon) After...

2004-11-03T16:22:00.000-05:00

Well, the masses have cast their vote, and I daresay I predicted the final results pretty well. Not quite as well as KivaCom, but better than most of the MSM pundits expecting a tightly drawn Senate late into the night (libby media indeed). I had meant to post this last night, but at 3AM I couldn't bring myself to do it, and early this morning was sleepy time until about five minutes before I needed to be to school.

All in all, I do believe my predictions went pretty darn well...

In the Presidential contest, I do believe I predicted a Bush victory with 51% of the popular vote and 296 EVs. With effectively all of the results in, Bush has, surprise, a 51% majority of the popular vote, and 286 EVs (included IA nad NM, which Bush has won, but the media refuse to acknowledge - although it really doesn't matter). I missed but one state in my predictions - Wisconsin, which I believed would go for the Prez. It was close, but my hopes/predictions didn't quite pan out. Still, 98% isn't bad, no?

In the House, I predicted a GOP pick-up of 6 to 7 seats. I haven't been following individual races, and so was flying by the seat of my pants on a round number that was culled from various other sources following the races more closely. So I never really put a lot of faith in this prediction, and, frankly, never really cared, as the House was not leaving Republican control, and any gains would be token at best. But enough excuses. I have seen sites, from left to right, tallying the final net at break even to GOP +5. Fox News and CNN both have it at GOP +4, and it appears three races left to call. I'm checking those out shortly. Regardless, Democrats had decent showings in a few races nationwide, but nothing to be overly excited about, as they still lost overall net. Bean beat Crane in IL-8, a shocker to be sure, and Edwards held on in TX-17. Those two results alone held the GOP official gain down. Still, I was very close, and it remains clear that the GOP will have a sizable majority in the 109th Congress.

I'll come back in a bit with the analysis of my Senate predictions and the final results.

EDIT (on the third try (6:22 PM) - Blogger's being real finicky): I realized after I posted that I didn't even show the three races that were too close to call. The first is NY-27, a Republican-held open seat where the Democratic candidate looks to be just slightly squeaking by (the media's not called it - I think they've probably given up on calling anything altogether by now). That cuts the GOP net to +3. The other two are Louisiana seats headed to a run-off. One is a Republican-held open seat, LA-3; the other a Democratic-held (Chris John's seat), LA-7. Both had Republicans at the top in the final Election Night vote tallies, but history has shown that Louisiana run-off races cannot be very heavily judged by the November results. We can presume, though, that the Republicans will either hold one or sweep both, making the final GOP tally either +3 or +4. I wasn't off by much...

Senate predictions analysis will be a bit later than expected, but is on its way...

After Months of Anticipation...

2004-11-02T13:57:00.000-05:00

...the Day of Election is upon us. There will be no live-blogging the results from me tonight. I doubt I have the significant readership to justify such doings, and I'm still getting over this nasty influenza. I plan to plop down in the recliner about 3 o'clock and not move for 12 hours or so.

But I'll be watching closely the returns and taking notes. Sometime soon, possible early tomorrow morning if I'm feeling motivated, tomorrow afternoon if I'm not, I'll have a rundown of the results and how they match up with my predictions. I'll thank everyone profusely, and then, barring a runoff in Louisiana, this blog will be dead to the world.

In the meantime, KivaCom and MEA have predictions up, and for presidential predictions, the always-right Daly Thoughts provides us with the final result 12 hours early. RCP has slightly more polls up, but they really don't make a difference now...

For the record, here are my predictions:

Bush wins with 51% of the popular vote, 296 EVs.
Senate GOP +4 to 55-45, as detailed earlier.
House GOP +6-7, to whatever it is now plus six or seven...

Months of punditry and pontificating have come down to this one day. It's bound to be an interesting night. I'll see y'all on the other side!

2:42 PM Update: I can feel the tension running through the blogosphere right now. It's quite exciting. I forgot to mention previously - John J. Miller has his last-minute predictions up at NRO as well - I'm scanning them right now.

Time is running out...

2004-11-01T19:17:00.000-05:00

Apologies for the lack of posting the last coupla days. Went to the Bush rally yesterday in Gainesville, with Jeb, Laura, and Mel (<--first time I've actually seen the guy). Nearly 30,000 people showed, even in this liberal town (there were only a dozen or so token protestors).

Then, overnight, I was promptly taken ill with what appears to be the flu. It's all Bush's fault, right? ;)

Anyway, I've spent probably 85% of today in bed, and have been unable to post any new polls or amend my predictions. Just glancing at the polls on RCP, though, I see no reason to change my predictions. I'm less certain about my Florida, Colorado, and Alaska predictions, although at the moment I don't think there is a need to change them. If anything, Colorado and Alaska would switch, and the overall outcome would be the same.

For those still looking for a poll fix, just go to RCP. I'm too out-of-it right now to post polls and update pages. Hopefully I'll be better before tomorrow night.

Also, KivaCom and MEA have put up their final take on the races. Both are highly respectable and logical analyses, even if I disagree with their take on a few races, and I eagerly await KivaCom's take on the nine competitive races, coming tomorrow.

Take care, all...

FINAL PREDICTIONS

2004-10-30T20:18:00.000-04:00

Sorry for yelling. Here they are, the ultimate results of months of punditry, for your consumption.

Senate Analysis 2004 – 11/1/04 Final Senate Race Predictions

Over the past several months, I've followed the Senate races closely, noting every poll and watching every trend. I've made predictions and then rescinded them. I've made mistakes and occasionally shown some insight (odds are out as to which I've done more, though). As the election nears – just tomorrow, in fact, it's time for my final predictions.

So, here they are. Not only am I predicting the winner of each race, I'm predicting the margin of victory – closer races have vote tally predictions, too – just to give me more chances to screw up (predictions for less competitive races will probably be off significantly – possibly 5-10 points or so – but I consider that much less egregious than mis-predicting a close toss-up race). I'll post after results are in tallying up my right-wrong record.

By state, alphabetically:

Alabama: Incumbent Richard Shelby faces only nominal opposition. This is one of several races that haven't even been polled this cycle. Shelby by 65+%.

Alaska: This is possibly THE hardest race to call. If it were any other Republican candidate, this race wouldn't be close. If it were any other Democratic candidate, this race wouldn't be close. If it were in almost any other state, this race wouldn't be close. As it is, though, this race is a complete toss-up between Murkowski and Knowles.

Still, all the polls to date have shown Knowles leading, and although Murkowski has regained a substantial amount of ground, I don't know if she'll be able to pull back up to even or better by Election Day. Bush's coattails may help slightly, but I doubt many Alaskans will be hesitant to cast a split ticket. Third party candidates will probably siphon a few Republican votes, providing the final blow to Murkowski. I'm predicting a Knowles win, ever-so-slightly, 48%-47%, and he will quite likely be unseated in 2010 by a much less vulnerable and much more conservative Republican candidate. Dem. Takeover.

Arizona: John McCain is well-liked by many, and his opposition is faint. No polls are available of this race either. He'll be back for another term. McCain by 60+%.

Arkansas: Republicans were hoping to overturn Blanche Lincoln, but it's unlikely. One poll showed her challenger within 14 points, but that's the closest he's gotten, and her margin has widened substantially since then. Lincoln by 20%.

California: Incumbent Barbara Boxer faced significant opposition from Bill Jones, but in liberal California, he was never able to gain footing, and Boxer's lead has only widened as of late. Boxer by 17%.

Colorado: Next to Alaska, this is the hardest race to call. Ken Salazar is a very strong, centrist Democratic candidate, and has polled very well throughout the campaign, but oddball polls have been prevalent here, and I'm not sure which one to trust. Flipping a coin, I see Coors with the momentum and, with a 50% chance of being wrong, am predicting he will keep this seat Republican, 49%-48%.

Connecticut: Incumbent Chris Dodd faced little opposition. The latest poll had him up by 45, and he will likely exceed that margin on Election Day. Dodd by 47+%.

Florida: Florida is one of few presidential swing states also with a competitive Senate race this year, and both candidates seem to be aligning themselves with their party's presidential pick. I expect the result of the Senate race to be closely tied to the presidential results, and there will probably be very few cross-over votes.

With that said, then, polls have shown this race clearly very close throughout the campaign. Lately, though, Mel Martinez seems to have opened up a significant margin, leading in almost all recent polls. There are very few undecideds left in this race, and I believe the final result will be quite similar to recent polls, with the few undecideds breaking fairly evenly. I'm picking Martinez to win 50%-49%. Rep. Takeover.

Georgia: The fate of Zell Miller's seat was never really in question. Johnny Isakson has led for the entire campaign, and although his lead has shrunken slightly as of late, he will still cruise to a significant victory. Isakson by 21%. Rep. Takeover.

Hawaii: Incumbent Dan Inouye, along with having one of the strangest names in the Senate, has had little trouble in this election. No polls have been taken of this race, and none are really needed, although with the interesting state of affairs in the presidential race in HI, this race may be slightly closer than expected. Inouye by 60+%.

Idaho: Senator Mike Crapo does not have a Democratic challenger, and I'm not sure if this race will even be on the ballot. Crapo by 95+%.

Illinois: Ever since Barack Obama's stunning keynote at the DNC, and the decision of the IL GOP to import Alan Keyes as their candidate, this race has been quite unexciting. Obama by 48+%. Dem. Takeover.

Indiana: Incumbent Evan Bayh has faced little contest, and will be back for another term. Bayh by 33+%.

Iowa: Incumbent Chuck Grassley has faced only small opposition, and will be back in DC for another term. Grassley by 40+%.

Kansas: Incumbent Sam Brownback has had little contest in this race, and no polls are even available for it. Brownback by 65+%.

Kentucky: One of the dark-horse contests after Senator Bunning's erratic behavior of late, Democrats became suddenly enthused, and left-wing sites listed this race as a possible takeover after a partisan poll showed the race tied. State Senator Dan Mongiardo has made a valiant effort and gotten several breaks in this race, but will probably not be able to make it all the way in this conservative state, although he will make it much closer than anyone would have predicted three months ago. Bunning, 52%-43%.

Louisiana: I absolutely refuse to explain LA's election system again. If you're reading this, you should know about it. With that said, I've been vocally predicting all along that Vitter could get through this one without a runoff, and I still think I'm right. Recent polls have him between 48 and 51 percent, and over 10% of all voters are undecided. A majority of the undecideds are likely undecided between Democratic candidates, but nonetheless it should not be hard for Vitter to pick up a few points.

I'm predicting Vitter becomes Louisiana's first Republican senator, polling slightly north of 52% on Election Day, with Chris John winning the race for useless second with 23%, Kennedy with 20%, and Arthur Morell with 3% or so. Rep. Takeover.

Maryland: Incumbent Barbara Mikulski will cruise to another term by a considerable margin. Mikulski by 30+%.

Missouri: Earlier this cycle, Senator Kit Bond looked vulnerable to the DSCC, and many started pouring money into Nancy Farmer's campaign, hoping to effect an upset. Unfortunately, Bond's vulnerability never developed, and he has led significantly throughout the campaign, and will walk to another term. Bond by 25+%.

Nevada: Likewise in Nevada, Republicans were optimistic about defeating sitting Assistant Minority Leader Harry Reid, but a disappointing primary field drowned any prospects of an upset. Reid by 31+%.

New Hampshire: Incumbent Judd Gregg will cruise over his 81-year-old opponent, Doris “Granny D” Haddock. Gregg by 40+%.

New York: Chuck Schumer is facing two major opponents, one from the Republican Party, another from the Conservative Party of New York. It won't matter, as he'll crush both of them. Schumer by 47+% over his nearest opponent, 35+% over the combined total of both.

North Carolina: Back in the summer, Erskine Bowles maintained a significant lead in this race to replace John Edwards. I predicted then that not only would the race tighten, but his opponent, Richard Burr would come back and eventually win the race. So far, it appears my prediction will hold. Burr has led in six of the last eight polls, with the other two showing a tie. It will still be close, but I'm predicting a Burr win here, 50%-47%. Rep. Takeover.

North Dakota: Incumbent Byron Dorgan has faced little opposition, and no polls have been taken of this race. Dorgan by 60+%.

Ohio: The Ohio race has been polled substantially, but only because the state is a major presidential battleground. Challenger Eric Fingerhut has been unable to pick up steam, and Incumbent George Voinovich will cruise to a second term. Voinovich by 26+%.

Oklahoma: This race should never have really been close. If you look at the polling trends, Coburn led for most of the summer, which melted into a solid Carson lead for most of the fall, and has only recently, in the past several weeks, come back to a steady Coburn lead. Coburn, like DeMint in SC, has a nasty habit of shooting off his mouth when it would be politically expedient to shut up, which has hurt him significantly. He could probably be leading by ten points if it weren't for some stupid comments he's made.

Don't count out Sheila Bilyeu's impact on the race, either. The left-wing wacko has polled upwards of 5%, which would assumably have gone to Carson otherwise, perhaps making the difference in the race.

This race is hard to call numerically due to the large amount of undecideds still left, even a few days before the Election. I'm guessing they break about 3-out-of-5 for Carson, and Coburn pulls it out 49%-45%, with Bilyeu picking off 5%.

Oregon: Incumbent Ron Wyden will cruise to a second term. Wyden by 25+%.

