Taking too long? Close loading screen.
Connect with us


Our Election Forecast Didn’t Say What I Thought It Would



My editors are forever asking me to take the long Twitter threads I write and turn them into articles here at FiveThirtyEight. So I’m actually going to give that a try!

What follows are some follow-up thoughts on our election model, which was originally composed in the form of a V E R Y L O N G tweetstorm that I never published. (See if you can guess where the 240-character breaks would have been.)

In this thread … err, article … I’ll try to walk you through my thought process on a few elements of our model and respond to a few thoughtful critiques I’ve seen elsewhere. Before you dive in, it may help to read our summary of the state of the race, or at least skim our very detailed methodology guide.

But the basic starting point for a probabilistic, poll-driven model ought to be this: Is polling in August a highly reliable way to predict the outcome in November?

The short answer is “no.”

Polling in August is somewhat predictive. You’d much rather be ahead than behind. But there can still be some very wild swings.

You can see that in the daily threads that Nathaniel Rakich, one of our elections analysts, puts together. Here is what a national polling average would have looked like in elections dating back to 1976:

OK, I cheated a bit. I’m using a version that Nathaniel published last week, partly because this was the exact moment in the campaign when Michael Dukakis, the 1988 Democratic nominee, started to blow his large lead, which he never regained. Still, there’s some wild stuff there! John Kerry led at this point in 2004. George W. Bush had a 10-point lead at this point in the 2000 race, but, as we know, he didn’t win the popular vote that year. In other cases, the leading candidate won, but the margin was off by as much as 20 points (Jimmy Carter in 1976).

Now, as I wrote last week, there are some caveats here. Several of these polling averages were taken while one or both candidates were experiencing convention bonuses, and although there are ways to correct for those, every time you correct for something so your model fits the past data better, you raise the possibility that you’re overfitting the data and that your model won’t be as accurate as claimed when applied to situations where you don’t already know the outcome.

There are also decent arguments that polling averages have become more stable in recent years. In that case, the wild fluctuations in the polls from, say, 1976 or 1988 might not be as relevant.

Our model actually agrees with these theories, up to a point! The fact that voters are more polarized now (more polarization means fewer swing voters, which means less volatility) is encoded into our model as part of our “uncertainty index,” for instance.

But we think it’s pretty dangerous to go all in on these theories and assume that poll volatility is necessarily much lower than it was before. For one thing, the theory is not based on a ton of data. Take the five most recent elections, for instance. The 20041 and 2012 elections featured highly stable polling — 2012 especially so. But 2000 and 2016 (!) did not, and 2008 election polling was not especially stable, either. Small sample sizes are already an issue in election forecasting, so it seems risky to come to too many firm conclusions about polling volatility based on what amounts to two or three examples.

Meanwhile, other people have pointed out that the most recent two presidents, Trump and Barack Obama, have had highly stable approval ratings. But the president just before them, George W. Bush, did not. His approval rating went through some of the wildest fluctuations ever, in fact, even though polarization was also fairly high from 2000 to 2008.

That said, polls have been stable so far this year. Indeed, that’s another factor that our uncertainty index accounts for. But don’t get too carried away extrapolating from this stability. Case in point: Polls were extremely stable throughout most of the Democratic primaries … but when the voting started, we saw huge swings from the Iowa caucuses through Super Tuesday. Poll volatility tends to predict future volatility, but only up to a point.

Remember, too, that voters haven’t yet been exposed to the traditional set pieces of the campaign, namely the conventions and the debates, which are often associated with higher volatility.

Now, suppose that despite all the weirdness to come in the general election campaign, Biden just plows through, leads by 6 to 9 points the whole way … and then wins by that amount on Nov. 3? If that happens, then we’ve got more evidence for the hypothesis that elections have become more stable, even when voters are confronted with a lot of surprising news.

But, crucially, we don’t have that evidence yet. So some of the models that are more confident in Biden’s chances seem to be begging the question, presuming that polls will remain stable when I’m not sure we can say that yet.

Then there’s the issue of COVID-19. Sometimes — though people may not say this outright — you’ll get a sense that critics think it’s sort of cheating for a model to account for COVID-19 because it’s never happened before, so it’s too ad hoc to adjust for it now.

I don’t really agree. Models should reflect the real world, and COVID-19 is a big part of the real world in 2020. Given the choice between mild ad-hockery and ignoring COVID-19 entirely, I think mild ad-hockery is better.

However, I also think there are good ways to account for COVID-19 without being particularly ad hoc about it. If you’re designing a model, whenever you encounter an outlier or an edge case or a new complication, the question you ask yourself should be, “What lessons can I draw from this that generalize well?” That is: Are there things you can do to handle the edge case well that will also make your model more robust overall?

