Nate Silver's Blog, page 137

Republicans Could Do A Lot Worse Than Merrick Garland Under President Clinton — Or President Trump

Senate Majority Leader Mitch McConnell said Wednesday that Republicans won’t hold hearings on Judge Merrick Garland, President Obama’s nominee to replace Antonin Scalia on the Supreme Court. If McConnell follows through and makes everyone wait until next year, he and his party will be taking a serious risk. The political situation for Republicans could get a lot worse — and McConnell might no longer be in charge of the Senate.

The logic of this isn’t that complicated. Right now, the most likely person to become the next president is Hillary Clinton, who is on the verge of wrapping up the Democratic nomination. The second most likely person is Donald Trump. His path to the Republican nomination is more tenuous than Clinton’s path to the Democratic one. But Trump had a successful day of elections on Tuesday — and if it isn’t obvious how Trump will get to 1,237 delegates, it’s even less obvious how anyone else will become the Republican nominee.

Neither choice is particularly palatable to members of the conservative Senate majority, only one of whom (Jeff Sessions of Alabama) has endorsed Trump. But if their choice between Clinton and Trump is bad enough, Republicans have some further bad news: Democrats have a shot at winning back the Senate. They’ll need to gain four seats to do so if Democrats hold the presidency, or five if a Republican wins. That isn’t a trivial task, but Republicans are vulnerable because a number of their blue- and purple-state senators who won election in the Republican wave year of 2010 are now on the ballot again. Furthermore, Trump could have a negative effect on down-ballot races; so could Ted Cruz, or someone nominated after a contested convention. Although I wouldn’t call Democrats favorites to win back the Senate, a Democratic Senate is probably more likely than not conditional upon Clinton becoming the next president.

If you combine the view of betting markets with a bit of back-of-the-envelope math, it suggests that Republicans face these rough probabilities:

A 40 percent chance of President Clinton with a Democratic Senate.A 30 percent chance of Clinton with a Republican Senate.A 20 percent chance of President Trump (probably with a Republican Senate).A 10 percent chance of Cruz, John Kasich or some other Republican.

You can quibble with those odds, obviously, and particularly with how much of a shot Trump would have against Clinton. Prudence would suggest that Trump’s chances are not zero: There could be an economic collapse, a terrorist attack or a Clinton scandal that could swing the election to Trump — or even if not, he could continue to rewrite the political rulebook and make fools out of political prognosticators. I wouldn’t take Trump at even money, though. For now, we’ll stick with the betting markets’ view, which imply that he’d be something like a 2-to-1 or 3-to-1 underdog.

The next step is figuring out what sort of person might be appointed to the Supreme Court in each circumstance. Suppose that, from Republicans’ view, a reliably liberal justice (say, someone with Justice Ruth Bader Ginsburg’s views) would score as a minus 10, and a reliable conservative (say, a clone of Justice Samuel Alito) would score as a plus 10. A truly centrist justice, who would side with the conservative position half the time and the liberal one half the time in key cases, would be a zero.

You’d probably score the possibilities something like what I have in the table below. With a Democratic Senate, Clinton would have relatively free rein to pick a nominee, although her majority in the Senate might not be especially large, and Republicans could filibuster her choice if Democrats win fewer than 60 seats. She’d also have won a fresh mandate even if Republicans held the Senate, however, and could probably get someone to Garland’s left confirmed, if not another Ginsburg.

SCENARIOPROBABILITYSCOREPresident Clinton + Democratic Senate40%-9President Clinton + Republican Senate30-6President Trump!20+2President Cruz or some other Republican10+9Weighted average-4The Republican calculus for a Supreme Court justice

Trump is a harder case to fathom. One way he could win the general election is by pivoting dramatically to the center and railing against partisanship. The Supreme Court nomination would then be one of President Trump’s first chances to demonstrate his abilities as a pragmatic deal-maker. He could nominate a conservative, but he could also pick someone with moderate or eccentric political views. Or he could make an unconventional choice: Trump once said his sister would make a “phenomenal” Supreme Court justice. A Trump appointment might be better for Republicans than a judge chosen totally at random from a circuit court, but perhaps not by much.

If you take a weighted average of these probabilities, you come up with a score of about negative 4. That’s equivalent to a center-left nominee – someone a lot like Garland, perhaps. If they wait until next year, Republicans might do better, but they could potentially do a lot worse.

And that’s before considering that the Supreme Court nomination isn’t happening in a vacuum. Polling suggests that a majority of the public wants the Senate to hold hearings on the next justice. Thus, blocking the appointment of Garland could hurt Republicans at the margin and further reduce their chances of keeping the Senate. On the flip side, it could curry favor with the Republican base. But one of the apparent lessons of this election is that the Republican base is neither as large nor as influential as we previously believed. Republicans continue to double down on a set of political strategies that seems to be failing.

Clinton Is Following Obama’s Path To The Nomination

Hillary Clinton may have had a sense of déjà vu. Eight years ago, after being ahead all night in Missouri’s Democratic presidential primary — the Associated Press erroneously called the state for Clinton — she lost after Barack Obama surged ahead with late-reporting votes from St. Louis. This time around, the shoe was on the other foot. Late Tuesday night, Bernie Sanders led Clinton by about 2 percentage points in Missouri. But Clinton pulled ahead after midnight on votes from St. Louis City and St. Louis County.

Clinton has not yet been declared the winner in Missouri, but she leads Sanders 49.6 percent to 49.4 percent in unofficial results. A win there would complete a 5-for-5 evening for her: Clinton won Illinois narrowly and Florida, Ohio and North Carolina emphatically. She was already likely to be the Democratic nominee, but she became more likely after what was perhaps the best evening of her campaign.

It’s not that Sanders had a terrible night. OK, losing Florida and Ohio by such large margins wasn’t good. But because of the Democrats’ proportional delegate rules, losing Missouri by a few thousand votes would make essentially no difference to his delegate count versus winning it by the same margin. Sanders’s narrow loss in Illinois was pretty respectable given that polls had once shown Clinton with a giant lead there. And Sanders lost North Carolina by only 14 percentage points. I’m not being sarcastic or damning with faint praise: If Sanders had lost the rest of the South by 14 points instead of margins that were sometimes 40 points or more, his path to the nomination would be considerably more viable. Sanders seems to be making progress with African-American voters.