Pennsylvania: Democrats were quite optimistic about defeating incumbent Arlen Specter, particularly after a divisive fight in the Republican primary, but their hopes have failed to pan out. Specter has rarely polled below a double-digit lead, and will win re-election by a significant margin. Specter by 16+%.

South Carolina: Previously in the season, this race was considered to be, alongside Georgia, an assured Republican takeover. Jim DeMint led by double digits for most of the summer. However, as the race heated up, he made a few off-color comments, and suddenly the race tightened.

However, he never once lost the lead in non-partisan polls, and to this day enjoys a significant advantage. I'm predicting a DeMint win, 51%-41% over Tenenbaum. Rep. Takeover.

South Dakota: The sitting leader of either party should never be in electoral trouble. Not since the '50s has one been defeated in his re-election bid. Tom Daschle, the Senate Minority Leader, is up for re-election this year, and is being faced by John Thune, a former Representative. Thune lost to Tim Johnson, SD's other Senator, by just a few hundred votes in 2002.

Throughout the campaign, polls have been fairly sparse, and the lead has flip-flopped back and forth among the non-partisan polls. This race is quite clearly a complete toss-up, but Thune has been gaining momentum lately, and South Dakota is, I think, ready for a change. I'm picking Thune to knock off Daschle by a significant margin, 51%-48%. Rep. Takeover.

Utah: I'm running out of ways to say the incumbent has faced little to no opposition and will be re-elected. Incumbent Bennett by 40+%.

Vermont: Ditto on the Utah comment. Incumbent Leahy by 55+%.

Washington: In Washington, Republicans had high hopes of knocking off Patty Murray, and recruited a decent, if not exemplary, candidate for the job in Rep. George Nethercutt. Nethercutt, however, has been unable to gain the advantage, and has never polled within close range of Murray. She will be re-elected by a very un-close margin. Murray by 17%.

Wisconsin: Likewise, the NRSC hoped to defeat incumbent Russ Feingold here, but the late primary hampered their efforts. Feingold, although polling somewhat close, has never looked significantly vulnerable, and has increased his lead throughout the campaign. He will cruise to a substantial victory and another term. Feingold by 16%.

Republican pick-ups: GA, SC, NC, SD, LA, and FL
Democratic pick-ups: IL, AK

Well, that's it. Republicans look to gain a net 4 seats, to a new advantage of 55-45 (Jeffords counted as a Democrat, of course). As I've continually stressed, I've tried to be as objective as possible throughout the campaign season, and we'll see whether my predictions are more my hopes as a Republican, or the result of my analysis of the polls.

'Twill be posted on FR momentarily. Please feel free to comment.

8:43 PM Update: 'Tis here. Enjoy...

8:54 PM Update: Nope. Stupid FR Mods moved it here. Not that you care...

10:31 PM Update: Now here, hopefully to stay. Yeesh... Also, thanks for the heads-up on Mikulski. Just another case of journalistic sloppiness on my part...

7:28 PM, 11/1 Update: This is when these predictions were supposed to be posted. Luckily I had the foresight to throw them out early, as, had I not, they likely wouldn't have gotten done (reference above post). Thanks for the heads-up on Senator Gregg, Dirk. How embarrassing...

The Countdown has Begun...

2004-10-30T13:14:00.000-04:00

...until Election Day. Many voters have already voted, and polls at this point are really rather worthless. I have several to post, then I'll do page updates, and my final predictions will be posted this evening.

Firstly, I apologize. There was far too much partisan excitement in that Oklahoma post. Live with it.

This post, though, will be devoid of partisan content. Now then, on to the polls!

Zogby tracking results for 10/29:

Colorado: Salazar 50%, Coors 45%. Toss-up.

Florida: Martinez 46%, Castor 47%. Leaning Rep. takeover.

Wisconsin: Feingold 57%, Michels 37%. Safe Dem. retention.

Pennsylvania: Specter 57%, Hoeffel 28%. Safe Rep. retention.

Iowa: Grassley 70%, Small 24%. Safe Rep. retention.

Nevada: Reid 57%, Ziser 37%. Safe Dem. retention.

Ohio: Voinovich 59%, Fingerhut 32%. Safe Rep. retention.

UNH Tracking of New Hampshire: Gregg 63%, Haddock 22%. Safe Rep. retention.

In Louisiana, the Verne Kennedy tracking poll shows Vitter at 48%, John with 22%, and Kennedy with 13%. The standings for the Democratic challengers have been reversed, in the last tracking update, Kennedy was slightly leading John. Regardless, I still think Vitter will win without a run-off. Leaning Rep. takeover.

Most of the other polls have been of un-competitive races:

In Georgia, a Zogby/Atlanta Journal-Constitution poll finds Isakson under 50% but still leading strongly, 49%-36%. Probably Rep. takeover.

In Illinois, a Research 2000 poll (as yet unreleased - RCP) shows the race holding steady at Obama 67%-Keyes 25%. Easy Dem. takeover.

In Connecticut, a Research 2000 poll shows incumbent Dodd leading 66%-27%. Safe Dem. retention.

Another New Hampshire poll, this one also from Research 2000 (RCP) shows Gregg ahead 68%-19%. Safe Rep. retention.

And, finally, a poll from Utah shows incumbent Bennett leading 61%-23%. Safe Rep. retention.

That's all for now. Obligatory feedback link here.

3:12 PM Update: All the pages are updated. I have some other stuff to do but should be wrapping up and posting my final predictions, as I said previously, tonight.

Oklahoma News Flash!

2004-10-29T19:36:00.000-04:00

This update was so important I had to create a new post for it. And frequent readers here know by now I rarely ever create new posts for a single poll.

Sooner Poll released a few hours ago shows results as follows: Coburn with 44.4%, Carson with 35.1%, Independent Sheila Bilyeu with 4.2%, and 16.3% still undecided.

Wow. If this poll is correct, Coburn has opened up a nine-point lead, and has gained eight points since the last Sooner Poll. If it's not correct, his lead is still nearly two spans of the MoE. It's nearly guaranteed that Coburn is leading in Oklahoma, and I can say that he will very likely win this election.

And there was much rejoicing...

I'm not so sure this poll is dead-on; I highly doubt Coburn's polling this well, and, regardless, this result is still within the Leaning Rep. retention bin, so, logically, it will stay there.

Kudos to JohnnyZ at Politics1 for the heads-up, and a plug for his election-tracking site as a reward.

Still working on the page updates, BTW...

It is the last Friday...

2004-10-29T19:07:00.000-04:00

before the election. And due to my laziness yesterday I have a ton of polls to post. Twenty-six, to be exact. In 14 races. So, straight to it, then:

Zogby tracking results for today:

Colorado: Salazar 52%, Coors 43%. I've given up even hoping for sensible results here. Stays Toss-up.

Florida: Martinez 47%, Castor 45%. Toss-up.

Wisconsin: Feingold 57%, Michels 37%. Now Safe Dem. retention.

Pennsylvania: Specter 53%, Hoeffel 30%. Safe Rep. retention.

Iowa: Grassley 71%, Small 23%. Safe Rep. retention.

Nevada: Reid 56%, Ziser 36%. Safe Dem. retention.

Ohio: Voinovich 59%, Fingerhut 29%. Safe Rep. retention.

The only change effected by these polls is in Wisconsin. Michels has lost all chance at an upset, and Feingold finally breaks the 20% threshold. Thus, Safe Dem. retention.

And, onward! Note that these updates are considered later than the Zogby tracking postage, so any designation change that follows takes precedence. In rough order of competitiveness:

In Colorado, Mason-Dixon shows the race tied at 46%. Rocky Mountain News has a drastically different result, showing Salazar leading 48%-42%. I've been chided by readers for leaving this race at Toss-up, and the new polls seemingly show Salazar with the advantage.

But 'tis not so. I, as noted previously, now assign almost no weight at all to the Zogby tracking result. Rocky Mountain News' polling has been highly erratic this season, first showing Salazar +11, then Coors +5, and now Salazar +6. Unfortunately, they did not do an accompanying presidential horserace poll with the same sample (like Zogby) to compare with other pollsters' results. I don't frankly know what to make of this other than to discount the RMN poll as well in the figures. And, as if I needed another reason to doubt the RMN poll's results, it was a POS poll (in more ways than one apparently), which is usually a Republican partisan pollster.

So, discarding the new RMN and Zogby results, and the S+11 Gallup poll from early this month, the polls of this race this month have been rather evenly spaced. Three show Salazar ahead, three show Coors ahead, and two show a tie. I believe the true state of this race lies very near the result of the new Mason-Dixon poll. Toss-up, although I dare say Salazar has some momentum nonetheless, and could take it into the election. You'll just have to wait for my final predictions to find out!

In Florida, a number of polls have been recently released. Chronologically, Insider Advantage shows Martinez ahead 46%-44%; Quinnipiac shows Martinez leading 49%-46%; Mason-Dixon shows Martinez leading 47%-46%; and the Zogby poll shown earlier has Martinez leading 47%-45%.

The Zogby poll of Florida has been quite volatile, and is, after all, a Zogby poll, but even removing that one from the equation we have three discrete non-partisan polls showing a consistent Martinez lead. Although it will take a few more polls to verify, this may signal a major shift in the race. Although all three polls' leads are within the margin of error, the statistical significance of all three nearly-simultaneous polls within a couple points of each other nearly rules out these polls as anomalies. There seems to be a definite Leaning Rep. takeover trend in this race, and it will be hard for Castor to reverse in the next few days, as news out of the race has been generally positive for Martinez and desperate for Castor (<-- note: subjective opinion, make of it what you will). Then again, two non-partisan polls showing it tied or in Castor's camp could easily swing this one back to Toss-up. A very, very slim Martinez lead indeed, but that is, after all, all that is needed.

Update before the post even goes gold: the most recently released Strategic Vision poll (not yet on the website) shows Martinez ahead 49%-46%. It is automatically disregarded since, it is by definition a partisan poll, but the trend is unmistakable.

In North Carolina, two polls confirm that this is a very, very close race. Research 2000 shows Burr leading slightly 47%-46%, while Mason-Dixon shows the race all knotted up at 46%. Based on these two polls alone the call should be Toss-up, however...

Looking at the polls out of this race in the last month, we find only one showing Bowles ahead, and that was back at the very beginning of the month (10/5). The rest have ranged from Even to Burr +6. Clearly, Burr has the advantage here, although it is slight. This race could go either way, but at the present it is indeed Leaning Rep. takeover.

In South Carolina, a Republican poll from McLaughlin shows DeMint leading 48%-40% over Tenenbaum. Yes, it's partisan, but surely it has at least some nugget of truth in it? DeMint's lead is at almost a span-and-a-half of the MoE anyway. Regardless, whether this poll affected the designation or not, this race is Leaning Rep. takeover.

In Washington, a few polls were released and it was soon found they interested no one. SurveyUSA shows Murray ahead 55%-41%, and Mason-Dixon shows the same lead, but with slightly different numbers of 53%-39%. Strategic Vision (not yet up on site at last check) continues to differ with the non-partisans, showing Murray ahead 49%-41%. Again. One wonders if they're even polling, or just repeating the same numbers over and over. Regardless, it's technically partisan, and undeniably the oddball, and is effectively thrown out. This race is still Probably Dem. retention, as Murray has yet to crest the 20-point-lead threshold, although there is little doubt as to the eventual outcome of this race.

In Wisconsin, the afore-mentioned Zogby poll clashes with a partisan poll from Tarrance Group showing Feingold vulnerable and leading only 48%-43%. When confronted with the choice between a somewhat-sensible Zogby poll and a far off-kilter partisan poll, I choose to slice off several fingers and refuse to punditrate on this. This race will stay where we all know it belongs: Safe Dem. retention. Blech.

In Georgia, Strategic Vision (again not yet posted) confirms what everyone else already knew: Isakson is winning. This time it's reportedly by a 55%-40% margin. Partisan poll or no, this race is Probably Rep. takeover.

And, on to the Non-Competitive Races! What fun!

In California, a Field Poll shows Boxer leading 53%-34%. Unsurprising, and the designation remains Safe Dem. retention.

In Pennsylvania, along with the Zogby tracking poll, Quinnipiac shows Specter leading 60%-34%, his largest lead recently in the cycle, and Strategic Vision (you know the drill - not up yet) is a bit more conservative in putting Specter ahead 55%-35%. Safe Rep. retention, either way.

In Missouri, Research 2000 (no link - RCP) shows Bond ahead 56%-40%, while the Kansas City Star shows him leading 53%-36%. Both margins are closer than safe designation, but Bond is not in trouble this late in the race, and thus, Safe Rep. retention.

In New Hampshire the University of New Hampshire is starting a tracking poll tracking the Presidential horserace, gubernatorial race, Senate race, and both House seats. The first result shows Gregg leading comfortably 62%-23%. I doubt this result will change much in the next few days. Safe Rep. retention.

In Nevada, in addition to the Zog tracking above, Mason-Dixon polls the race and finds Reid ahead 59%-35%. Not surprising, and no change from Safe Dem. retention.

And, finally, in Iowa, in addition to Zog, a Research 2000 poll (no link - RCP ) shows Grassley cruising 65%-27%. Safe Rep. retention.

And that, friends, are all the polls for today. I'll spend a bit updating the pages, scrounge up some dinner, and then polish off my final predictions. But you'll have to wait till Sunday to find out what they are.

Ogg like feedback.

Early-Morning Update, Sort of...