As an aside, when testing models on historical data I think people should pay a lot of attention to edge cases and outliers. For instance, I pay a lot of attention to how our model is handling Washington, D.C. Why Washington? Well, if you take certain shortcuts — don’t account for the fact that vote shares are constrained between 0 and 100 percent of the vote — you might wind up with impossible results, like Biden winning 105 percent of the vote there. Or when designing an NBA model, I may pay a lot of attention to a player like Russell Westbrook, who has long caused issues for statistical systems. I don’t like taking shortcuts in models; I think they come back to bite you later in ways you don’t necessarily anticipate. But if you can handle the outliers well, you’ve probably built a mathematically elegant model that works well under ordinary circumstances, too.

But back to COVID-19: What this pandemic encouraged us to do was to think even more deeply about the sources of uncertainty in our forecast. That led to the development of the aforementioned uncertainty index, which has eight components (described in more depth in our methodology post):

  1. The number of undecided voters in national polls. More undecided voters means more uncertainty.
  2. The number of undecided plus third-party voters in national polls. More third-party voters means more uncertainty.
  3. Polarization, as measured elsewhere in the model, which is based on how far apart the parties are in roll call votes cast in the U.S. House. More polarization means less uncertainty since there are fewer swing voters.
  4. The volatility of the national polling average. Volatility tends to predict itself, so a stable polling average tends to remain stable.
  5. The overall volume of national polling. More polling means less uncertainty.
  6. The magnitude of the difference between the polling-based national snapshot and the fundamentals forecast. A wider gap means more uncertainty.
  7. The standard deviation of the component variables used in the FiveThirtyEight economic index. More economic volatility means more overall uncertainty in the forecast.
  8. The volume of major news, as measured by the number of full-width New York Times headlines in the past 500 days, with more recent days weighted more heavily. More news means more uncertainty.

Previous versions of our model had basically just accounted for factors 1 and 2 (undecided and third-party voters), so there are quite a few new factors here. And indeed, factors 7 and 8 are very high thanks to COVID-19 and, therefore, boost our uncertainty measure. However, we’re also considering several factors for the first time (like polarization and poll volatility) that reduce uncertainty.

Source : FiveThirthyEightRead More

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *


Toronto FC hoping to make MLS Cup run having spent much of 2020 far from home



On a recent Thursday in Hartford, Conn., Toronto FC goalkeeper Quentin Westberg pondered the dichotomy of wanting to reach MLS Cup on Dec. 12, but also desiring to see his family again. Meanwhile, Jim Liston, the team’s director of sports science, was planning a trip to Lowe’s to buy 15 garbage cans so players could have an ice bath after training. As for manager Greg Vanney, he was fretting about his team’s health and the lack of practice time their schedule was affording.

Such is the life of a team as it attempts to not only navigate its way through the COVID-19 pandemic, but has been forced to do it away from home.

Due to travel restrictions between the U.S. and Canada, TFC — like the league’s other two Canadian teams, Montreal Impact and Vancouver Whitecaps — set up a “home” base in the U.S. for the remainder of the season; Toronto were stationed in Hartford. (Vancouver Whitecaps took roost in Portland, ground-sharing with Timbers, while Montreal Impact split use of New York Red Bulls’ facilities in Harrison, N.J.) This was on top of nearly every team spending nearly a month inside a bubble back in July at the MLS is Back Tournament outside Orlando, Florida.

The Reds spent about seven weeks back in Toronto as they played a series of matches against Canadian teams. In mid-September, the remainder of the regular season — and the temporary move to Hartford — beckoned. The vagabond nature of the campaign is what led Liston to joke that he was willing to discuss “whatever five seasons” the team has been through so far. But for Vanney and the players, the campaign has required a special kind of focus.

“A lot of what we’ve done here, and what we try to preach here is just control the controllables, and don’t get too drawn into the things you can’t,” Vanney told ESPN. “Roll with it, and make the best out of whatever the situation is.”

Stream FC Daily on ESPN+
– 2020 MLS Playoffs: Who’s in, schedule and more
– MLS on ESPN+: Stream LIVE games and replays (U.S. only)

Toronto has largely succeeded in spite of its odyssey. While there was disappointment at missing out on the Supporters’ Shield to the Philadelphia Union, TFC went 7-3-2 during its Hartford sojourn and finished with the second-best record in the league. But the challenges have still been immense. Simply being out of one’s home environment is difficult enough, but the time spent away from family and loved ones weighs heavy on the psyche, even as Vanney has given players the occasional trip back to Toronto — under quarantine — to reconnect with loved ones.

“It’s just very different, very challenging and emotionally exhausting,” Westberg said of his experience while based in Hartford.