But a night that wasn’t quite as bad as it seems wasn’t what Sanders needed. Even a pretty good night wouldn’t have mattered for him all that much. Instead, he needed a stupendous night that redefined the campaign. Big wins in Missouri, Illinois or Ohio might have done that; so might have making Clinton sweat in North Carolina or Florida. Sanders didn’t come close to passing that admittedly high bar.

I’m intrigued by the parallels to the 2008 campaign perhaps because it’s where FiveThirtyEight cut its teeth. I spent a lot of time in the spring of 2008 arguing that Obama’s lead in elected delegates would be hard for Clinton to overcome. But Clinton’s lead over Sanders is much larger than Obama’s was over Clinton at a comparable stage of the race. At the end of February 2008, after a favorable run of states for Obama, he led Clinton by approximately 100 elected delegates. Clinton’s lead is much larger this year.1 Clinton entered Tuesday’s contests ahead of Sanders by approximately 220 elected delegates. But she’ll net approximately 70 delegates from Florida, 20 from Ohio, 15 from North Carolina and a handful from Illinois and Missouri. That will expand her advantage to something like 325 elected delegates.

Sanders will need to win about 58 percent of the remaining 2,000 or so elected delegates to tie Clinton. Since the Democrats allot delegates proportionally, that means he’d need to win about 58 percent of the vote in the average remaining state to Clinton’s 42 percent, meaning he’d need to beat Clinton by around 16 points the rest of the way. Sanders would also have to overcome Clinton’s huge lead in superdelegates, although that’s probably the least of his worries. (If Clinton goes from winning the average state by double digits to losing it by the same margin, something cataclysmic will have had to have happened, likely sending her superdelegates scurrying for the exits.)

The second half of the calendar appears more favorable to Sanders than the states that have voted so far. Pretty much all of the South has voted, other than Maryland (if you consider it a Southern state), so Clinton doesn’t have many more delegates to rack up there. Not very much of the West has voted, and it will probably be a good region for Sanders. New York has lots of delegates, and could be interesting for Sanders, as could California. Pennsylvania could theoretically be a good state for Sanders, although it appears less promising for him after Clinton’s big win in Ohio.

Sanders can’t afford to merely come close in these states, as he did on Tuesday. Nor would narrow wins suffice. He needs to win these states going away to make up for his delegate disadvantage.

There’s no particular reason to expect he will do so. Instead, the Democratic race appears fairly static and fairly predictable along demographic lines. Even after Sanders’s dramatic, poll-defying win in Michigan last week, few Democratic voters decided upon or changed their vote this week, according to exit polls, and those who did broke about evenly between Clinton and Sanders. Polling averages quite accurately predicted the outcomes in the Democratic race on Tuesday, allaying the concern (one which worried us a lot!) that Democrats were experiencing some sort of existential polling crisis.

We’re fond of sports metaphors here at FiveThirtyEight. If the Republican race is Calvinball, with everyone making up the rules as they go along, the Democratic race is more like — zzzzzzz — golf. Clinton entered Tuesday night with the equivalent of a four-stroke lead with four holes to play. Then on the 15th hole, when Sanders already needed a minor miracle, she birdied while Sanders bogied. It’s not that it’s mathematically impossible for Sanders to win; Clinton could have some sort of epic meltdown. But she controls her own fate while Sanders doesn’t really control his, and she has quite a lot of tolerance for error.

Sanders has run a good campaign, and the fact that he ran competitively with Clinton in diverse states such as Michigan, Missouri and Illinois is more impressive in many ways than his early successes in Iowa and New Hampshire. But around 15 million Democrats have voted and, simply put, more of them seem to want Clinton as their nominee.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2704525/fivethirtyeightelections_2016-03-16-014825.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

Elections Podcast: Ides Of March Reaction

Another big night of primaries, and another late-night elections podcast conversation. We discuss Donald Trump’s wins in Florida, Illinois and North Carolina; Marco Rubio’s exit; what comes next for John Kasich after winning Ohio; and if Tuesday’s results truly sealed the nomination for Hillary Clinton.

http://c.espnradio.com/s:5L8r1/audio/2704525/fivethirtyeightelections_2016-03-16-014825.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

You can stream or download the full episode above. You can also find us by searching “fivethirtyeight” in your favorite podcast app, or subscribe using the RSS feed. Check out all our other shows.

If you’re a fan of the elections podcast, leave us a rating and review on iTunes, which helps other people discover the show. Have a comment, want to suggest something for “good polling vs. bad polling” or want to ask a question? Get in touch by email, on Twitter or in the comments.

How Will Tonight’s Primaries Affect The Presidential Race?

For this week’s politics Slack chat, we consider what the Republican and Democratic races might look like Wednesday morning, depending on what voters do in Ohio, Florida, Missouri, Illinois and North Carolina today.

micah (Micah Cohen, politics editor): Today may (will?) end up being the most pivotal day of the 2016 primaries. So, let’s define the range of possible outcomes. Can Donald Trump become the presumptive GOP nominee? Can Clinton become the presumptive Democratic nominee?

natesilver (Nate Silver, editor in chief): Isn’t Clinton already the presumptive Democratic nominee? Presumptive not meaning 100 percent, but in the range where something would have to go seriously wrong for her to miss out on the nomination.

clare.malone (Clare Malone, senior political writer): That’s really presumptive of you to say, Nate. But also probably accurate. I don’t necessarily think that this is the day that Trump becomes the assumed GOP nominee, though. I think the Ohio question is too big to ignore.

micah: What if he goes 5-for-5?

clare.malone: Well, then …

natesilver: Just a bit more context first. Clinton has won 58 percent of the pledged delegates so far, to say nothing of her advantage among superdelegates. By contrast, Trump has won 43 percent of delegates so far, and is opposed by party elites. I’m skeptical elites would deny Trump the nomination if he had a substantial plurality of delegates, but Trump and Clinton still come into the evening in rather dramatically different positions.

