Nate Silver's Blog, page 101

CARMELO NBA Player Projections

CARMELO NBA Player Projections

FiveThirtyEight’s CARMELO* identifies similar players throughout NBA history and uses them to develop a probabilistic forecast of what a current NBA player’s future might look like.

How good willbe?RANDOM PLAYERCarmelo Anthony

NEW YORK KNICKS
SMALL FORWARD
32 YEARS OLD

WEIGHTED AVERAGE OF PAST THREE SEASONSBADAVG.GOOD

PERCENTILE


VITALS

50TH




Height
6’8″




Weight
230




Draft position
3








SCORING






True shooting %
53%




Free throw %
82%




Usage %
31%








TENDENCIES






3 pt. frequency
24%




FT frequency
31%








PASSING/BALL HANDLING






Assist %
20%




Turnover %
10%








DEFENSE/REBOUNDING






Rebound %
12%




Block %
1.1%




Steal %
1.4%




Defensive /-
-0.8

WINS ABOVE REPLACEMENT PROJECTION

CONFIDENCE INTERVAL

PROJECTION

CATEGORY:GOOD STARTER5-YR MARKET VALUE:$75.5m’132014’15’16’17’18’19’20’21’22’23’24051015209.22.76.44.93.92.41.70.50.20.19.22.76.44.93.92.41.70.50.20.1PERFORMANCE OF THE 10 MOST COMPARABLE PLAYERS01020251Vince CarterYEAR: 2010SIMILARITY: 5401020252D. WilkinsYEAR: 1993SIMILARITY: 5201020253Tom ChambersYEAR: 1992SIMILARITY: 4801020254Kobe BryantYEAR: 2011SIMILARITY: 4701020255Clyde DrexlerYEAR: 1995SIMILARITY: 4701020256Alex EnglishYEAR: 1987SIMILARITY: 4601020257Chris WebberYEAR: 2006SIMILARITY: 4501020258Dirk NowitzkiYEAR: 2011SIMILARITY: 4401020259James WorthyYEAR: 1994SIMILARITY: 43010202510Antawn JamisonYEAR: 2009SIMILARITY: 41

THE FINE PRINT
2014
’15
’16
’17
’18
’19
’20
’21
’22
’23




Offensive /-
4.6
3.6
2.8
2.6
2.0
1.6
1.5
0.7
0.1
-0.3


Defensive /-
-1.0
-2.2
-0.2
-0.8
-0.6
-0.7
-0.6
-0.8
-1.0
-1.0


Total /-
3.6
1.4
2.6
1.8
1.4
0.9
0.9
-0.1
-0.9
-1.2















Value
$29.8m
$9.3m
$24.8m
$25.4m
$22.0m
$14.3m
$10.4m
$3.4m
$1.4m
$0.9m


Minutes played
2982
1428
2530
2331
2072
1495
1088
519
345
323

*FiveThirtyEight’s Career-Arc Regression Model Estimator with Local Optimization (CARMELO) is a system that forecasts a player’s future performance. The similarity score is an index measuring how comparable one player is to another, scaled such that a score of zero is average similarity and 100 is the highest possible degree of similarity. Plus-minus is derived from Box Plus/Minus, a measure of the number of points per 100 possessions that a player contributed to his team, relative to an average NBA player. All player ages are as of Feb. 1, 2017. Minutes played is prorated to 82 games.

Additional contributions from Aaron Bycoffe and Neil Paine.

Sources: Basketball-Reference.com, ESPN and Jeremias Engelmann.

What’s New In Our NBA Player Projections For 2017-18

Welcome to the 2017-18 edition of CARMELO, FiveThirtyEight’s NBA projection system.

Or to be more precise, welcome to the initial 2017-18 edition of CARMELO. We have a lot of basketball-related projects on tap for the summer, which vary from exploring what makes teams succeed in the playoffs when they struggle in the regular season to trying to develop our own defensive metrics based on player-tracking data. So there’s a chance that we’ll give CARMELO and our other NBA products a more serious overhaul before the NBA regular season starts in October. Be prepared for these numbers to change, in other words.

In the meantime, the initial adjustments from last year’s edition of CARMELO are relatively minor. For a basic outline of the system, which projects player performance by identifying similar players since the NBA-ABA merger in 1976, please see our explanation from two years ago. Here’s what’s different this year:

BPM vs. RPM, revisited, again

We were fairly pleased with how our CARMELO player projections performed last year, with the system identifying breakout stars such as Giannis Antetokounmpo, Nikola Jokic and Rudy Gobert. But we also use CARMELO to make team projections, and we weren’t all that happy with those. In fact, according to the APBRmetrics message board, which tracks various projection systems, CARMELO went from having some of the most accurate team forecasts in its debut season in 2015-16 to some of the least accurate last year. What happened?

In 2015-16, our projections were based on a combination of Real Plus-Minus (RPM), a system that rates each player primarily based on how his team performs when that player is on or off the court, and Box Plus/Minus (BPM), a system that rates players using traditional box-score statistics such as assists and steals. Last year, we switched to using BPM only. Why? There are a lot of things to like about BPM, including that it’s considerably more transparent than RPM, and it can be calculated going back several decades, making for easier historical comparisons.

But as a predictive tool, BPM does not appear to be as accurate as RPM. Instead, BPM has trouble picking up on factors such as defense and team cohesion. That led CARMELO to overrate teams such as the Minnesota Timberwolves and underrate more defensive teams such as the San Antonio Spurs last year. If we’d run the numbers using RPM instead of BPM in 2016-17, our projections would have been above-average again as compared with the projection systems that APBRmetrics tracks, we discovered.

All of this stuff gets complicated, and discussions can quickly devolve into alphabet soup. But for better or worse, the choice of metric matters quite a lot. According to BPM, Russell Westbrook’s 2016-17 season was easily the greatest in NBA history. According to RPM, he was only the ninth-best player in the league last season.

The upshot is that in the short run, we’re using a blend of two-thirds RPM and one-third BPM for this edition of the CARMELO projections. In the long run, we’re interested in developing our own plus-minus stat (but no promises about that quite yet).