Westberg has arguably had it tougher than most. The TFC goalkeeper is married with four children, including a baby girl who was born in June. For that reason, Westberg and his wife, Ania, made the decision at the end of September that it would be better for her and their kids to head back to his native France so they could be surrounded by family. Westberg called it “the least bad decision,” but there are difficulties nonetheless.

“I’m a very even person, and this year has challenged me a lot,” he said. “I’m still pretty even, but I keep a lot to myself and for sure there’s some difficult days, seeing your family [struggle] from your absence.”

The inability to be home has affected the players and staff in other ways. In Toronto, there are ways of disengaging from the game. Being with friends, loved ones or even in familiar surroundings can be the best medicine in terms of forgetting a bad game or training session. But in Hartford, at the team’s hotel, that escape is nearly impossible even as players try to distract themselves by reading or taking online classes.

“You don’t really unplug,” Westberg said. “You FaceTime family, or this or that, but it’s too short. You’re 100 percent focused on your soccer, and your whole day basically relies on being ready for whatever soccer activity that you have next, whether it’s practice or game. It’s good for your physique, it’s optimal for the way you eat and the way you [train]. But mentally, you’re not as fresh as your body.”

That isn’t to say there are only negatives to the separation. There is also an us-against-the-world mentality that Toronto has adopted, given that their players and personnel are experiencing the season in a way that is vastly different than most other teams. The team staff has done what it can to make their surroundings a home away from home, whether it’s personalizing the locker rooms at Rentschler Field or having hotel staff brand the surroundings in TFC colors. The hotel went so far as to bring in a barista who could consistently give the players their coffee fix. Supporters groups have even sent down banners in a bid to convey the fact that the players are remembered.

The care that TFC takes for players has extended to families back home, with the club supplying meals to loved ones three times a week.

On the logistical side, Liston made sure that one of the gyms used at MLS is Back was brought to TFC’s hotel in Hartford, and he remarked that the food at the hotel is “arguably the best we’ve ever had on the road.”

There have also been efforts to create new routines. Assistant coach Jason Bent, aka DJ Soops, has been in charge of the pregame music selection for the past 18 months — no easy feat for a squad that has a considerable international presence. In Hartford, Bent has set aside Thursday nights to spin music in one area of the hotel. He’ll even go live on Instagram or Twitch for those who prefer to relax in their rooms.

“[We] opened it to players and staff and basically anyone that’s part of our bubble to come relax, listen to music and just enjoy each other’s company,” Bent said. “I enjoy making people happy so if it’s helping everyone even in the slightest, I have no problem arranging the set and spinning.”

For Vanney, the pandemic and operating outside of the team’s home market has meant any number of challenges. He said the team has used three different training facilities in Hartford, with varying field conditions. He recognizes that the trips home are vital for the mental health of his players and staff, but any breaks also mean less time spent on the practice field. The compressed schedule, which at times involved games every three or four days, has had an impact as well. Even the best-laid plans in terms of squad rotation were impacted as minor injuries began popping up.

“We end up with a lot of guys in different positions because they need special kinds of treatment or care to help them get fit and back to health,” Vanney said. “So it ends up being a lot of different things kind of going on all at once, and that’s been the challenge of it.”

Recovery from matches has been complicated by the fact that TFC doesn’t have access to the same level of facilities that it does at home — hence Liston’s emergency trip to Lowe’s to fashion impromptu ice baths for the players. Then there are the different ways the players occupy themselves on the road as compared to home, especially amid the pandemic.

“There’s really no life outside of the hotel,” Liston said. “[At home], you may go walk the dog in the afternoon or go for a walk with your wife or friend or girlfriend or family and you’re out and about. The recommendation [here] is to kind of stay put. So you’ve got a really active population and pro athletes, who we’re asking them to be sedentary the rest of the time, kind of stay in the hotel from a COVID and safety standpoint. That’s not optimal for recovery either.”

There are also the creature comforts of home that are no longer available on the road, which can impact sleep.

“Sleep is the number one tool for recovery, and that’s definitely been a challenge,” Liston said. “We do well-being questionnaires and the scores on quality of sleep, and hours of sleep, just drop.”



Tom Barlow and Brian White seal Toronto’s fate in a 2-1 win for New York Red Bulls. Watch MLS on ESPN+.

Another change has been same-day travel, which has drawn mixed reactions from the TFC players and staff. Vanney and Westberg are generally in favor, saying it reminds them of when they each played in France. Flying back the same night also means a training day isn’t lost. Liston has a different perspective in that he prefers arriving the day before, and then leaving the same day.

