“I do see a path for him,” she says. Great! Could she explain it to me? Because all I see right now, even in a best-case scenario for anti-Trumpers, is Trump getting within 150 delegates or so of 1,237 by June 1st and then Cruz trying to make a goal-line stand to deny him the clinching margin he needs in the final primaries on June 7th.

We have Trump needing 94 of CA's 172 dels for 1,237. So yes, will likely come down to 6/7. https://t.co/DmLIOdjDw8 https://t.co/osbtb31aiO

— Dave Wasserman (@Redistrict) March 16, 2016

New Jersey, Montana, South Dakota, New Mexico, and the big one, California, vote that day. Trump either wins the nomination outright that evening or he ends up in Cleveland very close to a majority, with 46-49 percent of the overall delegate pool. And remember, hundreds of delegates, including those earned by Marco Rubio, will be unbound even on the first ballot. All Trump would need to do to clinch if he’s short would be to wheel and deal with those few he needs to gain their support. He’s been canny about that so far in landing major endorsements. Why wouldn’t a skilled negotiator be able to persuade far less powerful people to make a deal with him?

There is nothing in the RNC’s rules that prohibits delegates from cutting a deal for their votes, and lawyers say it is unlikely that federal anti-corruption laws would apply to convention horse-trading. (It is not clear that even explicitly selling one’s vote for cash would be illegal.) To lure a governor, for example, the offer of a Cabinet post could be necessary, while a delegate may be swayed by a job as regional HUD administrator or a seat on the Postal Regulatory Commission. A crucial vote on a procedural question could be ensured with a state party’s website-design contract to a delegate’s cousin’s firm.

But why waste an ambassadorship on someone who could be bought for far less? Every delegate and alternate is already paying for individual travel costs to get to Cleveland. Most state parties tell delegates to expect to spend $3,000 out of pocket on airfare, hotel and meals, and for some it could prove an unexpected hardship. (Delegates are assigned hotels by state; some could end up paying for the La Quinta Inn, others stuck with a bill from the Ritz-Carlton.) As blogger Chris Ladd has noted, Trump’s slate in Illinois contains “a food service manager from a juvenile detention center, a daycare worker from a Christian School, an unemployed paralegal, a grocery store warehouse manager, one brave advocate for urban chicken farming, a dog breeder, and a guy who runs a bait shop.” Could some of them be tempted to flip their votes if a generous campaign, super-PAC, or individual donor picked up the costs of their week in Cleveland? Far-flung territories that are treated as states under RNC rules offer even richer opportunities for geographical arbitrage. Round-trip flights in July to Cleveland from the Northern Mariana Islands, which nine delegates are unbound after the first ballot, already cost more than $2,000 each.

If Kasich drops out before the convention, his delegates would be unbound too. Winning them might be a straightforward matter of Trump convincing Kasich to be his VP. And all of this will be happening amid very real pressure on the delegates not to give thugs a pretext to riot by denying Trump the nomination. Put all of that together and it’s hard to imagine how, barring an RNC rule change that all but steals Trump’s delegates from him ahead of the first ballot, he doesn’t win at this point.

Plus, denying Trump before the convention would require Cruz to continue to perform for month after month at the rate he’s been performing, without some decisive number of undecideds breaking for Trump in each state as as he begins to look more like the presumptive nominee. Granted, Cruz should catch a bounce now that most of Rubio’s supporters, like Haley, are headed his way, but that good news is countered by the reality that the primaries to come are in states that are poor matches demographically for him. That’s why the results in Illinois and Missouri were so disheartening: If Cruz was stronger than everyone thought in urban areas, you would have expected it to show up last night while conservatives were roiling to stop Trump. Instead Trump won both states. Even some states that are “supposed” to favor Cruz look shakier than thought for him upon examination. Remember when his strategists told Politico that he needs to win “decisively” in Arizona and Utah next week? Here’s the new poll out of Arizona:

The last poll out of Arizona from March 10 (MBQF):
Trump 37%
Cruz 23%
Kasich 15%

Trump also had a larger lead of 41%-24% among early voters

— Marcus (@MarcusHUSA) March 16, 2016

And in case all of that eeyorism isn’t enough, here’s some more bad news breaking as I write this. When I said Kasich winning Ohio last night was a pyrrhic victory, this is what I meant:

Kasich announces buying ads in UT. Cruz already playing there. If there's an anti-Trump vote to be split, they're splitting it.

— Shane Goldmacher (@ShaneGoldmacher) March 16, 2016

Kasich hasn’t a prayer of winning Utah. Neither does Trump, actually; Mormons are a solidly anti-Trump demographic. What Kasich could do by contesting the state, though, is hold Cruz under 50 percent. Utah’s delegates, 40 in all, are awarded proportionally unless the statewide winner wins an absolute majority of the popular vote, in which case it’s winner take all. Cruz could gain 40 delegates on Trump in one fell swoop — and here comes Kasich to try to stop him.

I think it’s hopeless, but then I would say that. If you want a more hopeful (but only slightly) scenario for holding down Trump’s margins the rest of the way, read John McCormack.