Projections for international players

In past years, we’d published CARMELOs for NBA veterans such as Paul George based on their NBA statistics and for rookies such as Lonzo Ball based on their NCAA statistics. But we didn’t have projections for rookies such as the Knicks’ Frank Ntilikina, who played in international leagues rather than the NCAA. (The Belgian-born Ntilikina grew up in France and played for the French professional team SIG Strasbourg this year.)

So this year, we’ve introduced simple projections for European draftees based on biographical data: specifically, their age, draft position, height, weight, position and home country. CARMELO identifies players such as Ricky Rubio, Evan Fournier, Dennis Schroeder and Tony Parker as being similar to Ntilikina, for instance.

‘Gap year’ and ‘draft-and-stash’ players

Another tricky case involves rookies who missed the entire NBA season, or almost the entire season, on account of injury or other factors in their first year. The Sixers’ Ben Simmons, the No. 1 overall draft pick in 2016, didn’t play at all last season because of a foot injury, for example.

We’ve improved how the system handles these “gap year” players. They are now compared only to other players who also sat out what would have been their first season, such as Blake Griffin (who missed the entire 2009-10 season) or Julius Randle (who played just one game in 2014-15 before getting hurt). This is a small sample of cases, so the projections for these players can be somewhat noisy. Still, we think this is better than essentially just ignoring the injuries, as we’d been doing before.

European and other international players, meanwhile, often play an additional season or two abroad even after they’re drafted by an NBA team. (The NBA team who drafts these “draft-and-stash” prospects can retain their rights for a couple of seasons under most circumstances.) Draft-and-stash players such as the Celtics’ Guerschon Yabusele are projected based on comparisons to other draft-and-stash players.

Better biographical data

Finally, we’ve switched to using NBA.com, instead of Basketball-Reference.com, as our primary source for player heights and weights. Players’ listed weights — and sometimes even their official heights — can change from season to season. Basketball-Reference.com generally lists players by the heights and weights they were listed under as rookies, whereas NBA.com keeps more up-to-date with the changes.

We’re also using more detailed data on players’ positions than in the past. While a player’s primary position is determined by Basketball-Reference.com, we’re now also accounting for players’ secondary positions based on additional positions listed at ESPN Fantasy Basketball. Thus, a power forward who also plays center will be treated slightly differently than a power forward who also plays small forward.

We hope you’ll enjoy this year’s edition of CARMELO, and we’ll notify you of further changes as we make them throughout the offseason.

Which GOP Senators Might Resurrect The Health Care Bill?

In this week’s politics chat, we game out how likely the Senate health care bill is to pass, senator by senator. The transcript below has been lightly edited.

micah (Micah Cohen, politics editor): The GOP health care bill in the Senate is “in peril” according to The New York Times, and Senate Majority Leader Mitch McConnell has reportedly told members of the caucus that he will delay a vote on the bill until after next week’s July Fourth recess:

MCCONNELL tells senators: He will delay the health care vote until after the recess to solicit more support from GOP senators

— Manu Raju (@mkraju) June 27, 2017

But, of course, a vote on the House’s version of an Obamacare repeal bill was delayed before it passed in that chamber. And the media declared that bill nearly dead too. So we’re taking an “it’s-not-dead-until-it’s-dead” approach here at FiveThirtyEight. And, accordingly, we thought it’d be useful to go through the Republican senators who appear most likely to vote “no” on the bill (its official name is the Better Care Reconciliation Act of 2017, fwiw) and talk through their incentives and whether they’ll actually vote no.

Everyone good on the plan?

perry (Perry Bacon Jr., senior writer): Yes.

natesilver (Nate Silver, editor in chief): Can we have a draft?

micah: No.

natesilver: I pick Dean Heller.

micah: We have 10 GOP senators, and then we’ll have a group of wild cards.

natesilver: And how are we evaluating them?

clare.malone (Clare Malone, senior political writer): By looks, the way voters do.

micah: They’ve all expressed some version of concern/reluctance about the bill, so we’ll judge each on how much we think they’re posturing vs. how likely we think they are to actually vote “no”?

clare.malone: Let’s do it.

micah: The scale is …

This person is telling the truth. They’re a likely “no.”
¯\_(ツ)_/¯
Bullshit. This senator is posturing and will likely vote “yes” if it comes down to it.

natesilver: Can we use emojis?

Mitch McConnell Isn’t Playing 13-Dimensional Chess

The Senate’s version of the GOP health care bill, which Majority Leader Mitch McConnell unveiled last week, hasn’t had the smoothest rollout. Some 22 million more people would go uninsured under the Better Care Reconciliation Act by 2026, the CBO announced on Monday, essentially unchanged from the House’s version of the bill. The bill has been criticized both by conservative Republicans, such as Kentucky’s Rand Paul and Utah’s Mike Lee, and by relatively moderate ones, such as Maine’s Susan Collins and Nevada’s Dean Heller. Local newspapers have given it negative coverage. The American Medical Association opposes the bill, as does the AARP, while private insurers have had a mixed reaction. It’s gotten a mainly negative response from conservative intellectuals and policy wonks, although with some exceptions.

All of this has cut against McConnell’s reputation as a strategic genius — to the point where some reporters and commentators have suggested that McConnell intentionally drafted a poor piece of legislation. The New York Times’ Jennifer Steinhauer, for instance, hypothesized that McConnell was hoping for the bill to fail so he could move on to tax reform and other priorities. Balkinization’s David Super argued that McConnell began with a bill he knew “pretty much everyone” would hate so that changes to the bill would make it look better by comparison.

I’d posit a simpler idea: This bill is exactly what McConnell wants because it’s right in line with his long-term goals. As Bloomberg’s Francis Wilkinson points out, the BCRA “will transfer hundreds of millions of dollars from poor and middle-class people, in the form of health care, to rich people in the form of tax cuts.” To be more specific, the bill would cut Medicaid spending by $772 billion over 10 years, according to the CBO, and reduce health care tax credits by about $408 billion. It would also reduce taxes and penalties by more than $700 billion, mostly in the form of “repealing or modifying tax provisions in the ACA that are not directly related to health insurance coverage, including repealing a surtax on net investment income and repealing annual fees imposed on health insurers.”