“I think [same-day travel] makes for a really long day,” he said. “And there’s definitely a negative impact on performance, taking three bus rides and a plane ride before your game. You’re getting home — it can be 12:30, but it could also be 1:30 in the morning, and that’s where you know our well-being scores and sleep hours and quality just disappear. When you have so many games in succession, you can’t make up the sleep.”

With the playoffs set to begin for TFC on Nov. 24, the end is in sight, even as it makes for a complex — and even conflicting — set of emotions.

“This is the tricky part. I miss them a lot,” Westberg said of his family. “But in a way I want to see them as [late] as possible in December, because obviously, there’s this idea that we want to do well in the playoffs and we want to keep going. TFC has a history of setting high standards and high expectations. It’s a heavy load to carry but also an exciting one.”

Win or lose, it’s a season they’ll never forget.


Continue Reading


Bettman: NHL is mulling temporary realignment



The NHL is considering a temporary realignment of its teams for the 2020-21 season due to COVID-19 travel restrictions, according to commissioner Gary Bettman.

Bettman said Tuesday that restrictions on travel across the Canadian border, as well as “limitations in terms of quarantining when you go from certain states to other states” within the United States, could mean the NHL creates a more regionalized alignment for its upcoming season.

“As it relates to the travel issue, which is obviously the great unknown, we may have to temporarily realign to deal with geography, because having some of our teams travel from Florida to California may not make sense. It may be that we’re better off — particularly if we’re playing a reduced schedule, which we’re contemplating — keeping it geographically centric and more divisional-based; and realigning, again on a temporary basis, to deal with the travel issues,” Bettman said during a 2020 Paley International Council Summit panel with fellow commissioners Adam Silver of the NBA and Rob Manfred of MLB.

The NHL board of governors has a meeting scheduled for Thursday which will provide a progress report and possible recommendations for a season format, based on talks between the league and the NHL Players’ Association. The target date for starting next season remains Jan. 1.

Bettman said the league is considering a few scheduling options for the 2020-21 season. Something that’s off the table: playing the entire season in the kind of bubbles the NHL had in Toronto and Edmonton, Alberta, to complete last season. But Bettman said teams opening in their own arenas is a possibility, along with a modified bubble.

“We are exploring the possibility of playing in our own buildings without fans [or] fans where you can, which is going to be an arena-by-arena issue. But we’re also exploring the possibility of a hub. You’ll come in. You’ll play for 10 to 12 days. You’ll play a bunch of games without traveling. You’ll go back, go home for a week, be with your family. We’ll have our testing protocols and all the other things you need,” he said.

Bettman also indicated that the NHL is exploring “a hybrid, where some teams are in a bubble, some teams play at home and you move in and out.”

The NBA’s board of governors unanimously approved a deal with the players’ union that sets the stage for a season that will open on Dec. 22 and with a reduced schedule of 72 games. Silver said that the commissioners are in communication on COVID-19-related issues, especially the NBA and the NHL, since the two leagues’ teams share arenas and, in some cases, team owners.

Silver said he senses that the NBA will have fans in many of its buildings this season.

“We’re probably going to start one way, where we’re maybe a little bit more conservative than many of the jurisdictions allow,” he said. “What we’ve said to our teams is that we’ll continue to work with public health authorities. Arena issues are different than outdoor stadium issues. There will be certain standards for air filtration and air circulation. There may be a different standard for a suite than there will be for fans spaced in seats.”

Silver said there will be standardized protocols that are consistent from arena to arena, such as proximity between players and fans: “In certain cases, for seats near the floor, we’re going to be putting in testing programs, where fans will certify that they’ve been tested — some within 48 hours, some within day of game.” While Silver supported a continued expansion of the NBA postseason through its play-in tournament, Bettman said that he’s not in favor of expanded playoffs or “playing with the fundamentals of the game.” The NHL had 24 teams in its postseason last summer.


Continue Reading


The Battleground States Where We’ve Seen Some Movement In The Polls



With apologies to The Raconteurs, the presidential race continues to be “steady as she goes,” with little sign of tightening despite a plethora of new polls. FiveThirtyEight’s presidential forecast gives Joe Biden an 89 in 100 shot at winning the election, while President Trump has just an 11 in 100 chance. This makes Biden the favorite, but still leaves open a narrow path to victory for Trump, for whom a reelection win would be surprising — but not utterly shocking.

At the same time, we also have fewer polls from live-caller surveys, which have historically been more accurate and have shown slightly better numbers for Biden, than polls that use other methodologies, such as polls conducted primarily online or through automated telephone calls. Nevertheless, while the overall picture has shifted only a little in recent days, a few battleground states have seen at least some movement in their polls, which has slightly altered the odds Biden or Trump wins in each of those places.

What election stories need to get more coverage | FiveThirtyEight Politics Podcast


Continue Reading