2004-10-29T06:21:00.000-04:00

I'm sorry I missed all of y'all yesterday - I was busy with other things.

Anyway, I have a load of updates to do, but haven't the time to do them right now, so I'll postpone until this afternoon.

I will, though, post yesterday's Zogby results, as we will have new ones by my next update.

In Colorado, Zog continues his odd results with Salazar 53%, Coors 42% - Toss-up
Florida: Martinez 45%, Castor 45% - Toss-up
Iowa: Grassley 68%, Small 27% - Safe Rep. retention
Nevada: Reid 57%, Ziser 34% - Safe Dem. retention
Ohio: Voinovich 58%, Fingerhut 27% - Safe Rep. retention
Pennsylvania: Specter 55%, Hoeffel 29% - Safe Rep. retention
and Wisconsin: Feingold 59%, Michels 36% - Probably Dem. retention, inching ever-so-closely towards Safe.

I'm also working on my final predictions. I'll post them some time this weekend.

Thanks to everyone who's feedbacked, and TGIF.

South Dakota Nonsense

2004-10-28T06:37:00.000-04:00

I've been alerted to a poll of South Dakota by Zogby from several sources, so I'll go ahead and post it.

It's being reported differently by different news outlets, so we'll go with the one linked on RCP. The Rapid City Journal reports Zogby polls showing Thune ahead 48.5%-45.5%. As I've noted before, though, I don't trust Zogby polls farther than I can throw them, and although this one isn't as blatantly off as his Colorado tracking results, I'm not quite sure as to the veracity of this one.

The results do match the last Rasmussen poll of the race, though. Also of note in the article is it refers to the KELO-TV/Argus Leader poll posted yesterday as a Mason-Dixon poll. Still looking for confirmation on that, though.

All in all, this race stays Toss-up.

Also, an unconfirmed Research2000 poll presumably leaked shows Burr leading 47-46 in North Carolina. If proven true, this wouldn't change the race from Leaning Rep. takeover.

And, finally, a confirmed R2000 poll shows Obama leading Keyes 67-25 in Illinois. As if anyone was really that concerned that this race had suddenly tightened. Easy Dem. takeover.

Some of the Zogby tracking results have been leaked, too, but I'll wait to post those until this afternoon. Thanks all who pass here. Feedback is a requirement.

Welcome!

2004-10-28T06:35:00.000-04:00

6:53 AM: I'm bumping this one back to the top for our new visitors from PoliPundit. Thanks especially to Jayson for the mention.

**********

Welcome, visitors from Daly Thoughts. I've gotten over 175 hits in the last half-hour from Gerry's mention.

Please look around, and hopefully you'll like what you see. I keep track of polls of the competitive Senate races, and summarize each day's new polls every evening or so. Past polls and posts are compiled into Race Pages. The pages for the competitive races are listed at right. Also, an Overview page lists each race, the latest poll, and its current designation in a table format.

A quick note about the designations:

Safe or Easy means a 20-point or more lead. Some races with incumbents have never even been polled, and many are no contest at all. These are also Safe. This close to the election, many races with 13-point, 15-point, or larger leads are designated Safe as well, as it would be mighty hard for a candidate to overcome that large of a margin in less than a week.

Probably signifies a 10-point or larger lead.
Likely signifies a 5-point or larger lead.
Slight or Leaning means the race is essientially a toss-up, but leaning slightly towards one candidate or another, using by just a few points and always within the margin of error.
And, of course, Toss-up means the race is a complete Toss-up. Usually recent polls will be split between both candidates.

Thanks again for stopping by and I hope you'll stick around. I always appreciate feedback, either in the form of comments on a post or e-mail (to which I always try to reply).

Cheers,
K1

9:04 PM Update: We've broken a thousand hits for the day! I'm amazed! Thanks loads to Gerry and all our new visitors. Please stick around and always, ALWAYS leave feedback, or I'll be forced to harm this kitten...

October 27, 2004

2004-10-27T18:31:00.000-04:00

Just a quick post and no updates tonight. Other things to do and not many polls today anyway.

Also, I want to thank all the visitors from Daly Thoughts again. My previous record for daily hits has been nearly doubled, with 5 hours still left in the day. I'm disappointed, though, that no one has left feedback of any sort. I'm always looking to improve the site, and, this late in the election season, gather tips for '06. The positive response I've gotten thus far means I'll likely be back for the next election cycle.

Anyways, on to the polls:

Zogby Tracking Results: - My comments from yesterday apply.

Colorado: Salazar 53, Coors 43. Toss-up.
Florida: Castor 48, Martinez 45. Toss-up.
Wisconsin: Feingold 57, Michels 38. Probably Dem. retention.
Pennsylvania: Specter 54, Hoeffel 33. Safe Rep. retention.
Iowa: Grassley 66, Small 27. Safe Rep. retention.
Nevada: Reid 59, Ziser 30. Safe Dem. retention.
Ohio: Voinovich 54, Fingerhut 34. Safe Rep. retention.

Also, the MRI tracking poll of Louisiana shows Vitter breaking 50 for the first time, at 51%. Kennedy still leads John 17%-15%. Still decidedly Leaning Rep. takeover.

And, in South Dakota, two new polls seemingly contradict each other. KELO-TV shows Daschle ahead 49-47, while Rasmussen (no link, on RCP) has Thune ahead 49-46. I'm slightly more inclined to trust the Rasmussen poll, as he is a reputable professional pollster, but that may be just me. Had either of these polls come out by themselves, I would probably swing this race to the leading candidate, but simultaneously, they clearly show a Toss-up race.

That's all for me today folks. Once again, e-mail and comments are always appreciated.

9:50 PM Update: When I said no updates today, I lied. Every Race Page has been updated and uploaded, along with the Overview page. I'm done now. For real. G'night, folks...

October 26, 2004

2004-10-26T20:46:00.000-04:00

Hate to post and run, but I've been working on the Race Pages for over two hours now (and yes, they're ALL updated, even with today's polls), and dinner awaits.

Still, several topics to cover today:

The Dick Cheney rally today was quite an awesome experience. After all, how many folks can say they were within feet of the Vice President of the USA. I must say he is quite more personable in person than the mainstream media makes him out to be, and a great orator to boot. Martinez didn't show, and neither did Congressman Crenshaw, who was said to possibly be there. I didn't quite get the seats I wanted - VVIP was apparently overbooked, so I was standing off to the side of the stage, but still in quite a good spot.

But enough about that. This isn't a personal blog, after all! I'm here to inform, and thus:

The Zogby polls from yesterday and today are part of Zog's Battleground Tracking Polls, and apparently he throws in a Senate question wherever there's a race in the ten battlegrounds. While battlegrounds such as Colorado and Florida have competitive Senate races as well, and this is a boon, other battlegrounds, such as Ohio and Nevada, have very uncompetitive races, and tracking polls such as this are quite useless. Still, we report every poll, and the tracking toplines will be posted here daily. Here's today's batch, in rough order of competitiveness. I'm listing only the topline and the state's designation. To get my brief analysis, you'll have to visit the Race Page. Without further ado:

Colorado: Salazar 50, Coors 44. Toss-up.
Florida: Martinez 47, Castor 47. Toss-up.
Wisconsin: Feingold 54, Michels 41. Probably Dem. retention.
Pennsylvania: Specter 53, Hoeffel 35. Safe Rep. retention.
Iowa: Grassley 65, Small 28. Safe Rep. retention.
Nevada: Reid 61, Ziser 31. Safe Dem. retention.
Ohio: Voinovich 55, Fingerhut 32. Safe Rep. retention.

Note this, because I'm only going to post this once: all Zogby tracking updates will be put up here.

More polls on the day (abridged-version analysis today):

In Florida, an SUSA poll shows Castor ahead by three, making the tally over the past week-and-a-half 2 with Castor ahead, two with Martinez ahead, and six with the race even. I don't know who to trust on this one, although, notably, SUSA's new poll shows five points of movement towards Castor, and Zogby's tracking finds three points movement towards Castor just yesterday. Interesting, but still Toss-up.

Another MRI poll of Louisiana shows Vitter just one point short of 50%, and shows Kennedy (with 17%) leading John (with 16%) for the second poll in a row. Otherwise, nothing new to report. Leaning Rep. takeover.

A SurveyUSA poll of North Carolina shows Burr leading Bowles 50-44. This is Burr's largest lead yet and five of the last six polls have shown him ahead. This is his race to lose now. Leaning Rep. takeover.

In South Carolina, Jim DeMint still appears to be cruising towards a victory in the Senate race, and the newest SUSA poll, showing him ahead 13 points, 52%-39%, confirms his current lead. Still, this race will stay Leaning Rep. takeover until another poll shows DeMint has reclaimed his previous double-digit lead. SUSA has had some whopping outliers in the past, and it will take another poll to show whether this latest poll or the one before it was the off one.

The newest Wilson Research tracking poll shows Coburn leading 41-38 in Oklahoma, constituting no movement at all in margin since their last poll. Still Leaning Rep. retention, and looking stronger by the day.

Just as I move Pennsylvania to a less safe designation, the polls seemingly contradict me. Four polls today: Zogby tracking shows Specter jumping ahead to an 18-point lead, 53-35, Mason-Dixon shows Specter ahead by the same margin, 50-32, and Keystone shows the race even wider at 52-29. SurveyUSA, in a poll released earlier today, shows the race closer at 51-38. Still, this is clearly no 7-10 point race, so I'm moving it back to Safe Rep. retention, where, barring major incident or injury, it will quite likely stay until Election Day.

In the Non-Competitive Races:

Even over a week since their last poll of the race, SUSA finds Lincoln's lead in Arkansas the same at 57-38. Safe Dem. retention.

In California, a new SUSA poll shows Boxer's lead at 15 points, 53-38 over Jones. While technically far below the Safe designation threshold, this close to the election, I'm inclined to leave this one at Safe Dem. retention.

Research 2000 finds little change in Indiana, as Bayh leads his challenger 65-31. Safe Dem. retention.

In Missouri, incumbent Sen. Bond's lead in the latest SUSA poll has increased significantly. He leads Nancy Farmer 60-34. Safe Rep. retention.

Two polls of Nevada. An older Las Vegas Sun poll shows Reid leading 63-29, while the Zogby tracking shows him ahead a mere 61-31. Neither, obviously, changes the designation from Safe Dem. retention.

Both SurveyUSA and Zogby shows little change in Ohio: Voinovich leads 55-32 in the Zogby tracking, 59-33 in the SUSA poll. Safe Rep. retention.

Seems a bunch of polls for the Non-Competitive Races, until one realizes that all the races being polled are in presidential battleground states. Pollsters can tack on a Senate horserace question with little added effort or cost to a poll of the presidential race in a state, and as a consequence, many do. See the Zogby tracking bit above.

All of the Race Pages have been updated, including even today's polls. I need to start compiling my final predictions, and will probably have those ready by Monday of next week. Ron at Politics1 has already given his predictions. If you're looking for predictions NOW, or looking to form your own, I suggest you check them out, but also keep in mind the polls of the races too...

Alas, forty minutes later, and now I am very hungered. Please send feedback. I will answer e-mail after dinner. For real this time. I still have e-mails from last week unread. Good night, all, and thanks for your patronage.

Today's Polls

2004-10-25T18:12:00.000-04:00

Note: I am starting typing at 5:13 PM. There are 18 polls in 14 races to add.

Hope everyone had a good weekend. I'm nearly over my cold and will be catching up today with updates and e-mail. I still have some unanswered e-mail from last week, so please be patient.

So, straight to it then. In order of competitiveness:

In Colorado, Zogby polls the race with Salazar leading 52-41. I have trouble believing this poll, though, as it also showed Kerry ahead of Bush by four points, when every other poll shows the President ahead a considerable margin. I think this is a definite outlier. Even so, the lead has flopped back and forth between polls in this race, and the call remains Toss-up.

In Florida, a couple of polls show this race clearly quite close. Zogby shows Martinez ahead 47-44, while a St. Pete Times/Miami Herald poll shows the race tied at 44. Out of seven polls in the last ten days, four have this race tied. Two have Martinez ahead, and one has Castor ahead, all within the margin of error. If I had to call this one now, I'd say Martinez would win by about 7 votes, but for now this race is a complete Toss-up.

Breaking from polls for a moment, Martinez will likely be appearing at a rally with Vice President Cheney tomorrow in Lake City, my hometown. I will be attending, and as an officer in the local Young Republicans, I'll be in VVIP seating, right behind the podium, on stage. I'm quite excited. But back to my non-partisan analytical self...

At last, a poll in the South Dakota race emerges. But, alas, it is a Republican poll from McLaughlin & Associates. It shows Thune ahead 49-45. Surely Tom Daschle must be quite frightened by now. Still Toss-up though, but another poll showing Thune ahead, this close to the election, will confirm my suspicions that Thune may just win this one.

In Louisiana, a poll from Renwick shows Vitter with a rather miniscule 39%, and Kennedy is actually leading John 18%-15% in the race for second. RCP has adopted a nifty system of showing, instead of Vitter's lead over his nearest Democratic opponent, how far he is from reaching the magical 50% in each poll. I like that idea, and will implement it in the Louisiana page. Regardless, this one stays Leaning Rep. takeover, although another poll showing Vitter this far from avoiding a runoff will change that.

