Let’s play FiveThirtyEight Family Feud. Twenty seconds on the clock.
Name a famous Super Bowl upset.
Name a famous NBA Finals upset.
I mean… umm, uh.
That isn’t an easy question. Didn’t Bill Russell’s Celtics once lose in the finals? (They did, exactly once, in 1958 to the St. Louis Hawks.) Wasn’t it kind of an upset when Dwyane Wade and Shaq led the 52-30 Miami Heat to a title in 2006? (It was, though not by much.) The NBA doesn’t lend itself to upsets. When each team has around 100 possessions per game, small differences tend to add up. And the differences really compound over a seven-game series. This has been particularly true in the NBA Finals. Whether by chance or because of the conditions under which finals games are played, underdogs have had an especially low success rate.
The 2015 NBA Finals, which begin Thursday night in Oakland, look like a compelling matchup. Our Elo ratings provisionally rate the Golden State Warriors as the third-best NBA team of all time based on their performance to date. (There’s no big secret here. A 79-18 record — that counts the playoffs — is pretty amazing against this season’s Western Conference competition.) But the Cleveland Cavaliers’ Elo rating has been rising and is at its high point for the season, 1712. Only three finals matchups started with a higher combined Elo rating, and all three involved Michael Jordan’s Bulls.
Maybe LeBron James really is good enough to overcome a middling supporting cast. But his odds aren’t great. The Cavaliers would be only about a 3-point underdog on a neutral court, according to Elo. Over a seven-game series, however — and with the Warriors having home-court advantage in a potential Game 7 — that adds up to just a 25 percent chance of the Cavs winning the series.improves the underdog’s (Cleveland’s) chances slightly. Without recalculating Elo ratings at the end of each game, Golden State’s chances would be 79 percent instead of 75 percent.">1 (That matches the odds according to another statistical system we’ve been using to handicap the NBA.) And that may be optimistic if history is any guide.
Elo’s a really simple formula, so there’s a lot it doesn’t account for. You can make a case — we could argue about this for a long time — that past NBA Finals experience matters and could help James. Any lingering effects from the injuries sustained by Stephen Curry and Klay Thompson could matter a great deal to the Warriors, of course. Then again, Cleveland’s also pretty beat up,Ewing Theory?">2 and James’s superpowers were nowhere near enough against the San Antonio Spurs last year.
The other thing is that underdogs have historically been bad bets in the NBA Finals. Here’s every NBA Finals matchup in history, with each team’s Elo rating going into the series and the probability Elo would have assigned to each team winning the series beforehand.
It’s not just your lack of imagination: It really is hard to find monumental upsets in the NBA Finals. By Elo’s reckoning, the biggest one came in 1974, when the Boston Celtics beat the Milwaukee Bucks in seven games. That’s mostly because the Celtics looked worn down during the stretch run, finishing the regular season 27-20 in their final 47 games and deflating their Elo rating. (Plus, the Bucks were no fluke, with both Kareem Abdul-Jabbar and Oscar Robertson on the roster.)
Overall, however, of the 39 series in which Elo would have given one team at least a 2-in-3 chance of winning, the favorite in fact won 35 times, or almost 90 percent of the time.
Why such a high success rate? It could be a statistical fluke; we aren’t looking at all that large a sample. But this phenomenon isn’t unique to the NBA: It also holds for the NFL playoffs, we’ve found. Elo ratings treat regular-season games and playoff games the same. But in both the NBA and NFL, favorites tend to be more dominant in the playoffs, especially late in the playoffs, than they are in the regular season.
The reason may be that some of the “noise” that affects teams in the regular season is absent in the playoffs. No team is coming off a back-to-back, for instance. Teams are going all-out to win, instead of potentially testing out new strategies or resting starters. And at least in theory — although there a lot of NBA fans who would dispute this when refs like Joey Crawford are often involved — the games are adjudicated by the best officials. When you remove some of the quirky circumstances that can cost teams games — bad refs, funky schedules — the best teams tend to prevail more often.
That’s bad news for James, but it will add all the more to his legacy if the Cavs pull the upset off.