If Trump went 5-for-5, though? Yeah, that’d be huge for him. It would make it very hard for anyone else to get a plurality of delegates. It would take us to the endgame of the race. Trump would be a heavy favorite. But I’d stop a little short of calling him the presumptive nominee. We’d still have to see how a one-on-one race between Trump and (presumably) Cruz played out, and also to what lengths the party would go to try to deny Trump the nomination at the convention.

harry (Harry Enten, senior political writer): If Trump wins all five states tonight, I think he will be very difficult to stop. The press will be tremendous for him. He’ll accumulate a lot of delegates. Sure, he could be stopped, but it’ll be quite hard to do so.

natesilver: The inevitability narrative will be intense, yeah. Especially on CNN and other networks, which tend to take ambiguous results as Great News For Trump! If he actually has a good night, they’re going to be positively euphoric.

micah: All right, so let’s say Clinton is already the presumptive nominee, what would Sanders need today to reverse that?

harry: I think Sanders would need to come within 10 percentage points in North Carolina and Florida.

micah: And win Illinois, Ohio and Missouri?

harry: Well, yeah. I think it’s difficult to understate Clinton’s delegate lead right now. I’m talking pledged delegates, not superdelegates.

natesilver: Yeah. Let’s remember that for all the talk about Sanders’s amazing upset in Michigan, he still only matched his delegate target there, while losing ground relative to his target in Mississippi. He needs to start substantially beating those targets to have a shot at the nomination.

micah: OK, let’s focus on the Republicans and Ohio for a moment. What do we expect to happen there? What are you seeing on the ground, Clare? On the GOP side, Ohio seems like the most fought-over state. But if Trump wins all the other states but loses to Kasich narrowly in Ohio, I’m not sure that should be seen as any less of a good night for Trump. After all, Kasich is a super-popular governor of Ohio — if Trump keeps it close, isn’t that a big deal?

harry: On the Republican side, the state seems to be moving away from Trump. He hasn’t led in any of the last six polls, and he tends to underperform his polls.

natesilver: I basically agree with that, Micah. I wouldn’t go so far as to say that Trump would rather lose Ohio to Kasich to keep him in the race. That’s too cute by half. Or maybe even too cute by whole. But it’s probably more important how Cruz does vis-a-vis Trump tonight, since he’s Trump’s main rival.

If Trump wins Ohio but loses Missouri and (less likely) Illinois to Cruz, that’s a worse outcome for him than losing Ohio but taking four of the other five.

micah: Clare, what’s the mood in Cleveland? Are there pro-Trump vibrations?

natesilver: FEEL THE VIBRATIONS.

clare.malone: www.clevelandvibrator.com — that is a real industrial company. But I digress.

natesilver: To my surprise, that site is SFW.

clare.malone: Thus far today, I’ve been only to districts that lean Democratic. My swing (in my swinging rental car) this afternoon will be to suburban communities further afield that tend to be more conservative or to pockets of the City of Cleveland that are white working-class and might be feeling some of those Trumpian vibrations. I will report more on that on our live blog of the primary results tonight.

But if ads are any indication of the mood, I logged my TV watching this morning from 8 a.m. to 8:30 a.m., watching local news, and I saw seven commercials. Three were Kasich ads or pro-Kasich ads, one was for Cruz, one was for Clinton, one was for Sanders and there was a sole Trump ad, attacking Kasich for spending too much time in Michigan, because muck Fichigan.

natesilver: One thing we’ll learn more about tonight is which Republican candidates perform well in extremely Democratic congressional districts. In Ohio, that doesn’t matter for the delegate math since it’s winner-take-all. But it matters a lot in Illinois, New York and California, where there are large numbers of delegates at stake in districts that are 75+ percent Democratic in general elections.

micah: All right, so if a lot tonight rides on how Cruz does in Missouri and Illinois, let’s talk about that. Harry, how do we expect Cruz to do there?

harry: Well, Missouri is the black hole of polling. We have had one — one! — poll taken on the Republican side since the beginning of the year. The demographics suggest it could be a good state for Cruz given that he has won or come close to winning the states surrounding it. Illinois is a weird state. Cruz should do well in the south, but I’m unsure he can really play at all in the Chicago area, where most of the population is. Trump has led in all the polls in Illinois. If Cruz can win in Missouri and Kasich in Ohio, I’d be very interested to see how the media spins it.

clare.malone: “Illinois is a weird state” #analysis. I mean, I can tell you what the media spin is: CONTESTED CONVENTION!

natesilver: Illinois is pretty normal, methinks. Maryland is a weird state. Look how strange and misshapen it is! Alaska is a weird state. Illinois is pretty representative of the country as a whole, by contrast.

Anyway, the media is likely to pay too much attention to Ohio (and Florida) and too little to Illinois and Missouri. Still, I think a Cruz win in Illinois would matter more for the narrative than one in Missouri. Perhaps appropriately, since Illinois is a big, diverse state.

clare.malone: Maryland is pretty weird looking. It looks like the disintegrating fingers of a yeti. If Cruz won Illinois/Missouri, he would continue to have the pretty strong narrative argument that he’s the only one who’s consistently won (huzzah for homonyms).

harry: One of the things I’d watch in Illinois is whether voters properly fill out the ballot. Remember, you need to vote four times on the Republican side. Once for the statewide contest and three times for three district delegates. Newer voters may have problems. It’s conceivable that a candidate might “steal” delegates in a district, even as the statewide vote in that district goes for another candidate.

natesilver: Illinois seems super high variance. If Trump underperforms his polls by a couple of percentage points, then all of a sudden Cruz is winning districts in southern Illinois and Kasich might win a couple in and around Chicago. If Trump beats his polls by a couple of points, though, he could sweep most of the delegates and the inability of other GOP voters to coalesce around an alternative will look like a disaster.

harry: I should note that I went back last night and checked our polls-only forecasts, and Trump outperformed our forecast by more than a percentage point a grand total of three times. Doesn’t mean it can’t happen in Illinois. On the other hand, nine of 15 times he’s come in under our polls-only forecast.

clare.malone: Sorry, I’m just struck here that Marco Rubio’s name is no longer on our lips. (Or, the lips of our keyboards.) It’s amazing.

micah: Quite amazing … and Rubio is making noise that even if he loses Florida he won’t drop out.

clare.malone: That could also just be Election Day posturing, right?

micah: For sure.

natesilver: Yeah. Candidates always say that so as not to discourage turnout on Election Day. The one question I have, if Kasich loses Ohio and Rubio loses Florida, is whether one of them becomes more inclined to stay in because the other one dropped out.

clare.malone: Meanwhile they’re weeping on their buses, in the bathroom.

micah: What are Rubio’s chances in Florida?

natesilver: Our polls-plus model gives Rubio a … 3 percent chance. So a Rubio win would be pretty close to Sanders-wins-Michigan territory.

harry: It looks grim for Rubio in Florida. I have more hope that a third Deuce Bigalow film will actually be decent. Sure, it could happen, but none of the polls show anything other than a clear Trump win.

clare.malone: I kinda thought that Jeb Bush would have said something by now, speaking of Florida. He was meeting with candidates the past couple of days (not Trump).