In North Carolina, a couple of polls show Burr still ahead of Bowles. Rasmussen shows him leading by four, 49-45. The John Locke Foundation, another poll of origins unknown to me, shows him ahead 43-42. Regardless, this race remains Leaning Rep. takeover. Did I not predict Burr would eventually lead, even when, several months ago, he was ten points down? Kudos to Burr for exploiting Bowles' most glaring weakness - his connection to the Clinton White House, where he served as Chief of Staff.

In South Carolina, a Mason-Dixon poll shows DeMint leading Tenenbaum 47-43. This race is quite clearly still Leaning Rep. takeover, albeit by a slightly smaller margin than previously. Methinks DeMint will keep this lead until Election Day.

In Oklahoma, several polls show a tight race. A 10/20 Consumer Logic poll shows Carson leading by an incredible margin of 47-40, a far cry from the results of many other pollsters in the race, such as Rasmussen, who, in a poll released one day later, finds Coburn leading 48-43, and the latest Sooner Poll showing Coburn ahead 38-37. The last Sooner Poll had Carson ahead 2, but by a margin of 43-41, so both candidates' support levels have fallen since that poll. I have no knowledge of past Consumer Logic polls or the Tulsa World newpaper that commissioned it, and its odd topline does not jibe with other polls of the race. I'm tempted to throw it out as an outlier or biased poll. Four of the five polls this week, and seven of ten this month, have Coburn ahead, so, whether or not the Consumer Logic poll factors in, I'm leaving this as Leaning Rep. retention, albeit closer to Toss-up than previously.

In Kentucky, a Bluegrass poll shows Bunning ahead 49-43. This poll makes two major points: while (a) this race is most decidedly closer than it was three weeks ago, it is (b) not nearly as close as the Democrats would like you to think it is. Likely Rep. retention.

A Zogby poll of Wisconsin shows Feingold leading Michels 54-42. While this is clearly not a top-tier competitive race, it is neither wide open enough to call Safe. Thus, Probably Dem. retention.

In Illinois, a Chicago Tribune poll shows Obama leading 66-19, a movement of four whole points towards Keyes since their last poll. This race is quite clearly a complete Toss-up.

Just kidding. Still, quite unexcitedly, Easy Dem. takeover.

In Pennsylvania, Zogby shows an interesting result: Specter leading by ten, 48-38. I doubt quite highly this will be a problem race for Specter, but as he has been consistently polling less than 20 points ahead, this is moved to Probably Rep. retention, a purely cosmetic change at this point in the race, and I do not plan to create a discrete Race Page for this race.

Other Non-Competitive Races:

In Ohio, Zogby shows Voinovich leading 58-30, consistent with the race's designation of Safe Rep. retention.

In Nevada, Asst. Minority Leader Harry Reid should have no trouble holding his seat and poised to take the reins in case Daschle takes a fall, as Zogby shows him ahead 61-30. Safe Dem. retention.

And, finally, in Iowa, Zog shows incumbent Grassley cruising over his opponent 66-25. Safe Rep. retention.

Done at: 6:07 PM. 1 hour, 1300+ words later, and all the new polls are up. Still to do: update race pages, and I should probably start compiling my final predictions. The election is but a week away! 'Twill be done later tonight, as I presently have other demands on my time.

I'm off. Thanks to those who have sent feedback; I hope to reply to e-mail tonight or sooner. There don't seem to have been any new plugs, but my readership is continually increasing, so I can only assume I'm beginning to accumulate a regular readership. Several folks have already acknowledged such. Thanks to them especially. Have a lovely day, all.

9:49 PM Update: Well, still I have not done the Race Page updates. Will work on them soon, but not yet. Now, off to bed...

{Insert Title Here}

2004-10-24T18:53:00.000-04:00

I'm excruciatingly tired, but I think I'm about over my cold. In case you couldn't tell, I never did get the page updates or article done Friday night.

Due to severe constraints on my time, I've elected to forgo the now long-past-due article and simply do a massive prediction-fest article on Election Eve.

I also have several new polls and updates, but I'll do those tomorrow.

Good night, all.

T-Minus 10 Days...

2004-10-22T21:01:00.000-04:00

10 Days till Election Day. Just a few new polls tonight:

In North Carolina, Mason-Dixon shows the race tied at 45-45. While it doesn't exactly show Burr ahead like a majority of recent polls, it doesn't show Bowles ahead either, and is well within the MoE from the other polls. While this race is quite clearly very close, it's still Leaning Rep. takeover.

In Colorado, Ciruli polls the race as Salazar leading 47-43. Apparently I moved this race prematurely, and it heads back to Toss-up. This race will certainly go down to the wire.

And in Pennsylvania, Quinnipiac agrees with the other polls, showing Specter ahead 55-37. Safe Rep. Retention.

I'm really tired, and fighting a head cold, but I'll try and get all updates and the article done before I go to bed, as I'll be gone until Sunday afternoon. Thanks to everyone who's sent feedback; you are to be greatly rewarded. Pity to those who have not; you are to be greatly punished. Have a good night, all.

Evening SUSA polls

2004-10-21T21:57:00.001-04:00

A few new polls from SUSA show some promising results for Republicans:

In Colorado, they show the race as Coors 50%-Salazar 46%, a shift of 2 points each way from their last poll. This poll clearly reconciles the race's current designation of Leaning Rep. retention.

In Oklahoma, their poll shows Coburn leading 47-41, with independent Bilyeu garnering 8% of the vote, leaving only 4% undecided. This is a shift of 5 points in favor of Coburn since the last SUSA poll, and is his first lead outside the MoE. Discarding the outlying Wilson poll and the partisan Republican poll, four of the six October polls of this race have shown Coburn ahead. With great trepidation, I'm moving this race to Leaning Rep. retention, no matter what Denny Hastert says.

And in Georgia, the poll shows Isakson leading by 16 points, 55%-39%. Although there is no question as to the outcome of this race, it must, by the numbers, remain Probably Rep. takeover.

I'm dead tired, and will wait to update the pages and finish the article until tomorrow. That shouldn't stop you from sending feedback though - my Inbox is waiting 24/7 for your correspondence. See y'all tomorrow, and good night.

October 21, 2004

2004-10-21T14:07:00.000-04:00

Some new polls, including those referenced yesterday, but not analyzed:

In Alaska, the race is extremely close, yet pollsters have all but neglected it. The same company has produced all the polls of the race. They (KTUU-TV) just released a new poll, showing Knowles leading by the ever-so-slight margin of 46.9%-45.3%. Third-party candidates appear to be having no effect on the race, and a mere 4.2% are undecided. An anonymous reader makes a good point in that Murkowski's positives have shot up 7.5 points since the last poll. It will be interesting to see if this shift significantly affects the election results.

A number of reasons induce me to leave this race at Toss-up, where it may likely stay until Election Day. First, Knowles' lead is a razor-thin 1.6%. I have no information on the sample size of MoE of this poll, and even though Knowles has led in every KTUU poll so far, this new poll shows his lead from a month ago cut nearly in half, and at the lowest point of the campaign. Also, the jump in Murkowski's positives lead me to believe she can very easily close the small gap between her and Knowles, although it will be interesting to see just how much effect it will have in the limited time remaining before the election. Finally, undecideds usually break ever-so-slightly for the incumbent (or evenly) in Senate races (My Election Analysis has a great article on this fact here), meaning that, no matter how much Democrats whine about polls showing Knowles ahead, this race is still wide open.

In Florida, an older UNF poll shows Castor leading Martinez just 38%-35%, with a huge 22% undecided, perhaps because leaners were not pushed for an answer. Still, no other recent poll has had more than 9% undecided, and both candidates' support, as a result, is severely understated compared to the other recent polls - 10-12 points for Martinez, 7-9 for Castor. This poll is quite confusing and rather confounding, along with slightly outdated, so I'm not going to change designation. Still Leaning Rep. takeover. The Martinez sign is swaying nice and proud in the wind outside my window. 3:44 PM Update: See below. Now Toss-up.

In Louisiana, a Southern Media/Opinion Research poll shows more of the same: Vitter-43%, John-18%, and Kennedy-11%. It's likely going to be Vitter and John in a runoff, if one is to occur. I'm still wagering against it, and my prediction is slightly buoyed by the 26% undecided in this poll. Still Leaning Rep. takeover.

In Kentucky, we find SurveyUSA with a poll showing Bunning ahead 53%-39%, a stark contrast with the 43-43 Democratic poll from just a few days ago. Bunning's support, a full ten percent higher in the SUSA poll, is over two spans of the MoE higher than the Dem poll. It's possible that one of the polls is an outlier, in which case I would blame the GHY poll, or that one of the polls is cooked, which would also point suspiciously to the Democratic poll. Perhaps I was too overzealous in shifting this race last time. It goes back to the second tier and Likely Rep. retention, although I'm sure we'll see some more polls of this race soon.

In Washington, I've discovered that Strategic Vision has been polling the race for months as coming up with much closer numbers than other polls. On 9/24, they showed Murray ahead 48-41. On 10/7, they showed Murray ahead 49-41. And their most recently released poll, released just yesterday, showed Murray ahead by the same margin, 49-41. Either every other pollster is wrong, or SV's methodology is flawed in some particular way that causes the race to show up nearly ten points closer than other pollsters. Regardless, this lack of movement, mirrored in the other polls (just with a ten-point larger margin) leaves this race Probably Dem. retention.

And, briefly, some non-competitive race polls:

In California, polls continue to show the race widening for Senator Boxer. The LA Times shows Boxer ahead 55-33, while SV has her ahead by a slightly smaller margin, 53-36. Regardless, still Safe Dem. retention.

In Nevada, Harry Reid continues to face little challenge. A new Mason-Dixon poll shows him ahead by 27 points, 59-32. Safe Dem. retention.

And, finally, in Indiana, SurveyUSA poll shows incumbent Bayh cruising 63-31. Safe Dem. retention.

And those are today's polls. Gotta run, but I hope to have the article done today or tomorrow. Feedback is always a good thing, and if you've e-mailed me, I should be replying this afternoon.

3:43 PM Update: Quinnipiac has released a new poll showing the Florida race a tie at 47-47. Out of this week's polls, this is the third showing a tie, and the other two are split between the candidates. Florida looks indeed like a Toss-up now.

4:04 PM Update: Overview page updated. I have an hour in which to update the Race Pages and start on the article. Then I'm gone for the evening.

4:22 PM Update: In what took perhaps the least amount of time in history, I've already updated all the Competitive Race pages (AK,FL, KY, LA, and WA). All that's left now is the Non-Competitive race page, which should be up within ten minutes or so.

4:32 PM Update: And, right on schedule, the N-C Races page is updated.

I'm done for now. Off to compile the article. Enjoy your evening.

October 20, 2004

2004-10-20T16:04:00.001-04:00

A few new polls, but no time for updates. all are at RCP

In LA, more of the same: Vitter 43, John 18, Kennedy 11.
In CA, the LA Times shows Boxer 55, Jones 33.
In FL, UNF polls the race at Castor 38, Martinez 35, with a massive 24% undecided.

I have to run now to get some Martinez signs for my yard and some other errands (whoops! Did I just let some partisanship slip through?). Expect an update later tonight, possibly tomorrow.

10:14 PM: Sorry, no updates tonight. Too much stuff to do. Got the Martinez sign up, though. ;)

Out of school at 1 tomorrow, so updates are imminent, along with the now-past-due article. A few new polls to satiate the incensed crowd:

In AK, the KTUU tracking poll shows Knowles ahead 47-45. This will probably move to Lean Dem. when I get to updating.

In NV, Reid leads Ziser 59-32 in a M-D poll.

In KY, a SurveyUSA poll shows the race with Bunning ahead 53-39. This race is really, really confusing. Bunning failed to show up for the latest debate between the candidates, without any notice and he still has not, to my knowledge, explained why. Strange happenings, indeed...

A Few Extra Polls, Just For Fun

2004-10-19T21:21:00.000-04:00

SurveyUSA is really hitting the phones this week.

They polled the Missouri race and found Bond still ahead 56-38, a slightly smaller margin than the last poll, but still not nearly enough to move this race from Safe Rep. retention.

And, the third Ohio poll of the day from SUSA shows Voinovich ahead 58%-34%, his smallest margin of the polls today. For the third bloody time, Safe Rep. retention.

By the way, I've updated all of the Race Pages except the Non-Competitive page, which I will have done in a bit.

10:08 PM Update: Alright, everything's updated, and there's no new polls out in the interval between. I'm not going to work on the article tonight - even as a political junkie and computer geek, I've seen enough polls and HTML for one day. Good night all, and thanks for sticking with us.

A Veritable Ton, No?

2004-10-19T14:58:00.000-04:00

And the polls just keep pouring in. RCP has added five more new ones since last night, which brings the total number of races to update up to 12, over a third of all the contested races this year.

In subjective order of competitiveness:

In Colorado, two new polls out. The slightly more dated is a Rocky Mountain News poll showing Coors ahead 45%-40%. The second is a USA Today/Gallup poll showing Salazar ahead 49%-48%. These polls raise some interesting opportunities for punditry. The last RMN poll, from a month ago, had Salazar leading by 11 points. That's a 16-point shift towards Coors. The last Gallup poll had Salazar ahead by a similar margin of 11 points. That's a 10-point shift towards Coors. I have no trouble believing both previous polls were off to some (large) degree, but this still signals a major, major shift in the race. Combined with the SUSA and Mason-Dixon polls from a week or so ago, all four average out to nearly a tie, but the slightest Coors lead. Combining several factors: the immense shift in the polls, the average of the polls, and the fact that I'm trying to clear out most of the Toss-ups before Election Day, I'm going to go out on a limb here and say this is Leaning Rep. retention, but by the slightest of margins and subject to move at the next poll.