To put it another way, the BCRA is less a health care bill than a tax cut (that will mostly benefit the wealthy), coupled with a trillion-dollar-plus reduction in federal government spending on health care (that mostly benefited the poor and the sick). Those goals — lowering taxes on the wealthy, trimming the welfare state and reducing the size of government — are at the core of Ronald Reagan’s philosophy of movement conservatism, and they’ve been the primary axis of political conflict between Democrats and Republicans for most of the past several decades.

To Super’s point, the bill does leave McConnell with some room for compromise. There’s more deficit savings than under the House’s bill, so McConnell could cut a deal to get Alaska’s Lisa Murkowski on board — Alaska would be especially harmed by the BCRA because of the cost of health care there — or increase the amount of spending on opioid treatment to help sell Ohio’s Rob Portman and West Virginia’s Shelley Moore Capito on the bill.

But these are matters of tactics, not strategy. McConnell’s strategy is fairly obvious: He wants to pass legislation that lowers taxes on the wealthy and reduces government spending to the largest extent politically practicable.

Republicans probably would have faced some backlash from their base if they’d made no effort at all to repeal and replace Obamacare. But it’s easy to imagine more popular ways to do it. Removing Obamacare’s individual mandate and employer mandates could have eliminated some of its least-popular provisions and allowed the GOP to claim victory. McConnell and House Speaker Paul Ryan could have dismantled the Obamacare exchanges but left Medicaid intact. They could have attempted a more comprehensive effort at health care reform, such as the bill proposed by Collins and Louisiana’s Bill Cassidy, which would have given states flexibility in what to do about Obamacare. None of those options would have done as much to lower taxes on the wealthy or reduce government spending, however.

Needless to say, the approach comes with enormous political risk for Republicans. The House’s bill was extraordinarily unpopular: On average, unfavorable views of the bill exceeded favorable ones by 25 to 30 percentage points. (For instance, 59 percent of Americans disapproved of the House bill, compared with the 32 percent who approved, in the most recent CBS News poll.) The Senate’s bill, since it’s substantially similar to the House’s bill, isn’t likely to rate much better.

The unpopularity of the bill isn’t surprising. Government spending on welfare programs has slowly (if not always steadily) increased throughout U.S. history, and while passing these programs is politically challenging, removing them is even harder once you can identify the groups (in the case of the BCRA, everyone from seniors in nursing homes to opioid addicts) who would lose out. Most people are happy with their own health care, so proposed changes to the health care system usually start out facing a headwind.

Related:

The bill also comes at an awkward political moment for movement conservatism. The candidates who ran on those ideas, such as Sen. Marco Rubio, flamed out badly in the Republican primaries. Instead, Donald Trump won. But the ideas in the health care bill also aren’t very compatible with Trump’s populism. Trump ran on repealing and replacing Obamacare, but he promised to replace it with “something terrific,” said he’d protect the “lower 25 percent” even if it defied Republican orthodoxy, and repeatedly criticized other Republicans who said they’d cut Medicaid. McConnell and Ryan have co-opted Trump’s mandate — narrow as it was — and tried to turn it into their own, at considerable risk to both themselves and Trump.

Nevertheless, these sorts of opportunities are rare. Republicans hold both branches of Congress and the White House, and these periods don’t always last for very long; Barack Obama and Bill Clinton lost their House majorities after two years in office, never to win them back again, for instance.

So McConnell has come out swinging for the fences. It still isn’t clear whether he’ll succeed. But it also isn’t all that complicated.

CORRECTION (June 27, 11:26 a.m.): A previous version of this article misspelled the names of New York Times reporter Jennifer Steinhauer and the web site Balkinization. The text has been updated.

Politics Podcast: The Health Care Episode

 












Subscribe: Apple Podcasts |
ESPN App |
Download |
RSS |
Embed |
New to podcasts?


Embed Code

Public health reporter Anna Maria Barry-Jester joins the FiveThirtyEight Politics podcast team to break down the Senate’s health care bill and its implications for the public. The team also discusses how likely the bill is to pass and whether a bill this unpopular will affect Republicans in the 2018 midterm elections.

You can listen to the episode by clicking the “play” button above or by downloading it in iTunes , the ESPN App or your favorite podcast platform. If you are new to podcasts, learn how to listen .

The FiveThirtyEight Politics podcast publishes Monday evenings, with occasional special episodes throughout the week. Help new listeners discover the show by leaving us a rating and review on iTunes . Have a comment, question or suggestion for “good polling vs. bad polling”? Get in touch by email, on Twitter or in the comments.

Will The Senate Pass Its Health Care Bill?

We’ve seen this movie before. Republicans, after a secretive drafting process, finally unveil a health care bill. But the initial reaction is tepid, with both moderate and conservative Republicans expressing “concerns” and demanding changes to the bill.

That’s what happened when the House Republican leadership released its health care bill, the American Health Care Act, in March. And that’s what happened on Thursday, when Republican Majority Leader Mitch McConnell released the Senate’s version of a health care bill, called the Better Care Reconciliation Act.

Soon after the BCRA was officially released, four conservative senators — Rand Paul, Ted Cruz, Mike Lee and Ron Johnson — issued a statement saying that they were “not ready to vote for this bill,” although they also said they were “open to negotiation” about it. This was hardly the only opposition to the bill, however. According to The Washington Post’s whip count, five other relatively moderate Republicans — Susan Collins, Lisa Murkowski, Dean Heller, Rob Portman and Shelley Moore Capito — said they had serious concerns with the bill, ranging from its defunding of Planned Parenthood to its cutting of funds for Medicaid. That would seem to put McConnell in a difficult spot: Assuming no Democrats vote for the bill, McConnell can afford to lose only two of his senators. But he already has nine senators objecting to the bill from various directions.

We also remember how that other movie ended, however. After a false start in March, when House Speaker Paul Ryan pulled the AHCA from the House floor, Republicans regathered to pass their bill in May by a 217-213 majority.

It’s tempting, therefore, to assume that the same process will play out in the Senate and that the BCRA will eventually pass after a few weeks of drama. When the ink is still drying on the bill, as it is right now, Republican senators have lots of incentive to express their concerns, whether to stake out a negotiating position or to posture before their constituents. McConnell is a skilled vote-whipper, however, and the Senate has generally toed the party line in support of the GOP agenda and President Trump. And even if the bill isn’t exactly the one that senators might want, it’s still in line with longstanding GOP policy objectives.

