Delegate Math and the GOP

[southsider2k5]

I now work for a cruise line in business development. At this time, I am an informed voter and occasional small time campaign donor only. I think the city council campaign I managed in 2014 was probably my last campaign in any professional capacity.

 

So I need to find someone else when I run for Governor? Damn. Glad to see you posting again!

[StrangeSox]

I now work for a cruise line in business development. At this time, I am an informed voter and occasional small time campaign donor only. I think the city council campaign I managed in 2014 was probably my last campaign in any professional capacity.

[Balta1701]

On the original thread subject

This is an example of ignoring the elephant in the room, which in this case is winner-take-all primaries. The Republican schedule is backloaded with 15 of them, many of which are in big states. Let’s do a little simple math and see why all the talk of a brokered convention is so much horses***.

 

Based on this table at Real Clear Politics (and who would know the rules better than them?), I put together this spreadsheet [link to a Google sheet]. It’s based on a few simple assumptions that we can all argue:

 

Trump will continue to collect delegates in proportional races at the same percentage he has in the past (about 44%). Since Trump’s support in the Republican race is growing, and since there are states with thresholds and other obscure machinery that could award him more delegates to the top finishers, this is a pretty conservative estimate.

 

Trump will win all the winner-take-all primaries except for Ohio, Kasich’s home state.

 

Any contest where the delegates are unbound (112 delegates) is fodder for a floor fight, so those don’t count towards anyone’s totals. This assumption doesn’t matter because the rat pack of Cruz, Rubio and Kasich will keep fighting for scraps, so none of them will end up anything near Trump’s total.

 

Unsurprisingly, since the primaries are set up so that the leader of the pack can clean up in the later parts of the race, Trump goes into the convention with 1,472 delegates in this scenario. That’s 239 more than he needs. And Cruz and Rubio combined only get to about half of the 1,237 delegates needed to win.

 

I threw this model together in 10 minutes this morning, so it is crude. But it’s probably good enough for the purpose at hand, which is to show that anyone talking about a brokered convention or a Cruz/Rubio ticket needs to show why Trump will s*** the bed in a pretty spectacular fashion in the next couple of months.

[NorthSideSox72]

On the original thread subject

The core difference in what "Balloon Juice" did and what I did here, is that they are assuming all the polls hold. I didn't even get into the polls aspect, which especially for states more than a couple weeks out are worthless. I looked at from a general trending perspective, without assumptions about the WTA states. Then pointed out (as they did, sort of) that Ohio and Florida are the difference makers, probably. The tone of the candidates and party and others has changed significantly in their statements about Trump in the past week or two, and might make some other shift yet again, so I personally see the polls for non-immediate states to be nearly useless.

 

I mean, clearly if the current polls all hold, he wins outright. Not really a question. But keep this in mind - Ohio has had one poll since September. Florida 3 this year so far. Nothing from California and other big states in weeks or months. Nothing for nearly any of the WTA states since Super Tuesday. Using that set of data is - as you the scientist should recognize - problematic.

 

[Buehrle>Wood]

Isn't Illinois winner take all?

 

http://abc7chicago.com/politics/why-the-il...atters/1226252/

 

"There also may be a possibility Trump can win the popular vote in Illinois, but not the delegates. The primary here is not winner take all; Democratic and Republican delegates are chosen by congressional district."

 

ABC says it's not. I'm pretty confindent it is. I get delegates are directly elected here but if it's winner take all it seems they'd still be bound to the winner.

 

The only thing I can think of is it might make a difference at the brokered convention. I.E. Rubio wins district 1 but Trump wins the state. Rubios delegates have to vote for Trump at the conventionn at least the first round. Am I understandino that right? ABC still seems wrong then.

[NorthSideSox72]

Isn't Illinois winner take all?

 

http://abc7chicago.com/politics/why-the-il...atters/1226252/

 

"There also may be a possibility Trump can win the popular vote in Illinois, but not the delegates. The primary here is not winner take all; Democratic and Republican delegates are chosen by congressional district."