In North Carolina, a SurveyUSA poll shows the race close, Burr with 47% over Bowles with 45%. I have the utmost respect for SUSA and their methods, so the 3% movement since their last poll (which had Bowles up 1), though within the MoE, is still significant. Also, this poll corroborates closely with the Rasmussen poll from two weeks ago, signaling no recent movement in this race. Noting my comments from the last post on the race: "I think we're starting to see the movement towards Burr I've been predicting, but his lead is still small, and probably quite soft, so I'll wait for one more confirmatory poll before moving this one from Toss-up," I think it's time to move this one to Leaning Rep. takeover as well.

In Oklahoma, polls continue to pour in even as I type this post. The latest Wilson Research poll has Coburn ahead 42-39, a movement of five points from their last poll (which, after some consideration, appears to be an outlier by about, coincidentally, five or six points). ~~Rumors are also flying about a Mason-Dixon poll showing Coburn up 48%-43%, but these are as yet unsubstantiated.~~ Using only the Wilson Research poll, then, even trying to move races out of the Toss-up designation, I am loathe to move this one. ~~If the Mason-Dixon poll pans out, though, it may slide to ever-so-slightly Leaning Rep.~~ 5:29 PM Update: Thanks to the real KQQL, the Mason-Dixon rumor is not only unsubstantiated, but completely false.

In Florida, pollsters seem to be making up for their forced hiatus due to hurricanes. The second and third polls this week are both out. Mason-Dixon shows the race tied at 45-all, while SUSA shows Martinez ahead 49-47. Four of seven polls this month show Martinez ahead, and another two show a tie. Indeed, this race seems clearly Leaning Rep. takeover. Also, quick note regarding the Sayfiereview.com/Insider Advantage polls posted a few days ago. Apparently, they were for subscribers to the website only, and were not meant for publication. They will be promptly removed and not considered in the race's designation.

In California, a SurveyUSA poll shows Boxer leading with a commanding margin of 57%-35%, a 22-point lead, her largest of the year. Indeed, Bill Jones seems to have conceded this race, retracting his personal loan to his campaign. This race has teetered on the edge for a while, but based on this news and Boxer's 22-point lead, this race moves to Safe Dem. retention.

In Washington, a SurveyUSA poll shows incumbent Murray leading Nethercutt by a 56%-38% margin. Apparently Nethercutt's latest ad ("Different") has done little damage, as this lead is nearly the same it was two weeks ago. Still, it's not quite at 20 points for Murray, and so stays Probably Dem. retention, although I think with the ineffectiveness of Nethercutt's ad, this race has ceased to be competitive.

In Wisconsin, the NRSC has pulled $1.2 million in advertising for their Senate candidate, but a new Rasmussen poll doesn't jibe with this movement. It shows incumbent Feingold leading Michels by "a mere" ten points, 53%-43%. This is surprising, to say the least, but I'm leaving this race at Probably Dem. retention for now, where it will probably stay. Michels, even if "only" ten points behind, has too much ground to make up between now and Election Day.

Ah, yes. Kentucky. As I mentioned late last night, Gavin, Hart, & Yang, a Democratic pollster, presumably commissioned by the Mongiardo campaign, polled the Kentucky race at 43% for Senator Bunning, and 43% for challenger Dr. Dan Mongiardo. Surprising, to say the least, as Bunning was 17 points ahead just over a month ago. How can this be, you ask? Reference my post on the topic a few days ago. This is the first poll post-debate, and I think many of Kentucky's heavily Republican voters are beginning to think like Dirk Wahlstein. Presuming, of course, this is a real poll (I have no doubt it is - just throwing that possibility out), the MoE is 4.4%, meaning Bunning could be anywhere from 8.8 points ahead to 8.8 points behind, although the probabilities of such margins would fit a Normal model (good article on the MoE and its statistical significance here), meaning the most likely result is that the race really is tied. Accounting for the fact that this is, after all, still a partisan poll, I will move this race all the way to Leaning Rep. retention. Pending another poll, this could go either way. This will be a (if not THE) race to watch in the coming run-up to the Election.

Some less-competitive races:

In Arkansas, SurveyUSA finds Lincoln ahead a solid 19 points, 57%-38% over Holt. Closer, but not anywhere near close enough. Still Safe Dem. retention.

In New Hampshire, two polls show a similar conclusion. Research 2000 (no link - on Politics1) shows the race as 67-17 for incumbent Gregg. Suffolk U shows the race slightly ~~closer~~ (make that slightly less of a blowout), at 52-26. Safe Rep. retention.

In Ohio, an Ohio Poll shows Voinovich cruising 62-35. Safe Rep. retention.

And, finally, in Pennsylvania, SurveyUSA shows Specter ahead by just 7, 48-41 over Hoeffel. This very unusual, as it is the smallest lead Specter has had yet, and the first time Hoeffel has had more than 40% support. Once again, I have the greatest confidence in SurveyUSA and their methods, and looking at the internals of this poll, I see no reason to doubt this result, beyond the odd topline. Nearly all the movement has been within the moderates, with the bulk of the movement going to Hoeffel. Interestingly, although Constitution Party candidate Jim Clymer is getting 6% of the vote in this survey, Specter's support among conservatives has fell but one point since the last SUSA poll (when Clymer was not named, and Specter was ahead 19 points - then again, in that survey, "Other" got 9% of the vote, whereas now it gets 3%). However, the statistic that jumps out most is the low number of undecideds. Only 2% of voters in this poll consider themselves undecided. Hoeffel, then, will have to convince at least 5% of the populous currently supporting Specter (which works out to over 10% of Specter voters) to vote for him instead. I do not see him doing that, and on the strength of that alone, am leaving this race Safe Rep. retention.

Phew! That's all the polls for now. On to the wonderful chore of updating the race pages.

A few last notes: the IL SUSA poll listed last night was a false alarm, in fact it was an SUSA poll from 9/14 that I had missed. Regardless, it was Obama 65-Keyes 24, and it changes the designation of the race very little. Mega-Easy Dem. Takeover. Also, interestingly, out of the tons of polls added, we still don't have one of the Alaska race. The latest is nearly two weeks old now. The race is a Toss-up, and junkies such as ourselves are thirsting for polls of the race, yet the polling companies don't seem to care. Ah, well.

I'm due for another article, and I'm working on it as we speak, now that this post is through. Should be interesting with the number of new polls and changes of designation in the past coupla weeks. Here's the plan: article today or tomorrow, another article in a week, then an article of final predictions Nov. 1 or 2. I'll post Wednesday morning here on my win-loss record, and then this blog will likely die, unless Vitter fails to reach 50%. I'm taking suggestions on what I should do in the blogosphere after the election.

Plenty of thanks for mentions today (we're getting more and more exposure!): Country-Fried blog, Bitware.com Senate analysis page, Socialstudiesworld.com, as always, KivaCom, and I continue to take a load of hits from Daly Thoughts Thanks guys!

As always, send feedback is always appreciated.

4:29 PM Update: Post finished. Longest post in history. Enjoy.

4:34 PM Update: Cruising through some old posts at Daly Thoughts, I came upon a poll of the NY Senate race, dated 9/14, showing Schumer ahead 61-13. There are also tracking results for August, June, and April polls, all showing similar margins. Just in case you might have wanted to know...

5:26 PM Update: Still haven't got to updating the pages, but I'm starting on that now. A spanking-new Fox News poll shows Voinovich leading by an even bigger margin, 58%-28%. A Safe Rep. retention indeed.

8:10 PM Update: The Overview chart has been updated and uploaded. I hope to have the Race Pages done within the hour.

October 18, 2004

2004-10-18T21:33:00.000-04:00

SurveyUSA has a major load o' polls out today. They've polled the FL, NC, PA, AR, IL, WA, and CA races.

I also have a Research2000 poll of NH, Mason-Dixon and Insider Advantage polls of FL, and a Rasmussen poll of WI.

Too busy right now, I will update tomorrow. I also need to start working on the new article - only 15 days till the election.

A few thanks: first, to all who send feedback (I'll try to get caught up answering e-mail tomorrow as well), to Sean at My Election Analysis for plugging my blog, and Ron at Politics1, the RealClearPolitics team, and KQQL for continuing to feed me links to the latest polls.

10:13 PM Update: The Mongiardo campaign has just released a GHY poll showing the race in KY tied, 43-43. I usually don't move races based on partisan polls, but I will probably move this one at least one designation. This may be the sleeper race of the year, and you can be sure we'll be watching it closely.

Some polls for your weekend...

2004-10-16T14:55:00.000-04:00

I want to apologize for not posting yesterday, I was embroiled in the midst of some other major work, although few seemed to notice - my Inbox is as empty as ever! But enough of that. There is two days' worth of updates to do:

In South Dakota, the first poll in weeks shows the slightest movement towards incumbent Daschle. Rasmussen shows the race tied at 49. Their last poll, from 9/29, had Thune ahead by 4. Disregarding any statistical noise, this seems to show that most of the undecideds are leaning towards Daschle, and that a great majority of SD voters have already made up their minds (only 2% are undecided!). This doesn't bode well for Thune, but statistically speaking, either candidate could easily be ahead right now. This race remains a Toss-up.

In Oklahoma, a Sooner Poll shows Carson ever so slightly ahead, 43-41 over Coburn, seemingly supporting the results of the recent Wilson Research poll. This is the second non-partisan poll showing Carson ahead, after a string of polls with Coburn up. The lead in this race has varied widely and moved between the candidates a number of times, and the large number of undecideds (14% in this poll) means that both candidates have a similar chances to win. Still Toss-up, and the momentum factor is up in the air now too. This is clearly anybody's race.

At long last, we get a current non-partisan poll of Louisiana. Although outdated by almost a week, this poll corroborates the other polls of the race so far (even many partisan polls). It shows Vitter with 47%, John with 20%, and Kennedy with 14%. This Vitter's highest total yet, but still puts him just shy of the magical 50% mark. 16% of the voters, however, are still unaccounted for (3% say they will vote for Arthur Morrell), and it will interesting to see if Vitter can snag just 3 more percent of the vote and avoid a runoff. I'm wagering he can. Leaning Rep. takeover, possibly moving to likely if another poll shows Vitter with higher numbers.

In Florida, several polls are out, every one of them showing a tied race. Strategic Vision's latest shows the race tied at 47-47. InsiderAdvantage has polled the race twice in the past week, first coming up with a 43-43 tie (Oct. 10-11) and then a 42-42 tie (Oct 12-14), or so it's said. I can't find the polls anywhere on the web, except Politics1, and a tip from KQQL, who's never steered me wrong before. I have no problem believing the results, anyway, as they both confirm this race's designation of Toss-up. 6:36 PM Update: Whoops! I just realized Florida is currently Leaning Rep. ~~retention~~ takeover. Since two of the poll results are as yet unsubstantiated, and SV is technically a partisan poll, I suppose I'll leave it there until I find a write-up for the IA polls. 9:44 PM Update²: Very well, done Dirk. You caught my error. The Florida seat is held by Bob Graham, a Democrat, so a Republican win here would be a takeover, not a retention. Thanks.

In Georgia, there are a few new polls as well. Strategic Vision shows Isakson ahead 15, 54%-39%, while Zogby teams up with the Atlanta Journal-Constitution for a poll showing a slightly larger spread, Isakson +18, 53%-35%. Both these polls confirm the last Zogby and Rasmussen polls' showing of a under-20-point race, thus Probably Rep. takeover, although barring some major event, this should still be a cakewalk for Isakson.

And, finally, a Research2000 poll of the Vermont race shows incumbent Pat Leahy cruising 67%-21% over his no-name opponent McMullen. This poll, like the IA FL polls, I cannot find anywhere other than on Politics1. Interestingly, a Research2000 poll from the same time period, commissioned by the same TV station, is on RCP, but neither RCP nor the poll write-up mentions anything about the Senate race. Still, Politics1 is, in my consideration, a trustworthy and reputable source, and this poll does nothing if confirm Vermont's designation as Safe Dem. retention.

Feedback is a requirement if you view this site. If you do not send feedback, my goons will hunt you down! Thanks to TahoeHorn at the hornfans.com Forum for a plug and plenty of hits, and thanks always to Kivacom for their unending support of my bloggage.

8:00 PM Update: I finally fixed the Overview table with CSS style sheets. SiteMeter says 97% of y'all have browsers supporting style sheets, so I trust you should have no problem viewing it. It's up and running without a hitch now. Enjoy.

8:30 PM Update: Race pages are updated and uploaded.

Kentucky Race

2004-10-14T20:56:00.000-04:00

I've been reading up on some of the current events in the Kentucky Senate race. Now, normally I don't use this blog to publish news on the races (KivaCom does a perfectly good job of that), but I think this is significant.

Politics1 has a much better account of the goings-on in the race, but basically, incumbent Senator Jim Bunning is destroying himself. Particularly notable is his outrageous conduct during the recent debate between the candidates. After Bunning's campaign basically set the rules of the debate, Bunning didn't even show up, telecasting instead from his Washington office. It turns out now that he was using a teleprompter during the debate as well. He's also been making some, uh, "odd" comments on the campaign trail.