Also, there were TWO Deuce Bigalow movies??

harry: He did a stint in Europe as a gigolo.

clare.malone: Oy. Can I ask a question about the post-March 15 universe real quick?

So, let’s say that Trump doesn’t sweep, that he loses Ohio, and that Kasich/Cruz both do well enough to stay in (Rubio’s toast, we assume). What does the next, say, month look like? Are people starting to game out plans for a contested convention? Is there another state that starts to look like the plum to bite into? If we were campaign strategists sitting in a hotel room, chain-smoking and applying Rogaine, what would we be thinking right now?

natesilver: I hate to be a broken record, but to me there’s a wide range of “well enough” from “Trump’s probably got this” to “Ohh, he’s in a lot of trouble here.”

harry: Note: Nate doesn’t actually hate being a broken record. He loves it.

clare.malone: “Well enough,” meaning?

natesilver: Kasich narrowly winning Ohio and Cruz finishing a distant second to Trump everywhere else would be good enough for them to both stay in the race, but would still be a pretty darn good night for Trump. But anyway — yeah, one thing that changes is that the primary calendar slows down quite a bit for the next month or so.

There’s Arizona next week, then there’s Wisconsin. Both states are pretty interesting. But comparatively speaking, it will be a breather.

Having a little more to think may lead to more talk about plans to contest the convention, or conversely to consolidate around Trump. You might see more speculation about a conservative third-party ticket, especially if Trump has a really good night.

clare.malone: Yeah, that’s the option I’m most fascinated by. To see if there’s anything to this splintering of the party. To see if they actually play out that scenario.

harry: I’d be very interested to see if Kasich basically camps out in New York. That’s where I would go if I were him. It has a lot of delegates and an electorate that could be favorable to him. New York’s primary is on April 19.

clare.malone: Long Island is lovely in the springtime.

natesilver: Hudson Valley >>> Long Island.

clare.malone: Oh, fo sho.

micah: Let’s jump back to the Democrats for a second before we wrap. If Clinton has a tough night — let’s say she loses Illinois, Ohio and maybe even Missouri, and wins North Carolina and Florida by less-than-expected margins — don’t things look pretty bad for her calendar-wise? From March 22 through April 9 there are a bunch of Sanders-friendly states. From Nate’s article earlier today:

FORECAST BASED ON DEMOGRAPHICS AND RESULTS IN PAST PRIMARIES“POLLS-ONLY” FORECASTDATESTATECLINTONSANDERSSANDERS WIN PROB.CLINTONSANDERSSANDERS WIN PROB.3/15Fla.67%32%4%63%34%Ill.544534524410Mo.544533494846N.C.68.... Va.4554675/17Ky.544532Ore.4455706/7Ca.... C.63369Demographic projections of the remaining Democratic states

natesilver: Yeah, the next little stretch is great for Sanders. I’m honestly not sure whether “the narrative” matters for the Democratic race, where voter preferences seem to be pretty stable. But if Clinton is perceived to have a bad night tonight, it could turn into a rough month for her.

harry: How many delegates are in those states, Micah? The states Sanders will do best in aren’t exactly delegate-filled: Alaska, Hawaii, Idaho, Utah and Wyoming.

clare.malone: Man, Hawaii seems like it could use some FiveThirtyEight eyes on the ground.

micah: Nice try, Clare. Yeah Harry, I don’t think Sanders will overtake Clinton in delegates, but it’ll be a rough few weeks for her campaign, making tonight pretty critical for her narrative-wise.

natesilver: For a change, though, I wonder if expectations for Sanders aren’t getting a little ahead of themselves. He’s losing Ohio by a pretty clear margin in the polls. In Illinois, both our polling and demographic models have Clinton favored. In Missouri, it’s a bit harder to say. Clinton could go 2-for-5, but she could also very easily go 5-for-5.

Another thing to keep in mind — on both the Democratic and Republican sides — is that the results have often gone contrary to momentum. A week ago at exactly this time, the talk was about how Clinton was starting to wrap things up and how Trump was in trouble. Then Clinton had a bad night and Trump had an excellent one. If you’re a journalist, it’s easy to over-adjust your expectations, especially when you’re living and breathing the campaign 24 hours a day.

harry: I’m cracking up here, and not just because I’m watching a great episode of “Sex and the City.” Sanders very unlikely to go anywhere. He has a ton of money and a solid constituency. But Clinton has the pledged delegate lead and the demographics behind her. Sure, something could change, but we’re really just playing out the string on a baseball season when one team has a 20-game lead with two months to play.

clare.malone: To Nate’s point about the media and adjusting expectations, I think that’s right — I’m almost growing tired of the countless hypotheticals at this point (see my own above, though, because I am a frail human, after all). I’m sure that Sanders will stay in, if only because, as we might forget at this point, he came into the race to pull it to the left, and the momentum that he picked up along the way was somewhat astonishing.

micah: And he HAS pulled Clinton to the left.

clare.malone: Yes. So, in many ways, he’s already won. He’s exceeded his MORAL* expectations.

*On the moral scale of Bernie Sanders (readers, feel free to have your own opinions about that).

micah: But he’ll lose.

clare.malone: He’ll lose, but maybe he’ll go to heaven, if there is one.

micah: All Bernies go to heaven.

natesilver: The governing rule so far on the Republican side has been the Principle of Maximum Annoyingness, which means an ambiguous outcome. From the standpoint of wrapping up everything with a neat bow, it would be easier if Trump swept everything tonight, or really took some lumps (losing three or more states, for instance). Most likely, we’ll be somewhere in between, however.

micah: All right, to close, let’s play our traditional Election Day headline game. Give me the New York Times headline for Wednesday morning for the Democratic race and the Republican race (Harry, you can do the New York Post).

harry: My New York Daily News headline: “Cleveland Choke: Trump Loses Ohio.”

natesilver: “TRUMP WINS FLORIDA; KASICH TAKES OHIO.” There’s a LOT of news tonight, so I think the headlines may be short and to the point.

clare.malone: “Trump, Kasich, Split Tuesday’s Prize.”

micah: “It’s A Mess.”

clare.malone: That’s a terrible headline.

micah: So is yours!

clare.malone: Don’t fire me.

harry: Let’s be clear: We all suck.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2704525/fivethirtyeightelections_2016-03-16-014825.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

Can Bernie Sanders Pull Off An Upset In Ohio?

