My first post on analyzing the House races was two weeks ago and remains the basis for my final post on predictions today. As is my second post I did last week. In order to save a little time in the writing of this, I am just going to note in this post if any race has changed significantly from these previous posts. I am still only going to review the top RCP House races that look like a risk to Dems or Reps, not all 50 RCP races!

Two week’s ago I predicted the Dems would pick up 6-9 seats. Last week I predicted the Dems would lose one seat in addition to the Reps looking better in one seat and modified the prediction to 5-8 seats. This is the final prediction, and sadly there are not a lot of new polls out at RPC today as I had hoped. But I need to get ready to take our two first graders out to dinner in celebration of their good grades! So onward we go.

I will only be applying the bias correction in heavily Rep districts, and possibly some 50-50 races. But I think districts that went for Kerry will see sufficient Dem voting to throw out and ‘cleanse’ their district. Liberals are mad, no two ways about it. So I see Kerry districts turning bluer, no matter what the GOTV effort can produce. One final change, I am will dispense noting the change in the RCP rankings – too many changes to keep track of. I will list the races in their current RCP rankings, making the top races in any category the least likely to stay in the category from the Rep point of view.

Probable Rep

NY-24: Open (Boehlert) – I have no idea why RCP put this race in the high risk, likely to go Dem category. There are no recent polls to make this judgement and the district is strong Bush country. Going out on a limb here – there is no wave. My baseline is the political map is roughly the same as 2004 (high turn out). Stays Rep. If I am right, we will see it here tomorrow night.

NY-20: Sweeney – This is another race RCP dumped into their endangered Rep list for no good reason. The RT Strategy/CD Polls are showing a clear, consistent bias to the left. And they use unproven voice recognition polling. If I dump these I get an average Dem lead of less than 3% in a district that went for Bush. So the bias correction brings this to a bit over 2% for the Rep. A good GOTV effort and this one stays Rep.

IN-2: (Chocola – R) – Some polls show a drift towards Chocola. The average of the last three polls is around 5% and the bias correction would put it in the Toss Up category. But that Reuters Zogby poll looks crazy next to the other 4 polls, which have an average of 3.75% lead for the dems. I think Chocola will pull this out with a 2% win.

FL-13: (Open – R) – I dealt with this one in a previous post – stays Rep..

NM-1: (Wilson – R) – This race moved from Probable Dem to Toss Up and now is moving into Probable Rep (following its fall in the RCP ranking). Taking the last three polls and averaging brings it down to 1.3% Dem lead. The trend is towards Wilson and the Survey USA poll shows a lots of momentum. Wilson hangs on.

27. OH-1: (Chabot – R) – Few polls, all outdated but trending right. Trend looks to be with Reps. If the Dem wave fizzles it will dry up in this race. Stays Rep

CT-4: (Shays – R) – I almost moved this into the Rep column, but the inflated generic number stayed my hand. And that is really what is different this time around. Strong turnout numbers and the evaporating generic ballot leads. The average of the last three polls puts this a tie. And we should apply a small bias correction due to lack of polls right now. Shays takes this one with 3-4%.

IN-9: (Sodrel – R) – This one won’t even be close. Ignoring the obviously tilted RT Strategies poll and conservatively using the other four recent polls shows a dem lead of 2%. Applying the bias correction I think Sodrel wins with over a 3% margin

IL-6: (Open – R) – As with the others, there is one outlier that is tilting the average and needs to be deweighted. But even if you include it with the other four and average the lead is a measely 1.6%. Using the bias correction for this heavily Rep district gives Roskam a comfortable win with a margin that should exceed 3%

AZ-5: Hayworth – Another RCP drop in race. Many of the top 30 races moved out of contention, so RCP dropped some more in on some last minute polls. This race is averaging a tie and requires a 5% poll bias adjustment I would assume. I think Hayworth pulls this out by a coule of points.

CA-11: Pombo – Another new race to replace all those races that drifted right and out of reach of the Dems. We have one poll in a solid district. RCP is grasping here. Stays Rep

OH-2: (Schmidt – R) – Another new race (see a trend here) where the last two polls show an average of a 1% lead for the rep. If it wasn’t for the one Survey USA poll there wouldn’t even be a case to make this race a risk – unless the generic ballot numbers showed a massive Dem wave coming. But since there is no wave coming, then I would say this race stays Rep.

MN-6: (Open – R) – This one has gone beyond the Dems grasp. This use to be in the top 20 RCP rankings – not anymore. There is a sign here, for those willing to look.

22. CT-2: (Simmons – R) – This one appears to be sliding away from the Dems even in the RCP rankings.

WA-8: (Reichert – R) – I could put this in the toss up category, if there was a big Dem generic ballot lead out there to push me to. There is not one and even RCP has moved this down their risk list.

24. VA-2: (Drake – R) – This race should never have been added to the RCP risk list, but it was. Why? One outlier poll and the big Dem generic lead which made all races without complete consistency all of a sudden a risk. The latest polls show why this was never in play, but hindsight without the fear of a large dem lead in the generic allows for more clear thinking.

Toss Up

CO-7: (Open – R) – This race I held in the Probable Rep for a while because the polls seem to be all over the place. The last four polls give an average of 8.74%. If I ignore the latest one (yes, not a good thing to do) the average drops to just above 6%. This is still a strong BNush district, so the bias factor can be applied making this a 1-3% lead for the Dem, something a GOTV effort could overcome. So I am moving this to toss up with the expectation it could easily go Dem.