Of course, one cannot discount the campaign of Democratic state Senator Dr. Dan Mongiardo. Mongiardo was gaining in the polls before most of this happened (gaining eight points from the Sept. SUSA poll to the October poll). Mongiardo was viewed by many as a potential upset possibility, and even with Bunning outraising him by a 5:1 ratio, he has run a quiet, effective campaign. After corresponding with higher-ups in the Mongiardo campaign, I can say they are VERY optimistic about their candidate's chances, and rightfully so. Their own internal polling (caveat lector) shows Mongiardo has a 3:1 favorable/unfavorable ratio, and of the undecideds only 12% approve of Bunning's performance. The DSCC has noticed this too, and has poured nearly a hundred thousand dollars into the race.

Clearly, this race has become quite competitive, and I have built a race page for it and uploaded it. The question, though, is how competitive the race is. Even Mongiardo's own internal polls show Bunning ahead by eight points. This race is teetering between Likely and its current designation of Probably Rep. retention, and I could make the case for both, as I'm sure it's tightened further since the last SUSA poll, and Mongiardo undoubtedly has the momentum advantage, but until I see another poll of the race, I'll leave it where it is.

Thursday, October 14, 2004

2004-10-14T20:30:00.000-04:00

Several polls out today, many with significant implications:

I've taken flak from several directions about the South Carolina race. Apparently DeMint didn't perform well at the last debate and he continues to make numerous, uh, "politically incorrect" comments. Tenenbaum has been able to capitalize on these gaffes and has made up significant ground in the race. SurveyUSA 's latest poll shows DeMint ahead by only three points, 46-43. This race is clearly no longer a sure thing, but DeMint, ever so slightly, still has the upper hand, and, according to reports, did quite well in the second debate. This race moves to Leaning Rep. takeover, and becomes a first-tier Competitive Race.

In Oklahoma, a Republican poll from CHS & A shows Coburn ahead 46-41. As a partisan poll, I'm loath to move the race, but even combined with other recent polls of the race, it clearly shows this race as a complete Toss-up.

A Research 2000 poll of Indiana shows Bayh leading 63-27. Safe Dem. retention.

Any hopes Republicans have had for an upset in Arkansas have been thoroughly dashed. A Zogby poll of the state shows incumbent Lincoln with a commanding 28-point lead, 60-32. Safe Dem. retention.

Also, Gerry has posted linkage to the overview table I created and posted yesterday, and so far the reaction has been rather favorable. I'll polish it up a bit tonight, along with dealing with some other site issues.

Finally, KivaCom has mentioned us again, and a new referral source has come in from About Conservative Politics (a part of the About.com web conglomerate site). Much thanks to them.

Feedback appreciated, and helping to spread the word earns you a gold star! Thanks to all our visitors for helping this site stay alive.

Random Delusions...

2004-10-13T18:48:00.000-04:00

Three new polls on the day...

There has been a major lack of polls out of the Louisiana Senate race lately, and almost all of the polls have been partisan of one stripe or another. Republican pollster Public Opinion Strategies has polled the race as Vitter-43%, John-17%, and Kennedy-14%, not a major change from past polls. It still remains to be seen whether Vitter will break 50% on Nov. 2, and this race stays Leaning Rep. takeover until further notice. I suspect that as Election Day nears, there will be more non-partisan polls of this race, and if the race goes to a runoff, there will be a ton of polls in the month-long interval after every other major race has been decided.

In Kentucky, a Democratic poll by Garin-Hart-Yang (no link, on RCP) shows Bunning leading by a mere 8 points, 47-39 over Mongiardo. Although Bunning is clearly losing ground, I'm stubborn about these partisan polls, or perhaps I'm just lazy, but Kentucky remains Probably Rep. retention, and I'll wait a bit more to make a full page for the race.

In Pennsylvania, Strategic Vision (technically a Republican partisan pollster) polls the race at Specter 52% - Hoeffel 33%, a 19-point lead identical to several other recent polls. Still Safe Rep. retention.

Also, I have at last caught up with updating all the pages (even including this post!). Updates include the Georgia, Louisiana, Oklahoma, and South Carolina pages, along with the all-important Non-competitive Races page.

A few days back I mentioned getting an e-mail from a reader suggesting I create a table like that at Gerry's ECB listing the races and their designation. After a bit of work, I've drawn up a prototype, which I've posted here. I'm still working on it, though, so consider this Rev. 0.1, or similar to a beta testing prototype.

Looking at all the races in table format effects some interesting conclusions, too. As of now, Republicans are poised to takeover four currently Democratic seats (GA, SC, LA, and FL), while Democrats look to take over only one Republican-held seat (IL, of course). The five tossups include three Republican seats and two Democratic seats. To balance out the Republicans' four takeovers, Democrats would have to win ALL FIVE toss-up seats, and to regain control of the Senate Dems would have to win all five tossups, FL, and LA. Both situations seem near-impossible, as Republican candidates in the toss-up races of CO, NC, OK, and SD all have a slight advantage in momentum. The only toss-up that could be argued to be trending in a Democratic direction is Alaska, and even that is dubious, as Murkowski has very nearly closed the gap on Knowles.

The best the Democrats could hope for is for the Toss-ups to split evenly, with each party holding their seats, which would result in a +3 gain for Republicans. As I noted earlier, I see Republican trends in four of five toss-up races, however, and I think that those four races (or any combination of four from the toss-ups, actually) will end up going Republican.

Call me crazy, but I am amending my overall predictions:

Republicans win: GA, SC, FL, LA, NC, SD, CO, and OK.
Democrats win: IL and AK.

All other seats remain the same, and Republicans net +4 seats, for a 55-45 (44+1) majority. The Democrats face a very dire situation and have some major catching-up to do before Election Day.

And, finally, KivaCom, back after an absence, has given us a stunning review of the latest article. Thanks!

Please send feedback to me at the House of E-Mail, and/or leave comments.

The Fellowship of the Polls

2004-10-12T21:01:00.000-04:00

A veritable ton of Senate race polls out today (and yesterday, and most of last week - these are the ones I've seen today).

First, in Oklahoma, Wilson Research's weekly poll shows Carson ahead by 2, 40-38. Their last poll had him ahead by an identical margin. It is now quite clear that this race is a lot closer than past polls indicated, and neither candidate has a major advantage. I suspect this one will be very close down to the wire, and expect the White House to take a major interest in swinging it for Coburn in this deep red state. Still, assuredly, a Toss-up.

In South Carolina, a Rasmussen poll shows the race considerably closer than in the past, with DeMint leading by just six points, 49-43. The last non-partisan poll had him up by double that margin, 50-38. Note that almost all of the movement has been on Tenenbaum's side of the ticket. It's possible that in the last few weeks she has solidified her base. The first of a series of broadcasted debates between the candidates is tonight, and if Tenenbaum holds her own, we may yet see a race. Still, I don't think it's close enough to move the designation. Still Likely Rep. takeover.

In Georgia, there are two new polls, both showing the race significantly closer than before. Rasmussen pegs the race at 54-42 Isakson, while the Atlanta Journal-Constitution's slightly more recent poll shows Isakson up 49-33. While both of these polls show still significantly comfortable leads, I don't think this race is still an Easy takeover. Complacency or overconfidence in the Isakson campaign, perhaps? If there were only one poll showing a 12-15 point race, it would probably be an outlier. But two in a row, from two different sources, moves this race to Probably Rep. takeover.

In Pennsylvania, Quinnipiac gives us a poll showing the race maintaining the status quo, with Specter ahead 55-36, making that Keystone poll look more and more like an anomaly. Still Safe Rep. retention.

And, finally, a poll of the Arkansas race from Opinion Research shows incumbent Blanche Lincoln with a massive 60-32 lead over Jim Holt. This is clearly a Easy Dem. retention. That last 14-point spread result is quite likely an outlier.

Alrighty, there it is. All of today's polls. I'll work on the pages tomorrow. As always, feedback is much-appreciated. Props today to Kevin for catching my mistake on the North Dakota race - I had it listed as Easy Rep. retention, whereas it should be Easy Dem. retention, seeing as the incumbent is, after all, a Democrat (which, by the way, confuses me highly, as North Dakota is a very solid Republican state). A wonderful Tuesday evening to you all.

News of the Blog...

2004-10-11T18:21:00.000-04:00

I've updated and uploaded the following pages:

Colorado
Florida
Georgia
Illinois
North Carolina
Oklahoma

The only pages left are Alaska, California, and the Non-Competitive Races page. I have a meeting until 9 tonight, but I might be able to finish them up before bed tonight. If not, my tomorrow afternoon is free.

As always, feedback is welcomed. Thanks to everyone who stops by.

Update: Alaska is up, and apparently California didn't need updating at all. All that's left is the Non-Competitive page, and I can deal with that tomorrow. G'night, all.

October 11, 2004

2004-10-11T14:44:00.000-04:00

Hope everyone had a great weekend and are enjoying your Columbus Day. Happy Thanksgiving to our Canadian readers (although I have yet to see any).

In Oklahoma, a new Sooner Poll takes a stab at the post-debates (the Senate candidates debate, not the presidential debate) race standing and finds Coburn leading by the smallest of discernible margins, 39.8%-39.2%. A massive 21% of voters classify themselves as undecided, with just 22 days until the election. I believe Coburn has reversed the flow of this race, and it appears the RNC is paying attention, too, as VP Cheney has visited the state on his behalf, and George H.W. Bush is scheduled to campiagn for him later this month. This race is a complete Toss-up.

This race is the complete opposite: in New Hampshire, the first poll of the race (no link - blogged on Politics1) shows incumbent Judd Gregg (R) cruising to an Easy Rep. retention over her 80-some-odd-years-old opponent, Doris "Granny D" Haddock with nearly three times the votes, 62%-21%.

I'll be working on the race pages this afternoon and this evening (I mean it this time!), but I'm not making any promises.

Colorado Update

2004-10-10T13:08:00.000-04:00

A couple polls out of the Rockies this weekend:

First, Gallup shows Salazar ahead a wide margin, 54-43. I have a very hard time believing this margin, as it undercuts nearly every other poll result ever of this race (except the RMN polls - which likewise have yet to be explained)

Secondly, Mason-Dixon finds the race at a much more reasonable margin of Salazar 46-Coors 44.

The last several polls have varied rather widely, some showing Coors ahead, others showing Salazar ahead, many by statistically significant margins...I call this one, still, a Toss-up, although I'm not quite as certain which candidate has the momentum advantage anymore.

I have a bunch of homework in CS and Stat, but should be done by the afternoon, and hopefully I'll have updated the pages by tonight. Not that anyone really cares...

Feedback always welcome.

Colorado Update and Site News...

2004-10-09T13:34:00.000-04:00

After yesterday's update, SurveyUSA released yet more polls.

Only one was a Senate race poll, that of Colorado. The last SUSA poll had Coors up 5, this newest one has the race exactly tied at 48-48. That leaves only 4% of voters undecided, not a very large number for either candidate to work with. This poll confirms, then, what many have been saying for a while: this is one of the closest Senate races this year and will probably come down to the wire. This poll supports, quite clearly, the current designation of Toss-up, although Coors still has the slightest of momentum.

Anyone who has cared to notice will see I haven't updated the race pages in several days, and for that I apologize. Other demands on my time in the real world prevented much activity on the blog front for a bit. I'll be working to catch up this afternoon. Thanks again to all who have sent feedback; I try to answer every e-mail personally, so don't be shy!

Friday Evening...

2004-10-08T23:32:00.000-04:00

...and only one new poll. I presume the debate (more on that later...) overshadowed basically the entire rest of political news.

And it's not even a considerably interesting one either. SurveyUSA polls the Georgia race as being, surprise!, 57%-35% for Isakson, very, very similar to the last SUSA poll, which has Isakson ahead 58-33. Easy Rep. takeover.

The debate, you ask? Cut and paste from my comments at Daly Thoughts:

Although I don’t think Bush totally destroyed, he did much better tonight than in the first debate, and Kerry certainly didn’t live up to the expectations he set for himself in the first debate.

All in all, a Bush victory. I think he’s stopped the bleeding and new polls will probably show him with a 2-3 point edge.

There it is. Feel free to comment at will. I'll be working on the site some tomorrow - we may or may not see some polls. For now, it's been a long day, and a long week, and I'm going to bed. G'night, all...

Florida Update

2004-10-08T06:12:00.000-04:00

Quickly, before I leave the house, a Florida poll:

Mason-Dixon polls the Florida race as Martinez 46%-Castor 41%. If we discard the Strategic Vision poll as partisan, this is the third poll showing Martinez ahead, all from reputable companies and all LV polls. This is also his largest lead yet, and just slightly outside one MoE. I do believe there has been significant movement it this race, and it becomes Leaning Rep. takeover, as Martinez now has the upper hand.

Have a nice day, all...

Polls. Just... Polls.

2004-10-07T21:51:00.000-04:00

I was goign to do another edit to the last post, but there's three more polls from SurveyUSA, not all of them particularly interesting.

The first, though, is quite interesting. SUSA polls the Oklahoma race with Coburn leading 46-44, with independent Bilyeu siphoning 5% of the vote. SurveyUSA has had fairly consistent polling this season, and I have no reason to doubt this result. Carson's poll numbers have been slipping of late, and, despite the media fury over Coburn's occupational history, I've been expecting this type of movement for a while. While I think this race definitely has a Republican lean to it, I don't like to move races more than one designation at a time, and this is just one poll. A Toss-up that could soon break big for Coburn.

