I was thinking about this earlier this morning but Erick Erickson and Brandon Finnigan got there before I did. I don’t think either of them is suggesting this as a serious, i.e. workable, solution. It’s more a matter of recognizing that the odds of stopping Trump have now grown so slim that some sort of unorthodox, high-risk solution will need to be found. (See also “Should Rubio offer to make Kasich his VP?”) Cruz isn’t going to beat Trump with Rubio still in the race. But by the time Cruz drops out, which won’t happen until March 2nd at the very, very earliest (and likely not even then), it’ll probably be too late for Rubio to beat Trump head to head. In fact, both Erickson and Finnigan are recognizing a terrible but nonetheless almost certain reality: The only man in the field with a chance of clinching a majority of delegates before the convention is Donald Trump.

Once you accept that fact, the conservative anti-Trump strategy follows logically: For the rest of the primaries, the goal must be not to elect this or that candidate but to simply deny Trump a majority and then take your chances at the convention. And the most obvious way to implement that strategy is with a Rubio/Cruz alliance. Erickson:

I think you should look at your polling in your state and see who is closest to Trump. Then, I don’t care whether you prefer the third place guy, cast your vote for the second place guy.

If Cruz and Rubio are too immature to stop fighting each other to attack Trump, we should force them.

If Marco Rubio is in first or second place headed into your state’s primary, vote for Rubio, even if you like Cruz.

In the same way, if Ted Cruz is in first or second place headed into your state’s primary, vote Cruz, even if you like Rubio.

Simple math: Trump is polling at around 33 percent in most states and Rubio and Cruz are polling at around 20 percent each. If you could add their totals together, you’ve got a non-Trump winner nearly everywhere. Granted, that would mean that neither Rubio nor Cruz can clinch the nomination before the convention either, but that’s secondary right now to the greater goal of preventing Trump from destroying the conservative coalition. Says Finnigan:

What it entails: Rubio is given free reign in caucus states, Florida, Virginia, Michigan, Illinois, Massachusetts and Vermont;

— Brandon Finnigan (@B_M_Finnigan) February 23, 2016

Cruz the exact same in Texas, Arkansas, Alabama, Georgia, Tennessee, Kansas, Mississippi, Louisiana, North Carolina, Oklahoma.

— Brandon Finnigan (@B_M_Finnigan) February 23, 2016

Let them deny Trump by consolidating their vote in the Mach states, then go after each other.

— Brandon Finnigan (@B_M_Finnigan) February 23, 2016

It’s possible that Rubio and Cruz will hold Trump under a majority of delegates simply by continuing on independently through the end of the primaries, picking up delegates here and there, but it’s no sure thing. And this is the important part: Even if Trump finishes short of a majority, he’ll still almost certainly end up as the nominee at a brokered convention if he gets there with a huge near-majority. If Trump lands in Cleveland with, say, 49 percent of delegates and Rubio and Cruz each show up with 25 percent, there’s no way the party will shank Trump by nominating one of the other two. There’d be an open revolt among Trump supporters; even Trump’s critics would howl that he’d been denied a prize that he deserved. Telling a guy who’s the closest thing to representing the will of a majority of voters that he can’t have the nomination when he’s running a populist campaign against an elitist establishment would be a party-breaker. If you want Rubio or Cruz to have even a chance of becoming the nominee, you need them to go to the convention either with more delegates than Trump or rough parity with him. That needs to start today.

But it won’t. The threshold problem: How many of Rubio’s and Cruz’s supporters would heed their candidate’s advice to vote for the other in a given state? Some would, but only some. What if, say, Cruz endorsed Rubio in Nevada and Cruz ended up with 20 percent of the vote anyway? How eager would Team Rubio be to carry their arrangement forward? To many Cruz fans, Rubio is an amnesty shill and a guy with dangerously interventionist tendencies abroad. To many Rubio fans, Cruz is a self-aggrandizing obstructionist with a complex about his own conservative purity. Not all of them are going to pull the lever for the other candidate even if it benefits their guy long-term. How long does the alliance last if there aren’t enough vote-switchers in the first few states to help the favored candidate there defeat Trump?

What about the fact that caucuses are a greater commitment of time and energy than primaries are? Under Finnigan’s proposal, Cruz fans would be asked to spend hours caucusing on Rubio’s behalf while Rubio fans wouldn’t be asked to make the same commitment for Cruz, which is arguably either unfair to Cruz fans or unfair to Rubio given that many Cruz voters obviously will decline to make this commitment for him. Beyond that, what happens if the alliance holds for a week or two and a delegate imbalance opens up between Rubio and Cruz — for instance, Cruz wins big in Texas with Rubio’s help, piling up a bunch of delegates, whereas Rubio wins in smaller states, leaving him behind Cruz. Team Marco would be paranoid that Cruz could void the deal at any time, having already banked his delegate windfall. The candidates would need to divide their states evenly in advance to try to ensure that they end up with an equal number of delegates, but you can only do so much on that score. Since many states award delegates by congressional district, it may be that Rubio overperforms in some districts that are “supposed to” go for Cruz or vice versa. If their delegate hauls end up out of whack, what happens?

Or what happens if one of them comes to the conclusion that the other hasn’t endorsed him as enthusiastically as he endorsed the other? For instance, if Cruz makes a big show of backing Rubio in a “Rubio state” and then Rubio shows up in a “Cruz state” and says, “I hope my supporters will back Ted Cruz here, okay, bye” and then skedaddles, Cruz won’t think much of that arrangement. And then there’s the question raised by Finnigan of how long a Rubio/Cruz alliance should last. Should it go all the way to the convention or, if Trump quits before then, all the way to the day that Trump quits? If Trump quits while there’s a delegate imbalance between Rubio and Cruz, what happens then? Don’t forget either that later states on the calendar are more “moderate,” with more centrist and center-right votes to be had, which would favor Rubio over Cruz if the race is a two-man contest between them at that point. Why would Cruz want to risk a head-to-head race with Rubio on that terrain? (Possible answer: Because he’s likely to be an also-ran by then if he keeps going the way he is now.) The fundamental problem here is that nominating a conservative may be the primary goal for some conservative voters but it’s obviously not the primary goal for Rubio and Cruz and their most devoted fans. The primary goal for Rubio is nominating Marco Rubio and the primary goal for Cruz is nominating Ted Cruz. There’s no way out of that prisoner’s dilemma, I fear, especially given the embittering nastiness between the Rubio and Cruz campaigns over the last few weeks. Which is another reason to think Trump’s destined to be the nominee.