 

ABC says it's not. I'm pretty confindent it is. I get delegates are directly elected here but if it's winner take all it seems they'd still be bound to the winner.

 

The only thing I can think of is it might make a difference at the brokered convention. I.E. Rubio wins district 1 but Trump wins the state. Rubios delegates have to vote for Trump at the conventionn at least the first round. Am I understandino that right? ABC still seems wrong then.

 

There was some conflicting info on Illinois, but the best source seems to be the Illinois Republican Party themselves. Here is their page on this, which explains that there are 15 WTA delegates to the winner, and 54 that are directly elected by district. So it's a hybrid, mostly direct but with a decent chunk of WTA delegates.

 

[NorthSideSox72]

Update to the math after the weekend results...

 

Total GOP delegates: 2,472

Party-level delegates - unbound: 168

So, voting-linked delegates: 2,304

50.00001% of 2,304 is rounded to 1,153

 

Add the 168 to that, and the "magic" number is 1,321 to "clinch" a nomination prior to the convention. Other outlets will state 1,237, which is correct in gross terms for a majority of delegates, but I am focusing on what is needed for a convention-proof majority because of the intra-party anti-Trump movement.

 

So, let's now look at where the delegates stand after the weekend contests:

 

Trump: 391

Cruz: 304

Rubio: 154

Kasich: 37

---

Carson: 8

Bush: 4

Fiorina: 1

Huckabee: 1

Paul: 1

 

TOTAL SO FAR: 901 awarded (39.1% of voting-linked total)

 

After the above, the total remaining number of voting-linked delegates in future states and territories is 1,403. So for each candidate to reach the magic convention-proof number individually, they would have to do the following in the remaining states in terms of delegates and percentage of delegates won:

 

Trump: 930 delegates (66.3% of total)

Cruz: 1,017 delegates (72.5% of total)

Rubio: 1,167 delegates (83.2% of total)

Others: lol

 

If the states were all proportionally allocated, the chances of any of them getting that bullet-proof majority by the convention would be extremely slim. I mean, as well as Trump is doing, even with the help of a big winner-take-all SC he only has gotten 43% so far. So the idea of any of the three managing 66%+ is far-fetched. And even though Cruz and Rubio may not have to get that extra 2 or 3% (because some of the party delegates may actually support them), it's still just too big a stretch for them to reasonably expect.

 

I'm going to assume for the rest of this exercise that none of the big three drop out prior to convention, so they will keep going after delegates. Clearly if one of Cruz or Rubio drops out, the math changes.

 

The wildcard in all this the list of winner-take-all states: FL (99), MO (52), Marianas (6), Ohio (66), V.I. (9), AZ (58), WI (42), DE (16), MD (118), IN (57), CA (172), MT (27), NJ (51), SD (29). Those combine to make up 802 of the remaining 1,403 voting-driven delegates.

 

Let's assume that each candidate wins the remaining proportional delegates (601) at more or less the level they have so far on completed proportional states (all but SC - so 851 total), including Kasich and Carson (because who knows how long they stay, they are pulling smallish numbers anyway). That would put the counts, leaving the Winner Take All states out:

 

Trump (applying 40.1% to remaining prop and totalling): 632

Cruz (35.7%): 519

Rubio (18.1%): 263

Kasich (4.3%): 63

 

Now, this is similar to a general election electoral college game. For each candidate, which of the WTA states do they need in order to win that convention-proof majority? Here are what delegate counts they need in the WTA states to do that:

 

Trump: Must win 689 of 802

Cruz: Must win 802 of 802 (lol)

Everyone else: Mathematically impossible

 

In other words, for anyone other than Trump and Cruz, the only way they can win the nomination outright is to substantially increase their proportional position going forward - probabaly more than is realistically possible. But look also at how difficult that path is even for Trump. If he loses even just CA, or just MD, or FL plus almost any one other state, and he can't clinch before the convention.