Bernie Sanders’s win in Michigan last week was a massive upset relative to the pre-election polls of the state’s voters, which had shown Hillary Clinton ahead by an average of 21 percentage points. In fact, Sanders may have pulled off the biggest upset in the history of primary polling, eclipsing the previous record from 1984, when Gary Hart beat Walter Mondale in New Hampshire despite having trailed him by 17 percentage points.

When you consider Michigan’s demographics, however, the result wasn’t all that shocking. Michigan Democrats are fairly liberal and the state has a lot of college students — both factors that help Sanders. We aren’t just making this up as we go along; last month, we published state-by-state targets for the Clinton-Sanders race based on a few simple demographic variables in each state: specifically, its racial composition, how liberal or conservative it was, and how rural it was. Those targets had Sanders ahead of Clinton by 4 percentage points in Michigan.

Does that mean we called the upset in Michigan weeks ahead of time? No, we weren’t quite that good or lucky. The targets were based on a hypothetical race in which Clinton and Sanders were each winning about half the vote and half the delegates nationally. Since Clinton is ahead of Sanders nationally, she still would have been favored in our model (although not by the blowout margin that polls suggested).

Either way, the big gap between polls and demographics makes us nervous, especially because three more Midwestern states are voting today, including Ohio, where Clinton leads Sanders by about 11 percentage points in the polls. Historically, a margin like that would be quite safe: hence our polling model’s conclusion that Clinton is a 97 percent favorite. But after what just happened in Michigan? I’d love to drop a few bucks on Sanders if a bookmaker offered 30-to-1 odds against him, as our polling model does.

Fortunately, even if the polls haven’t been great, the conditions1 are potentially favorable for making demographic forecasts of the Democratic race. In 2008, under similar circumstances, I made demographic-based predictions of the Democratic race — see here for my North Carolina prediction, for example — which often outperformed the polls.

Those predictions in 2008 were based on regression analysis. They took advantage of the fact that Democrats report their vote by congressional district, which makes the sample more robust; by the time North Carolina voted eight years ago, for instance, hundreds of diverse congressional districts had already weighed in. So we’re overdue to apply the same technique this year.

In contrast to the demographic benchmarks we set in February, which were based on polling data, these are based on actual votes so far, aggregated across congressional districts. We can then compare these votes against demographic and attitudinal variables in each congressional district. For a more technical description of the analysis, see the footnotes.2 But basically, we’re just looking for sensible variables that have done a good job of explaining the split in the vote between Clinton and Sanders so far. The ones we included in the model are as follows:

The share of African-Americans is the best predictor of the Democratic vote to date, with Clinton performing significantly better in congressional districts with more black voters.Clinton also performs slightly better in districts with more Hispanic voters, although the magnitude of the effect is considerably smaller than that for black voters.Sanders performs better in districts that express liberal attitudes on social policy3.Sanders performs better in districts with major colleges, as measured by the number of people employed in postsecondary education in each district.As other researchers have found, Clinton performs better in the South, even after controlling for other factors.4Sanders performs better in districts where more voters are in labor-union households.Clinton performs better in districts where voters are more in favor of gun control.5Sanders performs better in caucuses relative to primaries, other factors held equal.

This regression analysis6 models the vote by congressional district reasonably well. We can aggregate the congressional district projections to come up with state forecasts. Here’s what they would have said about the states to have voted so far:

RETRODICTIVE VOTE SHARE BASED ON DEMOGRAPHICSACTUAL VOTE SHAREDATESTATECLINTONSANDERSCLINTONSANDERS2/1Iowa40%59%50%50%2/9New Hampshire475238602/20Nevada485153472/27South Carolina663373263/1Alabama74257819Arkansas59406630Colorado42574059Georgia73267128Massachusetts46535049Minnesota39603861Oklahoma52474252Tennessee66336632Texas65346533Vermont37621486Virginia643564353/5Kansas46533267Louisiana76237123Nebraska445543573/6Maine376235643/8Michigan51484850Mississippi77228316How a demographic model has fit the Democratic race so far

Our demographic “retrodiction”7 for Michigan still has Clinton winning, but only barely — by 3 percentage points, compared with the actual 2-point win for Sanders. Especially under the Democrats’ proportional allocation method, that’s a pretty minor difference. The model’s retrodictions in Vermont and Arkansas are also pretty far off, as you can see, but that makes sense given potential home-state effects for Sanders and Clinton in those states.

Other results are a bit harder to explain. How did Clinton (barely) win the Iowa caucuses when she got crushed in other Midwest caucus states, like Kansas and Minnesota? How did Sanders lose Massachusetts after winning New Hampshire by so much? How did Sanders win Oklahoma by 10 percentage points?

I have my theories — Clinton’s ground game may have saved her in Iowa, for instance — but my goal isn’t to explain away every last bit of variance (in which case I’d be guilty of overfitting my model). Instead, it’s to have reasonably sensible demographic-based projections that pass the smell test when applied to future states. Here are those forecasts, starting with the five states that will vote on Tuesday:

FORECAST BASED ON DEMOGRAPHICS AND RESULTS IN PAST PRIMARIES“POLLS-ONLY” FORECASTDATESTATECLINTONSANDERSSANDERS WIN PROB.CLINTONSANDERSSANDERS WIN PROB.3/15Fla.67%32%4%63%34%Ill.544534524410Mo.544533494846N.C.68.... Va.4554675/17Ky.544532Ore.4455706/7Ca.... C.63369Demographic projections of the remaining Democratic states

The numbers in Ohio jump out, since they suggest — in contrast to the polls — a very close race between Sanders and Clinton. After accounting for the uncertainty in the forecasts, the demographic model gives Sanders a 42 percent chance of winning Ohio, much better than the 3 percent chance that our “polls-only” forecast gives to him.

The news isn’t as good for Sanders in Missouri. There, the demographic model concludes that polls showing the race to be essentially tied are slightly too generous to Sanders; it forecasts Clinton to win by 9 percentage points.

