Tactical Voting Recommendations, 4/26
I don't love tactical voting -- it is both vaguely distasteful voting for someone you don't much like for the sake of someone you do, and really hard to coordinate -- but I also want my candidate to win this primary, and that can (for most practical purposes) only happen if the frontrunner is prevented from reaching 1,237 delegates on the first ballot. The 2016 primary is on a knife's edge; Mr. Donald Trump is better-than-even odds to clinch the nomination... which means denying him every possible delegate is critical for any other outcome. Tactical voting is called for.
So, if you support Ted Cruz, John Kasich, a convention dark horse like Paul Ryan, or simply oppose Donald Trump, here are the votes I think you should cast tonight in order to maximize your chances of winning (and mine).
PENNSYLVANIA: Cruz. Cruz and Kasich are close in the polls, but Cruz has the edge and a superior ground game, which will help him convert polling support to actual support more efficiently. Still, Trump is far enough ahead that this is likely to be a bloodbath. Only question is how well the Trump slate does among the unbound delegates. (PA has an unusual hybrid primary; don't ask.) I expect Trump to win the statewide vote (17 bound delegates) and about 45 of the 54 unbound delegates, plus or minus 8.
RHODE ISLAND: Cruz. The state's proportional, so it *shouldn't* matter which non-Trump you vote for (there's no spoiler effect in a true proportional race), but Cruz needs to stay above the 10% viability threshold (he's currently on the bubble at 13%), or the delegates he would have won go to Trump instead. Trump should get 8-12 delegates here.
CONNECTICUT: Kasich. Trump will likely win all the delegates in the state, but there is still a remote chance that Kasich could win one or two congressional districts if the anti-Trump vote consolidates behind him, which would net him 3 delegates, and hold Trump below a 50% winner-take-all threshold for statewide delegates. There's a 20% viability threshold which Cruz has a very low chance of making, and which won't be useful if Trump hits 50% anyway, so Kasich is really the right call here. I still think Trump wins all 28 delegates, but on a very good night, he might be held to 19.
DELAWARE: Cruz. Trump wins all the delegates either way. There is no clear anti-Trump right answer and Trump leads both by 40 points anyway. Cruz then becomes the default anti-Trump answer, on the basis that, if Kasich is sufficiently humiliated on 4/26, he might drop out, making the anti-Trump campaign much easier. If Trump doesn't win all 16 delegates here, it's a yuge upset.
MARYLAND: Kasich. According to PPP, Kasich polls ahead of Cruz in every district of Maryland. This surprises me, as Maryland District 6 seemed like prime Cruz territory to me, but we have to trust the data over our instincts. Vote Kasich in Maryland. Trump should still win something like 32 or 35 of Maryland's 38 delegates, but every non-Trump win counts.
OVERALL: Look, 4/26 is going to be bad for anti-Trump forces, but we've always known that, literally since early February. If Trump wins every single delegate of the night, that's unfortunate but not entirely unexpected. Tonight's tactical voting is to minimize the damage, not to actually make any gains. Indiana is the race that matters, and they're on May 3rd.
The most interesting development of the week, of course, is the Cruz-Kasich alliance. I'm going out of town for a couple of days, but I'll comment about it when I get back (for you, gerv!). By then, we should have a little more data about how it's playing with the public.