 

Put this all together, and it comes down to this:

 

--The only candidate with a realistic shot at going into convention with a bullet-proof majority is Trump

--Even Trump's path, unless he picks up more proportional values, is a very tough road to get that majority if trends continue - in fact it looks less and less possible

--The party votes won't go to Trump if it's a convention situation, nor would Rubio's delegates or probably most of the stragglers with tiny amounts of delegates in carry

--The path above doesn't change much unless either one of the three big candidates drops out, OR one of them make very big moves in their Win%

 

What does all that mean? The nomination likely rests in the hands of Ted Cruz. Partly in an ability to win some of the WTA states. But also if he's willing to back Trump at convention, Trump clearly wins. If he isn't willing to do so, unless Trump absolutely romps the rest of the way, then a contested convention is on the docket.

 

[bmags]

I have to say, I'm now less convinced that the party will fall behind trump. It looks like there is a decent chance Rubio will stay in race just to take away from Trump.

[NorthSideSox72]

I have to say, I'm now less convinced that the party will fall behind trump. It looks like there is a decent chance Rubio will stay in race just to take away from Trump.

Agreed. The math makes clear that while Trump is still clearly the front-runner, the path to going to convention with the nomination fully in hand keeps getting steeper.

 

The wildcards are A) if Rubio drops out, or B) if Cruz was willing to hand his delegates to Trump at convention.

 

[Chisoxfn]

Agreed. The math makes clear that while Trump is still clearly the front-runner, the path to going to convention with the nomination fully in hand keeps getting steeper.

 

The wildcards are A) if Rubio drops out, or B) if Cruz was willing to hand his delegates to Trump at convention.

Your allocations of proportional states might be a bit off, as you are going off of historical levels and we have seen recently where Rubio has started to "not get enough" to qualify and thus the split ends up being more along two candidates vs. the multitude of candidates. Don't most upcoming states have like a 20% minimum vote threshold to get any delegates. That wasn't as common in some of the early states (and this is going off of the top of my head). So it could be a little overly punitive to assume a consistent delegate % racked up given current trends.

 

Another wildcard is that Cruz has yet to really prove he can win outside of strong evanglist states (and he has even lost some to Trump).

[NorthSideSox72]

Your allocations of proportional states might be a bit off, as you are going off of historical levels and we have seen recently where Rubio has started to "not get enough" to qualify and thus the split ends up being more along two candidates vs. the multitude of candidates. Don't most upcoming states have like a 20% minimum vote threshold to get any delegates. That wasn't as common in some of the early states (and this is going off of the top of my head). So it could be a little overly punitive to assume a consistent delegate % racked up given current trends.

 

Another wildcard is that Cruz has yet to really prove he can win outside of strong evanglist states (and he has even lost some to Trump).

That's all true, it's hard to predict easily, and it probably helps Cruz more than either Trump or Rubio.

 

For the March 8th states, MI and MS both have 15% thresholds, Hawaii 20%, and Idaho none. Hawaii is a caucus, and has a really small GOP population, so there is no telling what happens there. In MI, Looks like Trump and Cruz are both consistently about the 15% in recent polls, while Rubio is just barely above the line and in some polls actually below it - so that will be interesting to watch. Mississippi doesn't have polls from the past two weeks, but in late Feb, Trump was well above and Cruz & Rubio just snuck over too.

 

[StrangeSox]

Rubio is facing pressure within his own campaign to drop out before the Florida primary. Some polls have him down by 20 points to Trump, and they argue that this campaign is a lost cause and bowing out now is less damaging to his political future.

[NorthSideSox72]

Rubio is facing pressure within his own campaign to drop out before the Florida primary. Some polls have him down by 20 points to Trump, and they argue that this campaign is a lost cause and bowing out now is less damaging to his political future.

 

There are two key dates for Rubio. One is tonight - as he's very close to the cutoff line for delegates in three states (Hawaii is open caucus, but is also very small). He needs to get above the line in a couple of those tonight. The other of course is Florida, where in the latest poll I saw he's down 8 points. He needs to win Florida.

 

[StrangeSox]

The latest polls have Rubio behind by 16 points on average, and if some of his own advisers are pressuring him to drop out, his internal polling must be showing them the same thing.