So all of this probably ends with a big signing ceremony in the Rose Garden — and millions of people losing their insurance, right?

Well, maybe. Betting markets give roughly even odds that key provisions of Obamacare, such as its employer mandate, will be repealed before the end of the year. I don’t know what probability I’d assign to the bill’s passage myself. But I do think it’s worth pausing to take an inventory of some of the differences between the House’s situation and the one the Senate now faces. There are quite a few of these — some of which make the bill more likely to pass and others less so:

Republicans need a higher percentage of their members to vote for the bill in the Senate. In the House, the GOP needed 216 of its 237 members to support the bill, or 91 percent. In the Senate, they need 50 of 52 members, or 96 percent. This is a big difference. If McConnell can’t nail down the most critical votes — such as Collins, Heller, Murkowski and Paul — he doesn’t have a lot of backup options. This factor is harmful to the bill’s chances of passage.
Senators face less immediate electoral pressure than members of the House did. Whereas every member of the House will face voters next year, only eight Republican senators are up for reelection in 2018. And of these, only Nevada’s Heller, Arizona’s Jeff Flake and mayyyyybe Texas’s Cruz look as though they’ll face any real pressure in their general elections. It’s true that the AHCA/BCRA would have profound repercussions for people’s insurance for many years, so the Republicans up in 2020 and 2022 will eventually have to deal with the public fallout if the bill become law. Still, this factor is helpful to the bill’s chances for now.
The House had to start from scratch, whereas the Senate can piggyback on the House’s bill. This factor is helpful to the bill’s chances in various ways. McConnell can tell his members that the sky didn’t fall after House Republicans passed their bill. He can tell Cruz and Lee that if the bill was good enough for the House Freedom Caucus, it should be good enough for them. And he can argue that abandoning the bill now — after the House has already taken a tough vote on it — would leave the GOP in the worst of all possible worlds, with swing voters angry at the GOP for passing the AHCA in the House, and base voters angry at them for not getting health care over the finish line in the Senate.
Senate Republicans won’t get a “do-over.” One conceit about the House bill, which Ryan may have used to gain leverage with moderate members, was that it would be completely rewritten in the Senate. Just vote on something to keep the process moving, Ryan could argue, and we’ll fix the problems later. But, in fact, the Senate bill turned out to be extremely similar to the House bill. Furthermore, Senate Republicans won’t get a second chance to fix problems with their bill because the House is expected to approve the Senate bill’s “as is” if the Senate passes it, which would send BCRA right to the president’s desk. This raises the stakes and is harmful to the bill’s chances.
The Senate must use reconciliation (or Republicans will have to rewrite the rules). Republicans are hoping to use reconciliation to pass the BCRA, which would allow them to pass it with a simple 51-vote majority and avoid a Democratic filibuster. But the Senate parliamentarian has not yet ruled on the bill and will not to do so until a Congressional Budget Office score is released next week. If some parts of the bill don’t pass muster, Republicans will either have to rewrite the BCRA or adopt novel interpretations of the reconciliation rules. (Or if they so choose, they could eliminate the legislative filibuster altogether.) The more drastic the procedural step, the more likely that moderates such as Collins and perhaps traditionalists such as John McCain will object to it. This limits McConnell’s flexibility and is harmful to the bill’s chances of passage.
The Obamacare exchanges are in more trouble than they were a few months ago. Obamacare isn’t in a “death spiral” — but it is in a more precarious position than it was a few months ago, with insurers pulling out of markets or substantially raising premiums on a state-by-state basis. (In part, this is because of problems the GOP itself has helped to create.) Trump has claimed that this will create political problems for Democrats and force them to the negotiating table. Polls suggest that Trump is wrong about this: Instead, a Kaiser Family Foundation poll in April found that by a 2-1 margin, voters would blame Trump and Republicans for problems with Obamacare going forward. Nevertheless, this gives McConnell and Trump a certain sort of leverage over congressional Republicans. If Republicans will be blamed either way — for an unpopular bill or for the collapse of Obamacare — there’s no politically easy way out and the GOP might as well pass its bill, McConnell and Trump can say. Therefore, this dynamic is helpful to the bill’s chances of passage.
Trump and the bill have grown even more unpopular. Trump’s approval rating was 44 percent when the House first considered the bill in March and about 42 percent when it passed the bill in May. Today, Trump’s numbers are closer to 38 or 39 percent. But the really shocking numbers are on the bill itself. For instance, voters disapproved of the Republican bill by a 59-31 margin in the most recent CBS News poll. (Although the numbers vary some from survey to survey because of question wording, the CBS results are fairly typical of the consensus.) Furthermore, the bill’s numbers are getting even worse as the Republicans’ secretive drafting process hasn’t helped to sell the bill to the public. It’s unusual to see major bills that are this unpopular become law. Maybe Republicans will keep the blinders on, but this factor is harmful to the bill’s chances of passage.
There have been several special elections since the House vote. Special elections for vacant House seats were held in Montana last month and in Georgia and South Carolina this week. Republicans won all three. At FiveThirtyEight, we look at the margins in these races and not just the winners, which leads us to conclude that these were a pretty bad set of results for Republicans, who won by much smaller margins than the GOP typically does in these districts. But we don’t expect Republicans to see things that way. Instead, we expect these results to give Republicans a morale boost and to be helpful to McConnell’s chances of passing a bill.
The news cycle is slowing down, and health care is the lead story again. Paul, Collins and the other Republican objectors won’t kill the bill outright; most have explicitly said they’re open to negotiations. Those negotiations will take time, however. Meanwhile, we’re entering a moment when the political news cycle is seemingly slowing down, with no more special elections in the immediate future and the Russia story having been in something of a lull for the past week or two. There’s a reason that McConnell drafted his bill in secret: With a bill so unpopular, he doesn’t want his members to be exposed to public scrutiny about it. But now that voters and the media have fewer distractions, some of his senators will get an earful. This factor is harmful to the bill’s chances of passage.

Finally, there’s the wild card of Trump, who reportedly called the House bill “mean” but praised the very similar Senate bill on Thursday. I’m not going to score that in either direction.

