fusag wrote:
Probably depends on what polls you read. I follow fivethirtyeight as a poll aggregator; he's biased toward Obama, but I trust his methodology. He has New Jersey about a point more in favor of Romney than Connecticut, and New Jersey has far more electoral votes, so I think it's clear which state has a bigger role to play in determining the election. That said, New Jersey hasn't polled lower than +7 for Obama this election. That sounds like safely blue territory.
Right now, there probably are only about 7 states in play. Maybe taken over the whole campaign, there could be more. Even within each swing state, the campaigns are targeting very specific groups that are wavering or still elastic (rather, I should stay relatively elastic after 2/4/8/12 years of partisan politics has entrenched other blocs). Furthermore, given the difficulty in getting someone to switch sides, voter turnout is becoming an important factor. Some research says that the impact of negative ads is not to convince someone to switch sides, but to sow enough uncertainty that they don't vote for the person they would have (and there's other research that disagrees...). Hence candidates pushing for their supporters to vote early and not be convinced to not vote later in the campaign.
Right now, there probably are only about 7 states in play. Maybe taken over the whole campaign, there could be more. Even within each swing state, the campaigns are targeting very specific groups that are wavering or still elastic (rather, I should stay relatively elastic after 2/4/8/12 years of partisan politics has entrenched other blocs). Furthermore, given the difficulty in getting someone to switch sides, voter turnout is becoming an important factor. Some research says that the impact of negative ads is not to convince someone to switch sides, but to sow enough uncertainty that they don't vote for the person they would have (and there's other research that disagrees...). Hence candidates pushing for their supporters to vote early and not be convinced to not vote later in the campaign.
While I generally agree (I think the number of swing states are higher--heck, even PA is only a point out of the margin of error), for game purposes I have to be pretty inclusive as to what gets included. The game just isn't fun if there are only 7 states to deal with. Also remember that I did this around Labor Day, and at the time CT was more of a swing state than New Jersey.
I follow most of the electoral vote sites, but namely fivethirtyeight, Real Clear Politics, Electoral-Vote.com, and (to a lesser extent) Election Projection. (RCP, to me, has the best info in the most organized manner.) When I looked for the swing states, I basically looked at all four of these (plus CNN's map) and collected how many were "swing" and "leans"/"barely" for one candidate. If at least three of the four had a state it got added. I believe this came to about 14 states or so. Then to balance out regions (and the math), I picked the next closest states, which in this case was three additional states (I believe they ended up being Connecticut and Arizona and one more I don't recall, probably Minnesota). Yes, they may be +7 or +8 ahead, but things still change--look at both New Mexico and Missouri, who a month ago were swing states and now are pretty safe (with around +5 or +6 for their candidate). Even states like South Dakota and Indiana were within 5 points for a while.
And, like you said, part of the game is demographics. You can actually target specific demographics within each state.
I will probably do an after-election adjustment looking at actual polls, but I'm not sure--Since the game is a what-if scenario, I kind of wanted to use Labor Day as a rough "flash point" for the rest of the game. It's not quite there, since there will be events and cards from before that date, and I will probably end up dumping CT for NJ and another state, but we'll see.