IN-8: (Hostettler – R) – The first round on this I said this race was a lost cause, then a poll comes out showing Hostettler only down by 7. Since then an RT Strategies poll has come out that seriously deviates from the numbers this district showed for Bush in 2004. I am not seeing that much change in the electorate, especially with the generic polls out yesterday that put the lie to one sided turnout models. I am loathed to keep this in the Toss Up category because the polls could be much tighter now. It is at least a toss up.

OH-18: (Open – R) -I am keeping this in the Toss Up category with reservations. Just because Bob Ney is a thieving scumbag doesn’t make the Rep running a crook (otherwise all dems wouldn’t be allowed near young female interns). Most voters are much more sophisticated than blaming a party for one man’s digressions. Plus the polls (minus a late and wild Zogby poll) show a 8% lead for the Dems, which could be as low as 3% if the bias correction is applied in this strong Bush district. We shall see what kind of people live in OH-08, but my guess is they are not the type to paint with a broad brush.

NC-11: (Taylor – R) I am moving this out of the Probable Dem category because (a) the latest Reuters Zogby poll puts the race at 5% and (b) that is a 50% cut in Schuler’s lead since the last Reuters Zogby poll. I think Schuler is heading for a loss with that trend line, but I will park it in the Toss Up category for tomorrow.

OH-15: (Pryce – R) – No recent polls and this is a 50-50 district. Pryceis a savvy incumbent and the GOP GOTV effort should do some magic here. But since we have nothing I am pulling this out of the Probable Dem (where RCP has it) since there is no information either way.

NH-2: Bass – This is the third poll RCP has air dropped into the at risk seats simply because of some wild polling anyone should have know to keep out of simplistic averaging of polls (without any weigting due to methodology, sample size, etc). The UNH polls have fluctuated wildly over the course of a few days – demonstrating they are useless. I was going to dump them all, but that seemed harsh so I averaged them into one fuzzy picture of an 8.75% dem lead. Then to not be unfair I averaged the UNH combined score with the two other polls from the period since averaging 9 with one seemed ridiculous. The race is between 1% and 4.5% – and I am tempted to apply the bias just because of the obvious problems with the UNH polls. I have so many doubts about this I going to stick it in the Toss Up, but it doesn’t really seem as an at risk seat to me. If The Dems do turnout in larger numbers they will get this seat. If the meet or fall behind the Reps it stays rep.

PA-6: (Gerlach – R) – This race remains in the Toss Up category. The district went marginally for Kerry, and the lead is slim across the last four polls. The Democrat poll shows the tightest of the four results. A good GOP GOTV could salvage this one, so it will be interesting to see.

CT-5: Johnson – This is one of the few late movers in the RCP rankings which looked legit. The race is about 5% points ahead for the Democrat, and it was a 50-50 district in 2004. New Toss Up.

Probable Dem

AZ-8: (Open – R) – This open seat is trending Dem, even though the district is heavily Rep. If you average the last three polls you get a lead over 10% for the Dem, which cannot be offset by the bias correction of 5%. Stays Dem

IA-1: (Open – R) – This is Kerry country, so the desire for partisan cleansing is strong. I don’t buy the polls showing 20% leads for the dems – it will not be that lopsided. But I plan to watch this race as an indicator for the entire election. Kerry beat Bush 53-46. If there is a wave Braley will beat this 7% margin of victory. And if there is not one he will tie it. And if there is a Rep GOTV wave he will fall short. Here is a bellweather – an open Republican seat in Dem country.

26. FL-22: (Shaw – R) -This is a race that illustrates some real issues with the RCP rankings. The polls show a solid Dem lead and the district is went for Kerry. It should be a strong Dem pick, but RCP has it in lean Rep. Go figure.

Special Factors

1. TX-22: (Open – R) – This is Delay’s old seat, but the Reps are fighting back and all the handwringing voters will not make the effort to write in for Sekula-Gibbs (see, how hard is that?) is an insult to American voters. Stays Rep

3. FL-16: (Open – R) – I am moving this to Probable Rep because it is clear the Reps can hold this and probably will. No new polls, but Floridians get to Punch Foley to elect Negron. Without any evidence to the contrary this goes Rep.

PA-10: (Sherwood – R) – Nothing new in this race poll-wise. The latest non-Democrat poll shows the race -9 for Sherwood. Applying the bias gives us -4. I am holding this in the Probable Dem.

NY-26: (Reynolds – R) – Another race that looked at risk in the flash of media hysteria (which no serious person should partake in) which is clearly trending Rep.

PA-7: (Weldon – R) – There are no recent polls in this race, but plenty of bipartisan scandal muck now that we have found ACORN has forged a lot of registrations. My guess is Americans just aren’t going to buy into this. We have one late poll with Weldon only down 7. Given this is a Democrat poll we can apply the bias factor and see this race as basically a tie, where GOTV can pull it out for the Reps.

So where has a week brought us? This week we have 4 probable Dems and 8 toss ups. This shows a further slippage in the Dems grip on a House majority. If the turnout models are now showing 50-50 turnout then we split the toss ups and the dems will only pick up 8 seats. If they sweep the toss ups they get 12. And if the Reps sweep the toss ups…Well, let’s just not go there for the sake of the liberal egos out there.

And as I pointed out this is without any Rep pick ups of a Dem seat – which is highly probable. There is GA-12 to consider. That would drop the Dem opportunities to 3-7 seats. I think Dems could lose LA-3 as well. As before, the dems could do better – but it has become very, very unlikely with their lead shrinking in the generic ballot and turn out reporting (not predictions) showing the GOP equaling or surpassing Dems in early and absentee ballotting.

Want to make some history? This is the election to do it, so get out and vote tomorrow.