So to tally things up, that’s four “helpfuls” and five “harmfuls.” Furthermore, the first “harmful” — that McConnell can afford far fewer defections than Ryan had — is especially important. Overall, the Senate bill faces a somewhat different and probably also more difficult set of obstacles than the House one did. Therefore, I’d guard both against interpretation that the bill will necessarily pass the Senate because it passed the House. At the same time, Ryan and House Republicans overcame some of the same obstacles — and if that precedent isn’t dispositive, it’s at least highly relevant. Among other things, we know that early whip counts aren’t all that reliable and major legislation often follows a bust-and-boom cycle in which it seems to be falling apart only to come together in the end. Neither Democrats nor Republicans ought to be taking much for granted.

Let’s Play The Democratic Blame Game!

In this week’s politics chat, we sift through Democratic finger-pointing after their latest special election losses. The transcript below has been lightly edited.

micah (Micah Cohen, politics editor): Democrats are not happy today, having lost special elections in Georgia and South Carolina on Tuesday. And we’re here to talk about that unhappiness, to really dive deep into it, let it flow over us and try to understand it.

harry (Harry Enten, senior political writer): But what about the mail-in votes?

micah: So we’ll do five questions about the Democratic blame game (which is raging on the interwebs at the moment).

Question No. 1: Blame the blame game! Should Democrats even be playing the blame game?

harry: What’s the purpose of playing the blame game? To help the party succeed in 2018? Or is the purpose to throw blame around because it feels good? If the former, then it is always good to go over what could be done better in the future. If the latter, then it’s not worth it. There were questionable choices that were made by the Democrats in these special elections, but none of them were indefensible in my opinion.

natesilver (Nate Silver, editor in chief): The whole fact that Democrats talked themselves into the narrative that Georgia 6 was a game-changer — and that, therefore, they need to put their tails between their legs today — suggests to me that they’re pretty bad at politics.

perry (Perry Bacon Jr., senior writer): I guess I don’t want to tell Democrats or Republicans or anyone what they should be doing. But does a blame game make sense? When your party is losing key races a lot, some kind of debate about why is natural. I think looking at candidate selection, how the large amount of money raised by the campaign was spent, etc., seems natural. Some of the particular questions in this blame game are not great ones, though. We will get to that later, I suspect.

natesilver: Democrats have been competitive in four substantially red districts. We can debate what “red” means, because the districts are red in different ways. But they’re the sort of results you’d expect in an election where the House was in play. Heck, some of them are even consistent with the sort of results you’d expect in a massive wave election. We’re only five months into Trump’s term.

perry: In other words, Nate, you think there is really nothing to be blamed for? There were not real, unexpected “losses”?

micah: Wait a second, Nate. This was a red district, but it was also an eminently winnable race for Democrats. Sure winning it wouldn’t have been a game-changer, but shouldn’t they be doing some self-examination today? They lost by 4 percentage points in a district Hillary Clinton lost by 1.5 points. I know other non-Trump Republicans have done better there, but if Democrats are hoping to ride dissatisfaction with Trump to the House majority, shouldn’t they be concerned they couldn’t do that in a district we know already doesn’t love Trump?

natesilver: Was Trump on the ballot last night? No. When we modeled the House in 2010, we used a combination of the presidential results in the last two presidential elections, plus the previous House result, plus a bunch of other factors.

Using the last presidential race is just a shorthand that works fine in most cases, but wasn’t particularly good here. And we have a great example of that, given what happened in South Carolina, which was literally the opposite of Georgia in many respects: historically a swingy district that went very pro-Trump. Democrats did quite well there.

harry: I think Micah’s point is the argument why Georgia 6 was bad for Democrats, but I want to get at something I said last night. First off, as Nate points out, we usually use the past two presidential elections (with 2016 weighted more) to understand the lean of a district. Second, back in 2006, there was a district (California 50) that had the same partisan lean as Georgia 6 in the prior two presidential elections. The result was that Francine Busby, the Democrat, lost by pretty much exactly the same margin as Ossoff did last night. Did that mean Democrats were screwed come November 2006? No. They won big. Obviously, that doesn’t mean Democrats will be riding the wave in 2018, but it does show that sometimes these special elections that are built up to be the be-all, end-all aren’t. And keep in mind that the 2006 cycle, like this one, featured Democrats outperforming their baseline by a lot in other places.

natesilver: Certainly, we can debate the strategy in individual races. But basically it’s like if an obscure college football team goes and plays against Ohio State at Ohio Stadium, and loses 30-27 when they were big underdogs going in. It’s disappointing for them, but, at the same time, an indication that the team has bright things in its future and that Ohio State has a lot to worry about.

perry: But what if the coach of team x got the team all hyped up about how it could beat Ohio State? And told the booster club this would be a big win?

micah: Yes ^^^. Which gets back to Democrats being bad at politics.

natesilver: Well, sure. I think Democrats can be blamed for that, to some extent.

micah: It does seem like many people, including us, were telling people this race was a referendum (in part) on Trump, and now they/we are de-emphasizing that.

natesilver: The official FiveThirtyEight pre-election spin was that Georgia 6 mattered more in perception than in reality. Which I think I still agree with.

We were looking to see if a candidate won by 5+ percentage points, which neither one did, although Handel wasn’t far from it. On the other hand, Democrats did much better than we would have guessed in South Carolina.

micah: Hmmm. I think Georgia 6 suggests that marginal Trump voters (whom we’ve dubbed Reluctant Trump voters) — i.e., people who voted for him but had an unfavorable view of him — are still generally with the GOP. That’s backed up by our survey data too.

And that is bad for Democrats.

natesilver: But Micah, the relevant factor is that they were marginal Trump Republicans. If you had a district where you had marginal Trump independents or Democrats — you had a few of those in South Carolina 5 — the outcome might have been different.

perry: I agree with Nate and Harry that last night should not be over-read, by Democrats or, frankly, journalists. But could the Democrats have some kind of internal debate about, say, “should we direct our base towards spending millions of dollars on better causes than House races that will be hard to win?”

natesilver: Yeah, I’m not sure why Democrats felt they had to go narratively all-in on an idiosyncratic district with a weird candidate.

micah: OK, next question. (I’m stealing most of these from Perry.)

