270ToWin doesn't use a sensible model for its simulation...

[LiberadorHugo]

I've literally had elections where Louisiana goes Democratic and Maine goes Republican. If LA votes Dem, Obama will win at least 350 EVs. I just did another won where WA and IA went GOP, but Obama won anyway with NC, VA, and MO. I think the model is too state-centric and needs some fudge-factor to account for national trends.

[Upton]

1. Optimism is nice but Obama is not going to win 350 electoral votes

I'll settle for 270.

[LiberadorHugo]

2. I never said he will...

All I said is that I run a simulation on 270 to win where Obama won Louisiana but lost a few safe states, giving McCain a nice victory.

[Nicholas D Wolfwood]

3. It's not much of a model

It just takes current polling and assigns the percentage of times a candidate wins based on that. For instance, if Obama is leading in Massachusetts 85-12, McCain will come up as winning Massachusetts 12% of all times you simulate it. Way too simplistic, in my opinion, and that's what causes wonky results.

[LiberadorHugo]

4. It's not THAT bad...

I've never had Obama lose MA and I think he's lost NY once.

[Nicholas D Wolfwood]

5. That was just an example.

You're not the brightest bulb, are you?

[skipos]

6. Forget about 270towin

Look these maps to get an idea of where things are at
http://www.realclearpolitics.com/epolls/maps/obama_vs_mccain/?map=10
http://pollster.com/
http://electoral-vote.com/