In Illinois, the polls have been all over the place, with recent surveys showing everything from a 42-point lead for Clinton to a 2-point lead for Sanders. Our weighted polling average has Clinton up by 7 points there, and the demographic model is largely in agreement, forecasting a 9-point win for Clinton.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2704525/fivethirtyeightelections_2016-03-16-014825.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

Finally, both polls and demographics imply that Clinton is likely to win by blowout margins in North Carolina and Florida. If Sanders were to win or come close in one of those states, it would be an even bigger upset than Michigan and would suggest that something fundamental had changed in the Democratic race.

For clarity: These are forecasts based on the results so far, as opposed to benchmarks of what might happen in a hypothetical 50-50 race between Clinton and Sanders. If the candidates hit their forecasts on the nose in every state, Clinton would wind up winning by about 10 percentage points nationally. Thus, Sanders needs to substantially beat and not just tie these numbers to have a shot at the nomination. If you like, you can turn them into benchmarks by adding a net of 10 percentage points to Sanders. For instance, while the forecast in Connecticut is Clinton +3, the benchmark would be Sanders +7.

Since Sanders has lost ground to Clinton in the states to have voted so far, however, even that would not suffice for him to win the nomination; he’d have to beat these forecasts by something like 15 percentage points instead. It would be pretty shocking — but then again, Sanders has proven he can win when the odds are against him.

Elections Podcast: Violence At Trump Rallies

FiveThirtyEight’s Farai Chideya joins our elections podcast crew to discuss the incidents of violence at Donald Trump rallies around the country and look for historical precedents. Plus, what to expect from a big round of Tuesday primaries that could all but decide the Republican and Democratic nominees.

http://c.espnradio.com/s:5L8r1/audio/2703358/fivethirtyeightelections_2016-03-14-174758.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

You can stream or download the full episode above. You can also find us by searching “fivethirtyeight” in your favorite podcast app, or subscribe using the RSS feed. Check out all our other shows.

If you’re a fan of the elections podcast, leave us a rating and review on iTunes, which helps other people discover the show. Have a comment, want to suggest something for “good polling vs. bad polling” or want to ask a question? Get in touch by email, on Twitter or in the comments.

How FiveThirtyEight Is Forecasting The 2016 NCAA Tournament

Welcome to FiveThirtyEight’s forecasts of the men’s and women’s NCAA basketball tournaments. We’ve been issuing probabilistic March Madness forecasts in some form since 2011, when FiveThirtyEight was just a couple of us writing for The New York Times. While the basics of the system remain the same, we unveil a couple of new wrinkles each year.

Last season, we issued forecasts of the women’s tournament for the first time. Our big change for this year is that we won’t just be updating our forecasts at the end of each game — but also in real time. If a No. 2 seed is losing to a No. 15 seed, you’ll be able to see how that could affect the rest of the bracket, even before the game is over.

Live Win Probabilities

Our interactive graphic will include a dashboard that shows the score and time remaining in every game as it’s played, as well as the chance that each team will win that game. These probabilities are derived using logistic regression analysis, which lets us plug the current state of a game into a model to produce the probability that either team wins the game. Specifically, we used play-by-play data from the past five seasons of Division I NCAA basketball to fit a model that incorporates:

Time remaining in the gameScore differencePre-game win probabilitiesWhich team has possession, with a special adjustment if the team is shooting free throws.

These in-game win probabilities won’t account for everything. If a key player has fouled out of a game, for example, his or her team’s win probability is probably a bit lower than we’ve listed. There are also a few places where the model experiences momentary uncertainty: In the handful of seconds between the moment when a player is fouled and the free throws that follow, we use the team’s average free-throw percentage. Still, these probabilities ought to do a reasonably good job of showing which games are competitive and which are in the bag.

We built a separate in-game probability model for the women’s tournament that works in exactly the same way but uses historical women’s data. Thus, we’ll be updating our forecasts live for both the men’s and women’s tournament.

Elo Ratings

Otherwise, the methodology for our men’s forecasts is also largely the same as last year. But we’ve developed our own computer rating system — Elo — which we include along with the five computer rankings and two human rankings we used previously.

If you’ve follow FiveThirtyEight, you’ll know that we’re big fans of Elo ratings, which we’ve introduced for the NBA, the NFL and other sports. We’ve now applied them for men’s college basketball teams dating back to the 1950s, using game data from ESPN, www.sports-reference.com and other sources.

Our methodology for calculating these Elo ratings is highly similar to the one we use for NBA. They rely on relatively simple information — specifically, the final score, home court advantage, and the location of each game. (College basketball teams perform significantly worse when they travel a long distance to play a game.) They also account for a team’s conference — at the beginning of each season, a team’s Elo rating is regressed toward the mean of other schools in its conference — and whether the game was an NCAA tournament game. We’ve found that historically, there are actually fewer upsets in the NCAA tournament than you’d expect from the difference in teams’ Elo ratings, perhaps because the games are played under better and fairer conditions in the tournament than in the regular season. Our Elo ratings account for this, and also weight tournament games slightly higher than regular season ones.

Elo ratings for the 68 teams to qualify for the men’s tournament follow below.