Question No. 2: Blame Ossoff! How much was the Georgia 6 result about Ossoff as a candidate and his being too blah or too centrist?

harry: Who knows? is my answer.

natesilver: I’m kind of ambivalent about Ossoff. On the one hand, the basic metric that we usually look at is whether a candidate has been elected in the past, especially to another office in that state or district. Ossoff hadn’t been. On the other hand, I think of his performance as basically having been “fine” in Georgia 6, and he entered the race when few other Democrats were willing to do so.

Put another way, I think he was an average candidate in a race where Democrats could maybe have won with a great candidate, but could also have lost by 12 points with a candidate who wasn’t taking the race seriously.

harry: Yeah, there was nothing that Ossoff did that screamed “awful candidate.” He was milquetoast.

perry: I didn’t meet Ossoff. I wasn’t in Georgia. And I’m always loath to criticize candidates I have not seen in person. I read some more populist Democrats last night claiming the party needed a more populist candidate. Handel emphasized that he did not live in the district. He was obviously young and had little electoral experience. But he didn’t make any major gaffes, seemed to know the issues, got an endorsement from Bernie Sanders and received strong support from John Lewis.

I don’t see him as a clearly bad candidate.

micah: And what do we make of that “Democrats should have run a proud progressive” argument? I’ve seen that argument a lot!

natesilver: I mean, I think you can say he played it a little too safe. He played to not lose.

harry: Now, now, he wanted to connect everyone, Nate.

natesilver: For me, there are basically three prototypes of campaigns that Democrats will need to run in 2018: (i) anti-Trump; (ii) anti-Republican; (iii) anti-incumbent.

I think Georgia 6 ought to have been an anti-Trump campaign, given that Trump is a much bigger liability in Georgia 6 than the GOP overall is and that people are doing pretty well there economically.

For me, there’s lots of room for populist progressives to do well as anti-Republican and anti-incumbent messengers. I actually don’t think they’re ideal as anti-Trump messengers, however, which is what you needed in this district.

perry: I don’t totally think we have any sense if voters, as opposed to Democrats on Twitter, vote differently between a Bernie-Sanders-type Democrat and a Hillary Clinton one or can really tell the difference. Even in Vermont and Massachusetts, I think any Democrat would win those states, not just Elizabeth Warren and Sanders. In fact, both of those states have kind of non-populist Democratic senators right now.

harry: That’s interesting. Keep in mind that this is more of a Bloomingdales district than a Walmart one. It’s a well-off, well-educated district. Further, you know that Archie Parnell fella in South Carolina 5? He’s a former managing director for Goldman Sachs. Doesn’t exactly scream populism.

micah:

Question No. 3: Blame Nancy Pelosi! Repub licans ran a bunch of ads featuring the former Democratic speak er of the House, and som e analysts are asking today if she’s a liability for Democrats in districts like G eorgia 6.

Clare is doing some reporting on this, btw. So I don’t want to steal her thunder, but a lot of people are talking about this so I thought we should include.

perry: Do we really think voters are sitting around thinking about Nancy Pelosi when they go to the polls?

natesilver: Handel gained ground during the stretch run of the campaign, and this was a big part of the GOP’s message.

micah: Doesn’t the fact that she’s featured in so many ads (and also was in 2010 and 2014 if I remember correctly), suggest that Republicans know — or at least think — that it works?

perry: It’s possible we’re just capturing partisanship there.

natesilver: For sure — any Democratic leader would become villainized after a sufficient length of time. But I think there’s something to be said for giving voters symbolically a new look.

I also suspect that this is what’s behind some of the Democratic infighting this morning. The knives are out for Pelosi. She took quite a few defections in the leadership vote late last year, which suggests there’s a lot of dissatisfaction in the Democratic caucus.

micah: But gender could play a role in all this too in addition to her being from San Francisco and all that. But again, Clare is doing some reporting on this, so we’ll get more deeply into it separately.

Question No. 4: Blame the party brand! Is the Democratic brand toxic in certain areas? (Re p. Tim Ryan, who challenged Pe losi, said this in The New York Times ).

perry: Here’s the quote:

Representative Tim Ryan of Ohio, who tried to unseat Ms. Pelosi as House minority leader late last fall, said she remained a political millstone for Democrats. But Mr. Ryan said the Democratic brand had also become “toxic” in much of the country because voters saw Democrats as “not being able to connect with the issues they care about.”“Our brand is worse than Trump,” he said.

heynawl-enten: This is a historically Republican district, so I’m not sure this is the best test for that. Nationally, the Democratic brand is usually stronger than the Republican brand. And more people identify as Democrat than as Republican.

natesilver: Yeah … this is a district where the Democratic brand is their biggest liability, because it’s much more of a Republican district than a Trump district.

The Democratic brand has been surprisingly nontoxic in places such as Montana, South Carolina and Kansas, however.

micah: So why not-as-toxic in those places and more toxic in Georgia 6?

natesilver: Because Georgia 6 is a strongly Republican-identifying place, more so than South Carolina 5 is.

Handel and the GOP figured that out and it got them over the finish line.

perry: OK, I don’t have the data in front of me. But I think the Democratic brand is a problem in the South.

micah: OK, final question …

Question No. 5: Blame the media and national Democratic groups and operatives! This was a theory Harry and Clare floated last night: National Democrats dumped a bunch of money and energy into this race, and we saw record turnout. But did that have the effect of motivating both already motivated Democratic voters and unenthused Republican voters?

perry: In other words, is the South Carolina result a blowout if it became the big race with all of the attention?

micah: Exactly.

perry: I think that right. Turning a race into a nationalized, partisan battle is going to favor the party that has the most voters in that district/state.

harry: Nate thinks Democrats should have spent more money/energy in South Carolina.

natesilver: I do, yeah. Next year is going to be a very expensive, high-stakes election. Not a lot of candidates are going to fly under the radar. Democrats have to get used to competing everywhere.

harry: What’s the purpose of these special elections? To win now? Or to test messages/strategies/tactics for 2018?

natesilver: To test for next year, because the marginal vote in the House isn’t very important right now. (If a Senate seat were in play in a special election, different story.)

micah: OK, so let’s sum things up and then I have one final question:

Who’s to blame? Well, Nate doesn’t think Democrats should even be playing the blame game — they did “fine.”

natesilver: Overall, they should be very excited about their special election results, in fact. Georgia 6 was the worst of the bunch.

perry: They did fine and maybe blaming each other is not that important. But they should be worried that if Republicans campaign hard in every district, spend a lot of money, bash the Democrat enough, mobilize the base, Democrats will have a hard time gaining 24 seats next year.

harry: Overall, I think the special elections are a great sign for the Democrats. However, I always think lessons can be learned about what could have possibly been done differently.

micah: Relative to the weighted past presidential vote that we use, Democrats did worse in Georgia 6 than any other special election so far in 2017. So maybe it’s just a modest outlier in that regard. But Georgia 6 was also the 1. most competitive district to vote so far, and 2. highest profile race. In other words, maybe Georgia 6 is a closer approximation to the conditions in 2018 than other special elections. So, my question: Should Democrats, as the media has, pay more attention to the result in Georgia than to the results in other special elections?

harry: Here’s what I said two months ago, which I still think holds today …

… special House and Senate elections in the two years leading up to a midterm can go any which way. In any of the previous four cycles before a midterm, there’s at least one example of a candidate doing poorly in a special election — relative to the previous weighted presidential vote — only to have their party do well in the midterms.

So even if Ossoff won — even if he won comfortably — it wouldn’t be safe to assume a Democratic wave is building in 2018.

Instead, if you really want to use special election results to look ahead to 2018, you need to look at a bunch of them. While any individual special election may mislead, the average special House and Senate result compared to the past presidential vote provides a decent indication of the national environment heading into the following midterm election.

natesilver: We actually overshot a bit in Georgia 6 in that turnout was higher than it’s likely to be in the midterm, and spending was obviously much higher too. The most representative race of midterm-type conditions so far is probably Montana.

micah: I don’t know. I think people will be bonkers in the midterms.

natesilver: Ballpark, I’d say South Carolina 5 was roughly 60 percent as important as Georgia 6 in terms of helping us to forecast the 2018 midterms. And it’s gotten like 0.6 percent as much attention.

perry: I’m not the data expert here. But my general reaction is to not underreact to things anymore, since I under-reacted to the events of 2015 (the rise of Sanders and Trump.) So this does not change my overall view that Democrats have a strong chance of winning the House in 2018. But I’m not going to dismiss that Handel won by more than I expected and that maybe 2018 won’t follow the patterns of previous midterms.

 

Emergency Politics Podcast: The Results In Georgia And South Carolina

 












Subscribe: Apple Podcasts |
ESPN App |
Download |
RSS |
Embed |
New to podcasts?


Embed Code

The FiveThirtyEight Politics podcast team reacts to the special election results in Georgia’s 6th Congressional District and South Carolina’s 5th. How’d the polls do? And what do the results portend for the 2018 midterms?

You can listen to the episode by clicking the “play” button above or by downloading it in iTunes , the ESPN App or your favorite podcast platform. If you are new to podcasts, learn how to listen .

The FiveThirtyEight Politics podcast publishes Monday evenings, with occasional special episodes throughout the week. Help new listeners discover the show by leaving us a rating and review on iTunes . Have a comment, question or suggestion for “good polling vs. bad polling”? Get in touch by email, on Twitter or in the comments.

Where Can Democrats Win?

The special election in Georgia’s 6th Congressional District on Tuesday night, where Republican Karen Handel defeated Democrat Jon Ossoff by just under 4 percentage points, will have plenty of near-term consequences. Republicans are in the midst of trying to pass a health care bill, and Handel’s win should give GOP members of Congress a confidence boost. Turnout was high in Georgia 6 — higher than in the 2014 midterms — and Republicans may take away the lesson that they can hold on to red-leaning districts by enacting their agenda and turning out their base, even in places where President Trump isn’t all that popular. Is that conclusion correct? Perhaps not, given the health care bill’s unpopularity. But GOP congressional leaders Mitch McConnell and Paul Ryan will take the momentum boost now and deal with the consequences next year.

Meanwhile, Democrats are already engaging in a series of recriminations about why Ossoff faded down the stretch run in Georgia, and whether they should have poured more resources into the special election in South Carolina’s 5th Congressional District instead. (In that election, also held on Tuesday, Democrat Archie Parnell lost by a surprisingly thin 3.2 percentage point margin — narrower than Ossoff’s 3.8-point loss.) Because Republicans used Nancy Pelosi to hammer away at Ossoff, some Democrats could also use the Georgia result to challenge Pelosi’s position as Minority Leader. There could furthermore be knock-on effects on Democratic recruiting and fundraising.

But in terms of the implications for midterm elections in 2018, it’s much less clear that Republicans had a good night. For election forecasting purposes, the margins matter: that Ossoff and Parnell came fairly close to beating their opponents yields a different interpretation than if they’d been blown out. And South Carolina is an important data point, just as Georgia was, even if it received a fraction of as much media attention. The earlier special elections in Kansas’s 4th Congressional District and in Montana’s at-large district provide useful information about the political environment also.

Democrats have gone 0-for-4 in these races. From an emotional standpoint, the outcomes have been disheartening for Democrats. From an analytical standpoint, however, they’ve ranged between “not bad” and “pretty great” for Democrats as compared with their results from the 2012, 2014 and 2016 elections — consistent with the sorts of results Democrats would expect if they were on track to compete for the House next year.

All of these races have been held in substantially red districts, although how you measure redness is a key question for Democrats in how they formulate their strategy for 2018. As compared to the 2016 presidential result, for example, the Georgia 6 outcome was hugely disappointing for Democrats, while South Carolina 5 was a boffo performance. Last year, Clinton lost to Trump in Georgia 6 by only 1.5 percentage points. Since Clinton beat Trump by 2 points in the popular vote nationwide, that meant it was only 3 to 4 points more Republican than the country as a whole. And yet, Ossoff lost to Handel by 3.8 points, seemingly making no progress at all. By contrast, South Carolina’s 5th district was 21 points more Republican than the country overall in last year’s presidential election, so for Parnell to have come within about 3 points of Republican Ralph Norman counts as a huge improvement for Democrats.