 

The latest polls out of Ohio show Kasich with a much better shot, though.

Edited March 8, 2016 by StrangeSox

[bmags]

One good thing, it's unlikely rubio's constituents move to Trump in large numbers, as Bush/Christies did not appear to. But, as "winner momentum" begins to take hold, it's possible they do.

 

[NorthSideSox72]

One good thing, it's unlikely rubio's constituents move to Trump in large numbers, as Bush/Christies did not appear to. But, as "winner momentum" begins to take hold, it's possible they do.

Yeah, momentum is away from Trump and over to Cruz at the moment though. We'll see how tonight, and more so the 15th, go.

 

[Balta1701]

Donald Trump has won both Mississippi and Michigan handily enough that they were called as the polls closed.

[pettie4sox]

Trumpmania is gaining steam.

 

Time to jump on the bandwagon.

[Brian]

Trumpmania is gaining steam.

 

Time to jump on the bandwagon.

 

I'm good

[LittleHurt05]

Trumpmania is gaining steam.

 

Time to jump on the bandwagon.

 

'MERICA

[Buehrle>Wood]

Trumpmania is gaining steam.

 

Time to jump on the bandwagon.

So far he's survived:

-The entire Republican party

-The entire Democratic party

-The entire media

-The sitting president

-The pope

-Half the world calling him Hitler

-and lots lots more

 

You cannot Stump the Trump.

[StrangeSox]

I don't think that the Democratic party has really turned their guns on trump yet?

[southsider2k5]

Trumpmania is gaining steam.

 

Time to jump on the bandwagon.

 

:crying

[NorthSideSox72]

Trump and Cruz dynamics didn't change much last night. But it was a terrible night for Rubio, who it appears will get zero delegates on the night. Kasich continues to climb though, and could become more of a story going forward.

 

I'll re-do the math again today, but just show the numbers instead of all the text.

 

[NorthSideSox72]

Update to the math after the Tuesday results...

 

BASELINES

Total GOP delegates: 2,472

Party-level delegates - unbound: 168

So, voting-linked delegates: 2,304

50.00001% of 2,304 is rounded to 1,153

Total ending delegates to clinch: 1,237

Delegates to clinch a convention-proof nomination: 1,321

 

CURRENT STANDINGS - total so far: 1,050 awarded (45.6% of voting-linked total)

Trump: 464

Cruz: 363

Rubio: 154

Kasich: 54

---

Carson: 8

Bush: 4

Fiorina: 1

Huckabee: 1

Paul: 1

 

PERCENT OF REMAINING DELEGATES NEEDED FOR CONV-PROOF MAJORITY (1,254 voting-linked delegates remaining)

Trump: 857 delegates (68.3% of total)

Cruz: 958 delegates (76.4% of total)

Rubio: 1,167 delegates (93.1% of total)

Others: mathematically impossible

 

Again assuming none of the 4 drop out (which may not be the case)... 802 of the remaining 1,254 voting-linked delegates are WTA, so 452 proportional.

 

DELEGATE COUNTS PLUS REMAINING PROPORTIONAL USING % WON OF THOSE SO FAR...

Trump (applying 44.2% to remaining prop and totalling): 464 + 200 = 664

Cruz (34.6%): 363 + 156 = 519

Rubio (14.7%): 154 + 66 = 220

Kasich (5.1%): 54 + 23 = 77

 

NEED TO WIN IN WTA STATES TO CLINCH PRIOR TO NOMINATION

Trump: Must win 657 of 802

Cruz: Must win 802 of 802 (lol)

Everyone else: Mathematically impossible if proportional trends continue

 

Trump and Cruz's level of difficulty haven't changed materially. It's nigh on impossible for Cruz, and truly for everyone else, barring big changes in their proportional performance. Of course, this doesn't obey that Cruz and Kasich are trending up, and Rubio down. But as Trump is basically running flat lately, the general parameters around him remain unchanged. He needs to win most of the big remaining WTA states and big chunks of the smaller ones as well, to win prior to convention.