RATINGSPROBABILITY OF…TEAMREGIONSEEDELOCOMPOSITEFINAL 4CHAMPSKansasSouth1209794.545.1%19.1%North CarolinaEast1207593.943.615.0VirginiaMidwest1205292.530.49.8Michigan StateMidwest2207891.833.98.9OklahomaWest2197290.032.06.8VillanovaSouth2204591.322.46.4KentuckyEast4201490.715.94.4West VirginiaEast3195689.316.23.4PurdueMidwest5193888.713.02.7OregonWest1203388.022.62.6Texas A&MWest3191586.812.42.4XavierEast2197387.79.91.8ArizonaSouth6195389.06.01.8DukeWest4191087.312.11.7MarylandSouth5187687.46.31.3IndianaEast5193887.45.81.1Miami (FL)South3193387.14.91.0Iowa StateMidwest4186786.56.41.0BaylorWest5183785.56.01.0TexasWest6178884.75.90.9UtahMidwest3188786.65.30.8Wichita StateSouth11189386.62.70.7CaliforniaSouth4187186.54.00.7IowaSouth7190485.93.20.6VanderbiltSouth11184685.62.40.5GonzagaMidwest11191686.03.20.5WisconsinEast7189684.82.90.4Notre DameEast6183284.42.60.3ConnecticutSouth9187285.32.10.3CincinnatiWest9179483.73.20.3ButlerMidwest9181584.22.50.3Seton HallMidwest6191484.51.80.2Virginia CommonwealthWest10179883.12.20.2DaytonMidwest7178882.41.60.1SyracuseMidwest10177282.71.30.1PittsburghEast10178782.31.20.1Saint Joseph’sWest8181481.61.10.1ProvidenceEast9182482.50.80.1Northern IowaWest11175180.20.8Stephen F. AustinEast14182481.00.4ColoradoSouth8175681.50.4YaleWest12179280.21.0Texas TechMidwest8177781.30.4TulsaEast11169079.90.2MichiganEast11176879.60.3Southern CaliforniaEast8173381.40.2Arkansas-Little RockMidwest12173478.90.2South Dakota StateSouth12173578.60.2TempleSouth10173078.50.2North Carolina-WilmingtonWest13172277.70.2Oregon StateWest7174077.60.2IonaMidwest13175978.20.1Green BayWest14166776.20.1Stony BrookEast13166377.10.1ChattanoogaEast12161076.6HawaiiSouth13173778.0Fresno StateMidwest14170876.6BuffaloSouth14161375.7Cal State BakersfieldWest15163575.00.1Middle TennesseeMidwest15163875.0North Carolina-AshevilleSouth15155374.2Weber StateEast15162373.3Florida Gulf CoastEast16154471.4SouthernWest16139268.0Austin PeaySouth16147768.8HamptonMidwest16148868.6Holy CrossWest16142066.9Fairleigh DickinsonEast16141766.72016 NCAA Tournament Team Ratings

Note, however, that Elo is still just one of six computer rankings that we use for the men’s tournament. The other five are ESPN’s BPI, Jeff Sagarin’s “predictor” ratings, Ken Pomeroy’s ratings, Joel Sokol’s LRMC ratings, and Sonny Moore’s computer power ratings. In addition, we use two human-generated rating systems: the selection committee’s 68-team “S-Curve”, and a composite of preseason ratings from coaches and media polls. The eight systems — six computer-generated, and two human-generated — are weighted equally in coming up with a team’s overall rating.

We’ve calculated Elo ratings for men’s teams only. For women’s ratings, we rely on the same composite of ratings systems that we used last year. You can find more about the methodology for our women’s forecasts

2016 March Madness Predictions

Ted Cruz Might Still Be Able To Stop Donald Trump

The good news for the Republican “establishment” is that there’s a man who might be able to stop Donald Trump. The bad news is that it’s Ted Cruz, someone they may dislike almost as much.

Cruz, who won Idaho on Tuesday while finishing second to Trump in Michigan, Mississippi and Hawaii, is within striking distance of Trump. So far, 29 percent of Republican primary voters have voted for Cruz as compared with 35 percent for Trump. Meanwhile, Cruz trails Trump by 100 delegates: not a trivial gap, although only one more than the 99 at stake in winner-take-all Florida next Tuesday.

The problem for Cruz is that Florida and the rest of the calendar probably aren’t as favorable to him as the states that have voted so far. (Florida looks like a Trump state as Marco Rubio loses ground.) But Cruz does have a few things going for him. He’s tended to outperform his polling almost everywhere. He’s won states in all four major regions of the country. And he potentially stands to gain if Rubio and perhaps John Kasich drop out.

The exit polls in Michigan and Mississippi asked voters who they’d pick in a two-way race between Trump and Cruz, also giving them the option to say they’d sit out the race. Among Rubio voters, on average between the two states, about 75 percent said they’d still vote in a Trump-Cruz race, and of those, 80 percent would prefer Cruz to Trump. Kasich voters were somewhat more equivocal; 55 percent said they’d still vote, and of those, two-thirds would go to Cruz over Trump. Although this is the first time the exit polls have asked about one-on-one matchups, the results are consistent with national polls showing Trump losing ground as the field winnows, as well as exit polls in previous states showing Trump being unpopular with Republicans who aren’t already supporting him.

What would the rest of the map (and more importantly, the delegate tally) look like in a potential two-way race between Trump and Cruz? I’m not ready to predict that. Beyond Florida and Ohio, which vote on Tuesday, there isn’t much polling in the other states. Nor has the Republican race been all that predictable along demographic lines, with Trump having performed well in states as diverse as Massachusetts, Alabama and Hawaii.

What I will do, however, is “retrodict” how the race might have gone had it been a two-man contest between Cruz and Trump all along. Could Cruz have beaten Trump in South Carolina, for instance? To do this, I’ll redistribute support from Rubio, Kasich and other candidates1 to Cruz and Trump based on the exit poll answers I described above. To repeat, these assume that most of their support would have gone to Cruz but not all of it, and also that a fair number of voters (especially Kasich voters) would sit out the race without their candidate on the ballot.

Here’s how things might have looked:

ESTIMATED SHARE OF VOTEDELEGATES UNDER UNIFORM GOP RULESDATESTATETRUMPCRUZTRUMPCRUZFeb. 1Iowa40.759.3523Feb. 9New Hampshire57.341.0184Feb. 20South Carolina47.652.31037Feb. 23Nevada56.742.9226March 1Alabama55.442.43710Alaska43.456.6621Arkansas43.154.5830Georgia50.748.05517Massachusetts64.932.5346Minnesota34.765.2630Oklahoma38.560.0833Tennessee50.347.44213Texas34.162.724123Vermont53.343.0123Virginia49.948.93215March 5Kansas30.568.4632Kentucky45.952.41034Louisiana47.049.81430Maine39.059.7418March 6Puerto Rico29.864.2319March 8Hawaii50.448.4144Idaho34.462.4525Michigan49.245.73917Mississippi52.545.0308Total444558Share of delegates42%53%What if Trump and Cruz had run one-on-one?

These figures estimate that Cruz would have won South Carolina, Arkansas, Kentucky and Louisiana in addition to the states where he already beat Trump. He also would have won Minnesota and Puerto Rico, which originally went to Rubio. Several other states, such as Michigan, Georgia and Virginia, would have been close between Trump and Cruz. Trump would be fairly dominant in the Northeast, however, and would still have won Nevada easily.