And yet … the results aren’t all that surprising if you zoom out and take a wider view. In 2012, Barack Obama came considerably closer to Mitt Romney in South Carolina 5 than he did in Georgia 6. And Republican incumbents were re-elected to the House by wider margins in Georgia 6 than in South Carolina 5 in both 2014 and 2016. South Carolina 5 has also much more recently elected a Democrat to Congress; John Spratt served there until the 2010 midterms, while Republicans have held Georgia 6 since Newt Gingrich’s win there in 1978.

To some extent, Montana — where Democrat Rob Quist lost to Republican Greg Gianforte by 6 points last month — also fits the South Carolina 5 pattern. It went strongly for Trump in 2016, but less so for Romney in 2012 — and Obama nearly won there in 2008. It has also been reasonably competitive in past Congressional races.

REPUBLICAN LEAN IN PARTISAN INDEX*


DISTRICT
G.O.P. MARGIN IN SPECIAL ELECTION
2016 PRESIDENT
2012 PRESIDENT
2016 HOUSE
2014 HOUSE




Kansas 4
+6.8
+29.3
+29.4
+24.6
+22.6


Montana
+6.1
+22.7
+17.6
+8.8
+13.4


Georgia 6
+3.7
+3.6
+27.2
+16.2
+21.2


South Carolina 5
+3.2
+20.6
+15.4
+13.3
+6.8


Average
+5.0
+19.1
+22.4
+15.7
+16.0


Which districts are reddest depends on your benchmark

* Result relative to national popular vote, also adjusted for incumbency in the case of U.S. House incumbents.

One lesson for Democrats would therefore seem to be to look at a mix of indicators for the competitiveness and partisanship of a district, rather than focusing on the 2016 presidential result alone. Trump’s popularity will be a key factor, but so could the long term partisan lean of the district and how it has voted for Congress in the past. Local issues, particularly how the new health care bill might affect the district, could also play a role.

Each of the special elections so far have also come with their quirks: For instance, Georgia 6 had a very high turnout and tens of millions of dollars invested by each party, whereas South Carolina 5 had a much lower turnout and very little investment. But for that very reason — because individual races can be determined by flukish and unpredictable circumstances — Democrats would be wise to avoid the mistake they made in 2016, when Clinton campaigned in too narrow a range of states and didn’t properly consider the uncertainty in the outcome. Well-educated Sun Belt districts such as Georgia 6 could be the Democrats’ path back to a majority. But so could places such as South Carolina 5 that had once been more Democratic, or districts in Ohio or Pennsylvania or upstate New York. It’s much too early to know.

The historical constant is that midterm elections have usually yielded a backlash against the president’s party — and the results of the special elections so far suggest that Republicans’ risk is higher than average.

As compared to the 2016 presidential results, Democrats have outperformed their benchmarks by an average of 14 percentage points so far across the four GOP-held districts to have held special elections to date. As compared to the 2012 presidential election, their overperformance is even larger, at almost 18 points. They’ve also outperformed their results from the 2016 and 2014 U.S. House elections by roughly 11 points, after one accounts for the fact that the special elections were open-seat races rather than being held against incumbents.

DEMOCRATIC SWING IN SPECIAL ELECTION RELATIVE TO BENCHMARK*


DISTRICT
2016 PRESIDENT
2012 PRESIDENT
2016 HOUSE
2014 HOUSE




Kansas 4
+22.5
+22.6
+17.8
+15.8


Montana
+16.6
+11.5
+2.7
+7.3


Georgia 6
-0.1
+23.5
+12.5
+17.5


South Carolina 5
+17.4
+12.2
+10.1
+3.6


Average
+14.1
+17.5
+10.8
+11.1


Democrats continue to substantially outperform their benchmarks

* Result relative to national popular vote, also adjusted for incumbency in the case of congressional incumbents.

How might this translate for Democrats next November, when all 435 seats are up for grabs? The results simultaneously suggest that an impressively wide array of Republican-held seats might be competitive next year — perhaps as many as 60 to 80 — and that Democrats are outright favorites in only a fraction of these, perhaps no more than a dozen. To some extent, this configuration is a result of Republican-led gerrymandering in 2010. Republicans drew a lot of districts where their members are safe under normal conditions, but not in the event of a massive midterm wave.

In order to win a net of 24 seats next year — enough to flip the House — Democrats may therefore need to target dozens of Republican-held seats and see where the chips fall. They can variously attempt anti-Trump, anti-Republican or anti-incumbent messages depending on the district.

Related:

The 2018 midterms will be strange in that a “pretty good” year for Democrats might yield a gain of only 15 seats for the party, whereas a “very good” year — if the political climate is just a few points more Democratic-leaning — could produce a 50-seat swing instead. Because they run the government, Republicans and Trump will have more influence on the macro-level political environment. But Democrats will have at least as much say on the district-level environment based on where they recruit strong candidates instead of giving Republican incumbents a free pass.

In that sense, Tuesday night poses the biggest risk to Democrats if it discourages them from recruiting quality candidates and providing them with enough money to run credible campaigns in districts as diverse as Georgia 6, South Carolina 5 and Montana. And the biggest risk to Republicans will be if they get too caught up in the narrative and ignore that the results in special elections so far indicate a lot of downside risk for the GOP next year.

Politics Podcast: The Most Special Special Election

 












Subscribe: Apple Podcasts |
ESPN App |
Download |
RSS |
Embed |
New to podcasts?


Embed Code

This week on the show, the FiveThirtyEight politics podcast team games out the special election in Georgia’s 6th Congressional District. Is the race between Democrat Jon Ossoff and Republican Karen Handel really that big a deal? Plus, in “good use of polling vs. bad use of polling,” Nate explains that chocolate milk does actually come from brown cows. Finally, they speculate on the possibility of a Mike Pence presidency and what it would mean for the Republican Party.

You can listen to the episode by clicking the “play” button above or by downloading it in iTunes , the ESPN App or your favorite podcast platform. If you are new to podcasts, learn how to listen .

The FiveThirtyEight Politics podcast publishes Monday evenings, with occasional special episodes throughout the week. Help new listeners discover the show by leaving us a rating and review on iTunes . Have a comment, question or suggestion for “good polling vs. bad polling”? Get in touch by email, on Twitter or in the comments.