Wargaming the Electoral College

"What, me worry?" (AP Photo/Evan Vucci)

Let’s get ready to ruuuuuuummmmmmmmmmmble!

Most of the convention noise should have filtered out of the polls by now and it’s still too soon for the polls to have registered the new noise. By that of course I mean Hillary Clinton’s “unintentional” campaign appearance with the gay-hating, Taliban-loving father of the Pulse nightclub terrorist right behind her, and Donald Trump’s allusion of “Second Amendment people” to potential assassins. And then there’s Clinton’s “shocking pay-for-play scheme,” except that as Jane’s Addiction warned us almost 30 years ago, when it comes to Hillary, nothing’s shocking.

If you’ll remember, our base map last time around was RCP’s poll averaged result, with all Leaners removed — just showing Trump and Clinton’s Solids and Likelies. Five weeks later and there’s only one small change to the base map: Trump seems to have nailed down Nebraska’s stray EC vote, giving him all five of NE’s semi-splittable votes.

RCP No Leaners

The bad news for Trump is that with 13 weeks to go, he still doesn’t have a lock on the southern tier of red states. The bad news for Clinton is that she still doesn’t have a lock on the northern tier of blue states. But when we look deeper, you’ll see that the bad news is mostly Trump’s.

Let’s add RCP’s Leaners in to see what I mean.

RCP Leaners

RCP Leaners

There are a few changes from last month’s map, but two of them may prove vital. Clinton has shored up Maine’s stray split EC vote, which balances out Trump’s similar gain in NE. Trump has slipped in Arizona, Georgia, and North Carolina. Wisconsin has slipped out of the Leans Dem column, but Pennsylvania and its treasure trove of 20 EC votes has slipped in. So have Virginia and Colorado, even though just five weeks ago I said that I personally wouldn’t try to call those states yet, “even under the kind of duress involving an ISIS propaganda video crew, my most tender bits, and a finishing hammer.”

CO and VA could prove decisive in Trump’s attempt to turn enough of Obama’s blue states red to beat Clinton.

If we filter out polls of registered voters from 270toWin’s collection of recent surveys from Virginia, the remaining likely voter polls give Clinton an average lead of 44-39. Colorado’s likely voters give similar answers.

In our first go around, I showed you how Trump could win by making an endrun through the Rust Belt, where his economic message would seem to play best. So far at least, that message hasn’t been resonating outside his core supporters. PA’s likely voters are going for Clinton by almost double digits, and that’s before the “margin of cheating” from Pittsburgh and Philly gets factored in. OH is still within the margin of error, but the trend is a slight lead for Clinton. The best Trump has managed with OH likely voters is a tie in the Suffolk poll. Clinton is up by an average of about six in MI and WI, and is ahead comfortably enough in MN that nobody seems to have bothered to poll there in months. Putting outdated polls aside, Trump hasn’t campaigned in MN as though he thinks he can win it, and Clinton doesn’t seem concerned about losing there, either.

If Trump ever had a northern path to victory, that path has narrowed considerably.

Moving away from the upper tier, IA is anyone’s guess, although perhaps Trump’s safest bet for a pickup from 2012. FL polls are all over the place, and I maintain last month’s refusal to try and make any predictions there.

To help you visualize Trump’s challenge in this electoral environment, I’ve put together a map showing the actual 2012 results, minus today’s remaining battleground states.

2012-2016 Battlegrounds

2012-2016 Battlegrounds

Trump has lost 26 of Mitt Romney’s 206 EC votes, leaving him with a deficit of 90 on the way to 270. Clinton can count on only 247 of Barack Obama’s 332, but she needs only 23 more to win. If recent polls from AZ, GA and NC are to be believed, there’s a non-zero chance that Clinton could pick up her needed votes in any two of those usually reliable red states.

Of the 90 Trump needs in today’s battlegrounds, 64 of them must come from states Obama won in 2012. Let’s disregard those recent figures and assume AZ, GA, and NC go red again. Trump would still need to pick up FL and OH, plus MI and either WI, IA or NV. Or he’d have to pick up FL, OH, and every state other than MI.

Clinton would need only OH and any other state, or just FL by its little old lonesome — and then it’s “Hello, Madam President Pay-to-Play.”

Now let’s go to extremes.

If the election were held today (major caveat: the election will not be held today), Clinton’s top-end potential would be… formidable.

Clinton Top-End

Clinton Top-End

Clinton has the potential, however unlikely, of delivering a performance equivalent to her husband’s substantial wins in 1992 and 1996 — but without the aid of a billionaire third-party candidate like Ross Perot to split the conservative vote. In the event of a total Trump collapse, Clinton might even add IN and MS to her total, plus NE’s rogue elector. In that case, Clinton could outperform by 381-157. Even wilder, Utah voters could bolt to independent conservative candidate Evan McMullen, tipping six more votes Clinton’s way.

Assuming the polls, pollsters, experts, and Yours Truly have correctly labeled the battlegrounds, at this moment Trump has a top-end Electoral College performance of just 265. That would still give Clinton a win on par with either of George W. Bush’s.

Trump Top-End

Trump Top-End

If Trump wants to win in this field, he has got to flip CO, PA, MI, WI, or MN. Complicating things, he must do so (in all cases but PA) without losing NV, which I’ve given to him here despite polls showing Clinton with a modest lead.

That’s not to say this race is over — far from it. Even the Washington Post admits that “Trump’s floor in the polls is probably higher than you think.” And as David Burge noted earlier this week:

Even barring some terrible event, Clinton’s advantages in the Electoral College are not insurmountable. The next major planned event is the first of three presidential debates, scheduled for Monday, September 26. Trump has often been at his best during debates, and going off-the-cuff with history’s most-scripted candidate might be his best shot at regaining the lead he held just two weeks ago.

(All maps created with 270toWin’s iPad app, which is highly recommended. Perfect for fiddling with on the sofa while you watch the news. If you prefer to do your wargaming in a browser tab, the website is great, too.)