Overall, however, it’s likely that Cruz would have a delegate lead on Trump. I haven’t gone through each state’s delegate rules in detail, which would require estimating the vote by congressional district. Instead, I’ve applied something I call “uniform Republican rules,” which I introduced in a previous article. These are simplified delegate rules that give one-third of the delegates to the winner in each state, with an additional bonus for finishing with more than 50 percent of the vote, and distribute most of the rest proportionately. (The uniform rules also assign a small number of unpledged delegates in each state; these are not shown in the table.) Basically, they represent a compromise between winner-take-all and proportional methods that roughly reflects the rules in the average Republican primary and that approximates actual delegate allocations reasonably well.

Under these rules, Cruz would have 558 delegates to 444 for Trump. So instead of trailing Trump by 100 delegates, he’d lead him by roughly that amount instead.

This might sound too optimistic for Cruz, and certainly there are things a method like this can’t account for. (If Cruz had won South Carolina rather than Trump, for example, it would have changed media coverage of the race.) But keep in mind that Cruz is doing most of the work himself — he hasn’t trailed Trump by all that much in the states that have voted. So giving him even some of the Rubio and Kasich voters is enough to put him over the top in several states that he narrowly lost.

Perhaps we shouldn’t be grouping Rubio and Kasich together, however. Whereas Rubio’s numbers have been going downhill in Florida and other places, Kasich is even money or maybe a little better to win Ohio. And if he’s sustained himself this long on the basis of little more than a few second-place finishes in the Northeast, why would he drop out if he wins his home state?

The exit polls haven’t asked many questions about how supporters of the other candidates view Kasich, so we’ll assume that Kasich gets about as many votes from Rubio as Cruz does. (This makes a certain amount of sense: Rubio supporters have more in common with Kasich voters demographically but more in common with Cruz voters ideologically.) Specifically, we’ll assume that support from former Rubio voters goes 42.5 percent to Cruz, 42.5 percent to Kasich and 15 percent to Trump.2 We’ll also assume that 80 percent of them still vote, slightly up from 75 percent because we’ve given them another option. Kasich also gets to keep all of his own supporters, obviously. Here’s how that would shake out:

ESTIMATED SHARE OF VOTEDELEGATES UNDER UNIFORM GOP RULESDATESTATETRUMPCRUZKASICHTRUMPCRUZKASICHFeb. 1Iowa36.347.516.27193Feb. 9New Hampshire46.423.728.21534Feb. 20South Carolina41.937.820.230116Feb. 23Nevada53.133.313.22350March 1Alabama51.433.313.13890Alaska40.248.411.48190Arkansas39.744.013.912260Georgia46.236.615.947178Massachusetts54.517.325.63235Minnesota30.148.621.37245Oklahoma35.348.914.311300Tennessee46.237.014.539160Texas31.454.011.5251220Vermont37.918.739.7429Virginia43.432.523.030107March 5Kansas26.255.817.05303Kentucky39.238.520.727116Louisiana43.842.610.430140Maine34.149.415.25152March 6Puerto Rico25.538.630.04144March 8Hawaii45.238.215.51152Idaho31.052.513.25250Michigan38.428.627.9351111Mississippi48.438.410.627110Total47745275Share of delegates45%43%7%Kasich could cost Trump votes but give him states

Kasich wins Vermont under this math, but that’s the only state he gets. Meanwhile, the votes Kasich takes back from Cruz allow Trump to narrowly win South Carolina, Kentucky and Louisiana, when he would have lost them in a two-way race.

Overall, under uniform delegate rules, Trump wins slightly more delegates with Kasich running than without! Even though he wins fewer votes, the winner-take-all bonuses really help him. Cruz falls back behind Trump, meanwhile, while Kasich still isn’t doing that well, with only 75 delegates to date.

So if Republicans want to stop Trump from being their nominee, they should encourage Kasich (along with Rubio) to drop his bid? That’s one reasonable interpretation of this data, yes. Trump’s great weakness is that he loses ground as other candidates drop out. The most parsimonious strategy is just to get it to a one-on-one race.

However, I think the details make the case a bit more equivocal than that. Kasich could make a few good arguments for staying in the race:

O-H-I-O.Southern states, Kasich’s weakest region, are overrepresented so far.Cruz is weak in the Northeast, and even if Kasich can’t win those states, he might be able to prevent Trump from hitting winner-take-all thresholds in states like New York, where a candidate wins all statewide or congressional district delegates if he gets more than 50 percent of the vote.Rubio voters probably wouldn’t just split their votes 50-50; instead, they could behave tactically, going to Kasich in Connecticut (for example) but Cruz in Arizona.

Overall, I think it’s a close call between Cruz and Kasich having a better shot of stopping Trump as a tandem or Cruz being better off on his own. Kasich doesn’t appear likely to drop out before Ohio, in any event, so what happens there will probably go a long way toward determining his plans.

Rubio’s chances are hanging by a thread, however. He’ll presumably drop out if he loses Florida, but there’s an argument that his chances are slim enough in Florida that anti-Trump Republicans would be better off if Rubio dropped out immediately, which would give Cruz and Kasich better chances to win delegates in Ohio, Illinois, Missouri and North Carolina, all of which also vote Tuesday.

Either way, there’s likely to be some sort of endgame after March 15. Even if Trump wins both Ohio and Florida, Cruz might run relatively well against him in a one-on-one race from that point forward. Would it be too little and too late, especially with a lot of states in Trump’s northeastern base still left to vote? Possibly, but Republican “party elites” would finally have to make the very choice they were most hoping to avoid.

Check out our live coverage of the Republican debate.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2699899/fivethirtyeightelections_2016-03-09-150239.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

Elections Podcast: The Biggest Primary Polling Upset Ever

We convene a special gathering of our elections podcast team to discuss Bernie Sanders’s surprise victory in Michigan on Tuesday. By most measures, it’s the biggest polling miss in a primary in modern political history. But will the result in Michigan change the dynamics — or delegate math — of the Democratic race?

http://c.espnradio.com/s:5L8r1/audio/2699899/fivethirtyeightelections_2016-03-09-150239.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

You can stream or download the full episode above. You can also find us by searching “fivethirtyeight” in your favorite podcast app, or subscribe using the RSS feed. Check out all our other shows.

If you’re a fan of the elections podcast, leave us a rating and review on iTunes, which helps other people discover the show. Have a comment, want to suggest something for “good polling vs. bad polling” or want to ask a question? Get in touch by email, on Twitter or in the comments.