They also polled the Kentucky race, and, surprisingly, found Bunning ahead by a margin of a mere 11 points, 51% to 40%. While this is really nothing to be worried about for the Bunning campaign, it is still a major shift since SurveyUSA's last poll, a result of +19 for Bunning. This race moves to Probably Rep. retention, from its previous Safe designation, but I'll wait to create a discrete competitive race page for it until another poll confirms these results.

SurveyUSA also, inexplicably, polled the Illinois race, which makes for the third poll of this race today. Obama leads 68%-23%, a result uncannily similar to, well, every other poll of the race. Ever. Safe Dem. takeover. (<-- good catch Josh)

Thanks to Gerry for posting the presidential toplines at his blog, leading me to find these new polls. That's it for me tonight. I'll have a ton of updating to do tomorrow. As always, please alert me to new polls, comment on the posts, or send me feedback. Have a good night...

More polls for your evening...

2004-10-07T20:42:00.000-04:00

I realized I had already posted the SUSA Pennsylvania result last night, before today's post. Ah, well, it really matters not.

A few new polls:

Rasmussen checks in with a poll of the North Carolina race, and gets a result similar to the partisan poll from last post: Burr ahead by 2, 47%-45% over Bowles. I think we're starting to see the movement towards Burr I've been predicting, but his lead is still small, and probably quite soft, so I'll wait for one more confirmatory poll before moving this one from Toss-up.

Strategic Vision, historically a Republican pollster, releases the third poll in as many days of the Florida race, this one having Castor ahead 48-45 (thanks, KQQL). This race is clearly very fluid, and although Martinez has made up ground in the past few weeks, he still does not have a solid lead. Remains Toss-up. Interestingly, the same poll asked voters if they approved of the way the President responded to the hurricanes. A vast majority, 79%, said yes. Only 16% responded no.

In Alaska, KTUU-TV checks in with their monthly tracking poll. They have Knowles ahead, 48.1-45.2. Still, my comments from the article apply - I doubt Alaska is going to elect a Democrat. Still Toss-up.

The race pages will be updated tomorrow. I've also found that Strategic Vision has been doing a lot of polls that haven't been posted to RCP. More particularly, they've polled for several months in Pennsylvania, Florida, California, and Georgia. Those will be added to the pages as well, although I won't write up any new comments unless a change of designation results.

Thanks to all who have stopped by, and particularly those who've commented or e-mailed me. I'm always happy to get feedback, so please, be my guest. I'm off to do some real work now. In the meantime, enjoy your evening...

Update: And just when I thought I had it all covered, in waltzes another poll (Thanks Gerry of ECB). Gallup shows Salazar ahead of Coors 54%-43% in the Colorado race. This poll, particularly compared to other recent ones (most notably the SUSA one from just two days ago) absolutely screams outlier. Its questionable result in the presidential horserace (calling the race tied, when every other recent poll of CO shows the Prez ahead a solid five points) suggests the same. This race remains a Toss-up.

Would you like some polls with that?

2004-10-07T15:46:00.000-04:00

Polls today:

Quinnipiac polls the Florida race, and finds Martinez leading 48-47 among likely voters, but falling behind 44-40 among registered voters. I'll be using the LV result, making this the second straight poll showing Martinez ahead, albeit by the slimmest of margins. Still, considering the very slim margin, this one stays a Toss-up until another poll shows Martinez ahead.

A week-old partisan poll from "Am View" (I'm unfamiliar with the company) polls the North Carolina race as 47-45 in Burr's favor (no link - only recently posted on RCP). This is a partisan poll, is slightly out-dated, and no write-up is available, so it won't influence the call. Even if it could though, this race would still be a Toss-up.

A few polls from some non-competitive races:

In Pennsylvania, Keystone finds Specter ahead 44-35, a margin of only 9 points, way different from past polls of the race. SurveyUSA chimes in with a more recent poll corresponding with past results: Specter 54-35, a margin of 19 points. Interestingly, Hoeffel's numbers in the Keystone poll are similar to his numbers in other polls - Specter's support in the Keystone poll is a full ten points lower than in the SUSA poll. With the SUSA poll backing me up, I'm going to currently disregard this poll as an outlier and leave Pennsylvania Easy Rep. retention.

And just what we all need - more polls from Illinois! Research 2000 find Obama ahead 69-24, while Rasmussen finds Obama with a mere 64-20 lead. Keyes has gained six and seven points since the last poll! We may yet see a competitive race in IL. Or not. Easy Dem. takeover.

That ought to do it for now. I'll update the race pages later. Having trouble with my connection at the moment...barely squeezed this post through...

No rest for the weary...

2004-10-06T21:26:00.000-04:00

And, more SUSA polls are pouring out of the gates.

In Pennsylvania,SUSA pegs the race as solidly for Specter 54-35, a 19-point margin similar to past polls of this race. No change in designation - still Easy Rep. retention.

In Indiana, a similarly non-competitive race, they show the race as overwhelmingly for Bayh 66-28, a decisive 38-point margin that leaves this race with only one possible designation: Easy Dem. retention.

I'm almost done updating all the race pages with the article quotes and links. It will be nice to not have anything in particular to do.

I've noticed that SurveyUSA seems to be polling the Senate races a lot lately. Basically every state they've ran a poll in that had a Senate race going, they've polled the race, even in horrifically non-competitive races such as Indiana or Missouri. Any ideas why?

10/6/2004 Update Article

2004-10-06T18:57:00.000-04:00

Well, at long last, the 10/6 Article is up.

You can view it here in HTML.

It's also posted here on Free Republic.

I'll be updating the sidebar on the blog and pages with the links and posting the comments onto the pages in a bit...In the meantime, enjoy, and feel free to utilize any part of my work - just put in a little plug for the blog, OK?

Polling Update - 10/6

2004-10-06T15:53:00.000-04:00

SurveyUSA apparently released some more polls right after I posted yesterday's last analysis.

None of them are in majorly competitive races, though:

They find Boxer leading Jones 53-36 in California. This 17-point lead correlates quite nicely with other polls of the race, and the designation stays Probably Dem. retention, although I have little doubt Boxer will clean up.

Their poll of the Washington race shows Murray ahead 57-38, a margin seven points wider than the last SUSA poll. Apparently Nethercutt's new ad using Murray's Osama comment has not made much headway in this liberal state. This race becomes second-tier competitive, at Probably Dem. retention.

And, finally, SUSA polled the Missouri race and finds Bond ahead 59-34, a ten-point larger margin than the last poll of the race, clearly solidifying this race as Easy Rep. retention.

I'm working on the article and there is a very good chance it will be done today. And about bloody time, too!

And it continues...

2004-10-05T18:24:00.000-04:00

And immediately after posting the last post, I am alerted to find that SurveyUSA has a few new Senate polls out:

In Ohio, they show the race as, not surprisingly, Voinovich 58%-Fingerhut 34%. This margin is but two points larger than the last poll of the race, and it doesn't make the slightest inkling of difference in the designation. Easy Rep. retention.

In the North Carolina race, Bowles is still ahead 47-46. This makes very little room for error for either candidate, and is the third poll in a row with Bowles ahead only one point. This means the race could go either way before election day, and on the strength of recent movement towards Burr, becomes a Toss-up, although if the election were held today, Bowles would win by the slightest of margins.

Polls, polls, polls, polls...

2004-10-05T17:59:00.000-04:00

Several polls today.

In Oklahoma, Wilson Research shows the race tightening, with Carson ahead 43-41. That's a three-point narrowing from the last Wilson Research poll. Nonetheless, it's the eighth straight non-partisan poll showing Carson ahead or tied, and, as such, this race remains Leaning Dem. takeover, although I expect the next poll to be similar or closer and this race to become a toss-up.

In Wisconsin, an older (9/27) Harris poll shows little movement in the race, pegging Feingold ahead 52-39. This is the second straight poll showing Feingold ahead more than ten points, and as such this race becomes Probably Dem. retention, although it will be interesting to see if Michels, according to many an "aggressive campaigner," can narrow the gap.

In South Carolina, a Republican poll a few days old (10/1 - no link, on RCP) shows DeMint ahead, not unexpectedly, 46-36. Still Likely Rep. takeover.

Non-Competitive Races

2004-10-04T16:24:00.001-04:00

The Non-Competitive Races page is up, if you're into that kind of thing.

While compiling the Non-Comp. Races page, I also came across several recent polls of these races that were not reported:

In Connecticut, a 9/29 Quinnipiac poll shows the race as Dodd 69% - Orchulli 24%. Easy Dem. retention.

In Nevada, a 9/16 Mason-Dixon poll pegs the race as Reid 54% - Ziser 34%. Easy Dem. retention.

In Ohio, a 10/2 Columbus Dispatch poll has the race Voinovich 56% - Fingerhut 34%. Easy Rep. retention.

In Oregon, a Research2000 poll shows the race as Wyden 58% - King 31%. Easy Dem. retention.

In Pennsylvania, a 9/16 Keystone poll shows Specter ahead 51-25, and a 9/29 Mason-Dixon poll shows him ahead 53-31. Clearly Easy Rep. retention.

And finally, in Utah, a 9/30 Salt Lake Tribune poll shows Bennett leading 62-19, decisively Easy Rep. retention.

Enjoy...

Florida Update, Etc...

2004-10-04T16:24:00.000-04:00

The first poll of this race in almost two weeks shows the results I've been expecting all along. SurveyUSA shows Martinez ahead 50-46 with a mere 1% undecided. While I am still slightly wary of any Florida polling results, even if this poll is an outlier or slightly off, it is still indicative of some movement in this race. The long period in between this and the last poll (a Gallup) only strengthens this perception. This race is now a complete Toss-up.

This is the only new poll for the competitive races. Another article is due now whether I like it or not, and I still haven't worked on the Non-Competitive Races page. Luckily, I have most of this afternoon and this evening off...

Also, it appears the ad-free period on my FreeServers account has ran out. I apologize for the ads, but hey, beggars can't be choosy...

And, finally, KivaCom has given us another mention. Thanks to them and anyone stopping by. Spread the word!

Weekend Polls

2004-10-03T17:30:00.000-04:00

I'm back, and I find several polls and pages need updating. The polls are:

In South Dakota, a Zogby poll shows Daschle up 48-46, while a slightly newer Rasmussen poll shows Thune ahead 50-46. The Rasmussen result is almost exactly the same as their result from nearly a month ago, so I think it's safe to assume there has been very little movement in this race. This is a statistical Toss-up. It will be interesting to see if the President makes some stops in SD, as surely this race is high on the White House's priority list.

A Sooner Poll in Oklahoma polls the race as 44-38 Carson, a result 4 points higher for Carson than their last result. I think Coburn is having some trouble pulling out of this hole he seems to be in, and this race stays Leaning Dem. takeover.

And, finally, a Mason-Dixon poll of the South Carolina race has it pegged as DeMint 50-38 over Tenenbaum, the same margin as the last nonpartisan poll of the race. This surprises me not in the least, and this race remains Likely Rep. takeover.

As always, please alert me to any new polls. Thanks to KQQL for the SD and SC results.

Quickly, a few polls...

2004-10-01T14:23:00.000-04:00

Before I run off to pack, there's a few new polls out, both in the Carolinas.

Mason-Dixon polls the NC race as Bowles 45-Burr 44. This is the exact same result as the last poll (which was a Republican partisan poll). While there has clearly been some movement, Bowles is still technically ahead, so this race yet remains Leaning Dem. retention.

In South Carolina, a Democratic partisan poll from GSG (listed on RCP) shows Tenenbaum up 46-43. This is way out-of-whack with other polls of the race. Assuming this poll is at the very bottom end of the MoE for DeMint, and the last poll (SUSA) was at the very top end, it is still at the very edges of being statistical plausible. I doubt highly these results, and the designation changes not. Likely Rep. takeover.

I will update the Race Pages on my return. Have a lovely weekend.

Blog News

2004-10-01T14:20:00.000-04:00

I still haven't done the analysis article, and I'll be out of town (and offline) all weekend. I also expect a lot of polls tonight and over the weekend, so I'll try to catch up Monday. By then it will have been two weeks since the last article anyway.

In the meantime, how about the debate last night. Not Senate-related, but political business still. I'll say what I've been saying elsewhere: "I thought it came out pretty even.

"Even in that both candidates bored the hell out of me. I heard the same thing from both sides over, and over, and over, and over.

"Kerry needed a really strong performance to stay in contention in this race. The fact that he didn't outright lose means he's still surviving. Bush missed a chance to put him away for good last night.

"Still, Kerry needs a majorly impressive performance in the next two debates to have any chance at winning."

So there it is. Anyone want to comment, have at it.

New polls

2004-09-30T19:30:00.000-04:00

A couple new polls -

Jim Holt is polling strangely close to Blanche Lincoln in a new SurveyUSA poll. Lincoln leads 54-40. Although polls of this race are hard to come by, I'll wait for one more showing a margin of less than 20 points before moving this one from Easy Dem. retention.

SurveyUSA also did a poll of the NC Senate race asking voters who they thought won the 9/27 debates. The results were very close, with Bowles' 37% to Burr's 33%. 25% thought there was no clear winner. This poll, though interesting, doesn't affect this race's designation of Leaning Dem. retention.

Odds and Ends...

2004-09-30T16:48:00.000-04:00

I'm sure you know where I'm going with this, but I have a ton of stuff to do tonight and will likely be working until dinner. After that, I should start on the new article, but it likely won't be posted until late tonight or, God forbid, very early morning tomorrow.

Also, KivaCom, one of my main sources for news of the races and on the blogroll at right, has given us a mention. Although I whined earlier about having little recognition, I never expected success (even minor) this quick (this blog is exactly 7 days old). Ah, the power of the blogosphere. Thanks to those stopping by, and I hope you like what you see - feel free to comment...

And, finally, only one new poll out - another Ohio poll showing Voinovich up 64-34. Clearly Easy Rep. retention. No link, blogged on Politics1.

Update: A quick googling brought up this link, which gives little information, but some stats about the poll.

Race Pages, Reloaded

2004-09-29T17:36:00.000-04:00

The last Competitive Race Pages are up:

Oklahoma
South Carolina
South Dakota
Washington
Wisconsin

I have some other things to do tonight, but I should have the Non-Competitive Races page up tomorrow, and, as I said earlier, the next article up tomorrow.

New Article Soon...

2004-09-29T13:47:00.000-04:00

While I have a ton of other things to do, I've set a deadline for myself to have the latest article done by at most tomorrow night.

I will be working on the race pages this afternoon and should have those done by tonight.

Thanks again to all who've stopped by and welcome to those who are currently stopping by.

Update: In the meantime, I've converted my two previous articles to HTML format and uploaded them to the server. They can now be linked to from this blog and any race page. Here they are:

September 5, 2004 Article - the one that started it all.
September 20, 2004 Article

North Carolina Poll

2004-09-29T13:32:00.000-04:00

After the weekend news cycle, the influx of polling has tapered off a bit.

Only one new poll: a Republican poll from POS (no link - listed on RCP) showing Bowles up a mere 1 point - 45-44 over Burr.

No sample size is given, and, again, this is a partisan poll. As such, NC remains Leaning Dem. retention until another poll.

Alright! Some recognition!

2004-09-28T21:51:00.000-04:00

Thanks to Gerry for the mention.

For the record, ECB2004 is the best race tracking blog on the web. Be there, or be square.

Race pages

2004-09-28T19:04:00.000-04:00

I've been working for the last several hours on web pages summarizing each Senate race and archiving my commentary, almost exactly in the fashion of Gerry's ECB. I'll be uploading them to a Freeservers.com webserver, and linking to them in the sidebar.

The first batch is up:
Alaska
California
Colorado
Florida
Georgia
Illinois
Louisiana
North Carolina

Sure, the colors don't quite match, but I'll fix that later. I should have the rest up tomorrow. Enjoy!

New polls

2004-09-28T13:56:00.000-04:00

Two new polls out today.

A Democratic poll from GQR (no link - listed on RCP) shows Vitter with 44, Kennedy with 18, and John with 24, not much change from the last LA poll. If the Democrats' votes are combined, they still trail Vitter 44-42. This race remains Leaning Rep. takeover. <--thanks for noticing that.

A poll from Wilson Research (link shows Carson up 44-39. I continue to be surprised by the poll numbers out of this race, as Oklahoma is a stauch Republican stronghold, but I suppose based on the polls, this race is a Slight Dem. takeover.

Random bloggage

2004-09-27T16:41:00.000-04:00

There has been lots of changes in the past week and I'm planning to do a new summary article for FR and revise my predictions a bit. I should be working on it tonight.

In other blog news, in case you can't tell, this blog is still very, very new, and it gets kind of frustrating when I don't think anyone reading what I've wrote (no matter how worthless my opinion may be...) Please check in, even if you're just passing through. Anyone can leave a comment on my posts.

Also, if you have your own blog, please put me on your blogroll. I'm working on building up mine and I'd be happy to put anyone's blog on it. Just leave a comment or e-mail me (here would be good), and, presuming your blog is of adequate substance, I'll add it.

And thanks to anyone who stops by.

NRO weighs in

2004-09-27T16:30:00.000-04:00

John J. Miller publishes his monthly analysis of the Senate races.

He seems to be a bit more conservative on his predictions than yours truly, listing four races as toss-ups and predicting a Republican gain of a mere two seats. Still, good stuff. Enjoy.

New polls

2004-09-27T16:05:00.000-04:00

I apologize for being out for the weekend (although it doesn't appear many folks view this blog in the first place), I was out electricity and Internet until about two hours ago due to Jeanne.

There have been several new polls out in my absence.

Mason-Dixon, polling for the liberal Sioux Falls Argus-Leader, has released a poll showing Daschle up 50-45. The poll also shows Thune's and Daschle's unfavorable ratings almost equal (at 34 and 37 percent, respectively). The poll also shows 18% of Republicans crossing party lines to vote for Daschle, while only 6% of Democrats polled for Thune, which contradicts the rhetoric from Thune's campaign claiming a large number of Democrats in his camp. This race slides back to Toss-up.

Research2000 (link) shows Bowles up 49-40 in North Carolina, little change from the last poll out of the state (a SurveyUSA showing Bowles up by 10). Republican candidate Richard Burr is just now beginning to spend money on television ads, and the latest polls polls out of the state showed Bush up 4 and 13 points. Although I think it will be closer than the polls show, this race still Leaning Dem. retention.

A Democratic poll out of Alaska (no link - but listed on RCP here) shows Knowles up 47-41. I am loath to change the race's designation based on a partisan poll, but I do think Knowles is slightly ahead at the moment. Against my better judgment, moved to Toss-up status.

SurveyUSA puts Coors ahead in Colorado 51-46 (link) 51-46, just outside one span of the MoE. There's clearly some movement towards Coors in this race (two of the last three polls have shown him ahead - RCP), but I'll leave it at Toss-up until one more poll shows him ahead.

A Badger Poll puts Feingold ahead 53-38 in Wisconsin, a figure identical to Isakson's lead in Georgia and Boxer's lead in California. I'm going to spot this race about halfway between this poll and the last one. Barring any major mishaps, I think Feingold will probably triumph by about ten points, a very Likely Dem. retention.

I do believe that's all for now.

Another Oklahoma Poll

2004-09-25T18:39:00.000-04:00

The latest Sooner Poll has Carson up 39.8% to 37%. Note, though, that this poll did not include liberal independent candidate Bilyeu, who has been drawing support from Carson, and has a large number of undecideds (23.3%, presumably, though, including Bilyeu voters).

Although I'm still not entirely convinced, the OK race becomes a Toss-up. Just too derned close to call. For now.

A Second Opinion?

2004-09-25T18:37:00.000-04:00

Fred Barnes has an article up that reads almost exactly like what I've been saying.

The coattail effect in NC, OK, SD, CO, and LA could be a major factor in these elections, and cannot be understated.

Modus Operandi

2004-09-25T18:00:00.000-04:00

I recently had my designation of the Oklahoma race as slight Rep. retention called into question.

Let me explain exactly what my designations here mean: I am not, like many of the other race jockeys out there, blindly classifying races and states based on the newest polls, effectively showing how the election would turn out if it were held that very day.

Only one problem: The election will be held on November 2nd, and that's not going to change. My designations are predicting how the race will turn out on November 2nd.

Take, for instance, the Oklahoma designation. I'm predicting a close Coburn win, (as is noted in my most recent FR article) because the movement in the polls up to this point have been decidedly towards Coburn. The last five polls have gone Carson +2, Carson +2, Carson +7, Carson +1, push. Also taken into account is the fact that this election is in, of all places, Oklahoma. Judging solely by the President's 30-point lead in the polls, we can clearly tell that Oklahoma is a very conservative, Republican state. Coburn is a conservative Republican, Carson is a liberal Democrat.

Also, up until this week, Coburn was cruising, while Carson's campaign was hitting roadbumps at every opportunity. After this week, though, things may change. I'm waiting for one more poll to see if Coburn has lost any support after the media frenzy this week over this latest crisis.

So, with that in mind, don't be surprised if I see the races a bit different from other folks, just take it for what it's worth.

South Dakota Poll

2004-09-25T10:17:00.000-04:00

Sure it's a partisan poll (link) but it corroborates perfectly with previous polls, showing Thune up 50-47.

This race remains Slight Rep. takeover.

The Easy Races

2004-09-24T19:55:00.000-04:00

A Strategic Vision poll (link) shows Boxer ahead in California 52-37. There's little question she will win, but this poll is just this side of "Easy" designation. This race remains Probably Dem. retention.

In Iowa, a Research2000 poll (link) shows Chuck Grassley ahead 66-15, leaving this quite clearly Easy Rep. retention

Another Florida Poll

2004-09-24T19:50:00.000-04:00

Another Florida poll from Quinnipiac (link) showing Castor up a mere 1, 43-42.

This is very nearly a toss-up, but will remain Slight Dem. retention until at least one poll shows Martinez up slightly.

Florida Update

2004-09-23T22:04:00.000-04:00

A new Gallup poll (link) shows Castor up, surprisingly, 51-45 over Martinez.

That makes the second poll at or above one span of the MoE, warranting a change of the Florida race to Slight Dem. retention.

South Carolina Poll

2004-09-23T21:59:00.000-04:00

A new SurveyUSA SC poll (link) has DeMint up 12 over Tenenbaum, 51-39, reinforcing the designation of Likely Rep. takeover.

Washington Update

2004-09-23T21:53:00.000-04:00

Two new polls for the Washington race.

The first, a Seattle Times/Elway poll (link) has Murray up 57-37 among RVs.

And a SurveyUSA poll puts Murray ahead of Nethercutt 53-41 among LVs.

No major surprises here, but the 20-point lead seems a bit high. I'm more inclined to trust the more reputable pollster's LV poll. Designation remains Likely Dem. retention.

Oklahoma Poll

2004-09-23T21:16:00.000-04:00

Another poll in Oklahoma (link) showing movement towards Coburn, just like yours truly predicted, no?

This race will stay at Slight Rep. retention for now, and you can bet we'll be watching it closely.

Current Race Standings

2004-09-23T20:52:00.000-04:00

For those too lazy to browse over to FR to read my analysis, as of right now, the Senate races stand as follows. However, for the actual analysis behind the categorizing, you'll have to read the article.

Open Seats - Republicans listed first

Georgia (Miller) - Isakson vs. Majette - Easy Rep. takeover
South Carolina (Hollings) - DeMint vs. Tenenbaum - Likely Rep. takeover
Louisiana (Breaux) - Vitter vs. Kennedy and/or John - Leaning Rep. takeover
Oklahoma (Nickles) - Coburn vs. Carson - Slight Rep. retention
Florida (Graham) - Martinez vs. Castor - Toss-up
Colorado (Campbell) - Coors vs. Salazar - Toss-up
North Carolina (Edwards) - Burr vs. Bowles - Leaning Dem. retention
Illinois (Fitzgerald) - Keyes vs. Obama - Easy Dem. takeover

Republican-held Seats - Incumbent listed first

Alaska - Murkowski vs. Knowles - Leaning Rep. retention

Alabama (Shelby), Arizona (McCain), Idaho (Crapo), Iowa (Grassley), Kansas (Brownback), Kentucky (Bunning), Missouri (Bond), New Hampshire (Gregg), Ohio (Voinovich), Pennsylvania (Specter), and Utah (Bennett) are all Easy Rep. retention.

Democrat-held Seats - Incumbent listed first

South Dakota - Daschle vs. Thune - Leaning Rep. takeover
Washington - Murray vs. Nethercutt - Likely Dem. retention
Wisconsin - Feingold vs. Michels - Likely Dem. retention
California - Boxer vs. Jones - Probably Dem. retention

Arkansas (Lincoln), Connecticut (Dodd), Hawaii (Inouye), Indiana (Bayh), Maryland (Mikulski), North Dakota (Dorgan), New York (Schumer), Oregon (Wyden), and Vermont (Leahy) are all Easy Dem. retention

So, if the races play out exactly as they are today, and the parties split the toss-up seats, the Republicans will net +3 seats and the Senate will be 54-46.

Colorado Update

2004-09-23T19:14:00.000-04:00

Two new polls are out for the Colorado race.

One, listed only on RCP, a partisan Republican poll, shows Salazar up 2 points.

The other, a Ciruli Assoc. poll (link) shows Salazar up one.

Excluding the Rocky Mountain News poll, all six polls of this race since August have been within one span of the MoE. This is quite clearly a complete Toss-up.

Wisconsin Update

2004-09-23T17:08:00.000-04:00

The latest Wisconsin poll from ABC News (link), along with showing the President ahead a massive 10 points, shows incumbent Senator Russ Feingold leading Tim Michels by a mere 6 points, 51-45.

And this after I had very nearly written off this race as an easy retention for Feingold. Reclassified as Likely Dem. retention until further polls can clarify this strange turn of events.

Post Number One

2004-09-23T16:55:00.000-04:00

Welcome all to my Senate Race 2004 weblog. Over the past month or so, I've been publishing my analyses of the major Senate races through FreeRepublic, and I've been meaning to start my own political blog for ages.

Senate Analysis Sept. 20
Senate Analysis Sept. 9

I intend to update this blog as much as possible with new polls and news from the Senate races. With any luck, I will expand into House races as the election nears.

For those of you looking for presidential race and electoral college analyses, I highly recommend Gerry's ECB2004.