The Cleveland Cavaliers haven’t been any good lately. And I don’t just mean their loss Sunday against the Atlanta Hawks, in which they became only the third team in NBA history to blow a 26-point fourth-quarter lead. They’re 12-13 since the All-Star break. They have one of the NBA’s worst defenses, having allowed 107.9 points per 100 possessions — in the same territory as the Orlando Magic and the New York Knicks. They haven’t won a road game against a Western Conference playoff team all season. But handicappers think LeBron James and company have a pretty good chance of winning their second-straight NBA title anyway.
Their view depends on their belief in the existence of Playoff LeBron, a superhero that transcends his already-formidable regular season form to carry his team to ever-greater heights. The good news for Cavs’ fans is that Playoff LeBron exists. He just might not be mighty enough to drag this team to a title.
On the basis of their regular-season record and point differential, this season’s Cavs have been in the same general vicinity as teams such as the Boston Celtics, Toronto Raptors and Utah Jazz. Those teams are variously 30-to-1 to 100-to-1 longshots to win the title, according to Vegas bookmakers. But the Cavs are nonetheless the second-favorite team to win the championship, with a 20 to 25 percent chance according to bookmakers.
Computer systems disagree. All of them have the Warriors as odds-on favorites to win the title, with the San Antonio Spurs as the next-best bet, and the Cavs as part of an undistinguished mass of teams beneath them. ESPN’s BPI puts Cleveland’s chances at just 4 percent. Basketball-Reference’s playoff odds also have them at 4 percent. And FiveThirtyEight’s Elo-based ratings1, which heavily weight recent play, have them even lower at just 2 percent.
Usually, Elo-type ratings mimic betting markets fairly well. We give the Warriors a 65 percent chance of winning the title, for instance, and the San Antonio Spurs an 11 percent chance — right in line with where markets have them. So what accounts for the huge difference on Cleveland?
One explanation is that this is all just sort of irrational: the Cavs are a marquee team and bettors just can’t stomach the idea that they’re just the Raptors with better uniforms. But I’m not sure I totally buy that; NBA betting markets are usually fairly sharp.
Instead, bettors expect the Cavs to find a higher gear in the postseason. This isn’t an idea they just came up with; it was already priced into their assessment of the Cavs before the year began. At the start of the NBA season, FiveThirtyEight’s projections forecast the Cavs to win 57 regular-season games. (They have 51 now, so they’ll finish with no more than 53 wins.) That forecast was almost the same as what Vegas gave them, which put their over-under at 56.5 wins. But we also gave the Cavs only an 11 percent chance of winning the title whereas Vegas put them at 5-to-2 against, or a 29 percent chance. In other words, handicappers and the computer models agree on “regular-season Cavs.” It’s just that Vegas thinks that “playoff Cavs” are different — and much better — whereas our Elo ratings make no such distinction.
But is there good reason to think that Cleveland can turn it up a notch?
It’s not hard to recall examples of defending champions that lollygagged their way through the regular season, only to show up as the best version of themselves in the playoffs. In 2000-01, Shaquille O’Neal, Kobe Bryant and the Los Angeles Lakers finished with a 56-26 record — better than the Cavs this year, but not by that much — before winning 15 of 16 playoff games and repeating as NBA champion. And Hakeem Olajuwon and the 1994-95 Houston Rockets finished at 47-35 before winning the title despite being the No. 6 seed. In his last season in Miami, James and the 2013-14 Miami Heat had an uninspired regular season, going 54-28. But they made the NBA finals before losing to San Antonio.
I’ve done a bit of cherry-picking there, however. Overall, it’s not clear if defending champs overperform by much in the playoffs. In the table below, I’ve looked at every defending NBA champion since the ABA-NBA merger in 1976-77. If the “higher gear” theory is correct, then they should systematically beat Elo’s expectations in the postseason, in which case their Elo ratings will rise over the course of the playoffs. On average, however, these teams’ Elo ratings increased only from 1644 to 1658 during the playoffs.2 So there’s a little something there, but in Elo terms, that’s pretty minor — not much more than a rounding error.3
|ELO RATING IN PLAYOFFS|
|2016||Warriors||73-9||+10.4||Lost NBA Finals||1788||1756||-32|
|2015||Spurs||55-27||+6.3||Lost 1st Round||1733||1721||-12|
|2012||Mavericks||36-30||+1.8||Lost 1st Round||1547||1525||-22|
|2011||Lakers||57-25||+6.0||Lost 2nd Round||1659||1624||-35|
|2009||Celtics||62-20||+7.4||Lost 2nd Round||1693||1653||-40|
|2008||Spurs||56-26||+5.1||Lost Conf. Finals||1662||1678||+16|
|2007||Heat||44-38||-1.2||Lost 1st Round||1505||1479||-26|
|2006||Spurs||63-19||+6.7||Lost 2nd Round||1685||1675||-10|
|2005||Pistons||54-28||+3.3||Lost NBA Finals||1613||1689||+76|
|2004||Spurs||57-25||+7.5||Lost 2nd Round||1734||1719||-15|
|2003||Lakers||50-32||+2.7||Lost 2nd Round||1655||1651||-4|
|2000||Spurs||53-29||+5.9||Lost 1st Round||1637||1625||-12|
|1996||Rockets||48-34||+1.6||Lost 2nd Round||1485||1497||+12|
|1994||Bulls||55-27||+2.9||Lost 2nd Round||1575||1607||+32|
|1991||Pistons||50-32||+3.1||Lost Conf. Finals||1552||1535||-17|
|1989||Lakers||57-25||+6.4||Lost NBA Finals||1637||1677||+40|
|1987||Celtics||59-23||+6.6||Lost NBA Finals||1676||1659||-17|
|1986||Lakers||62-20||+6.8||Lost Conf. Finals||1652||1647||-5|
|1985||Celtics||63-19||+6.5||Lost NBA Finals||1668||1685||+17|
|1984||76ers||52-30||+2.4||Lost 1st Round||1587||1572||-15|
|1982||Celtics||63-19||+6.4||Lost Conf. Finals||1686||1703||+17|
|1981||Lakers||54-28||+3.3||Lost 1st Round||1610||1595||-15|
|1980||SuperSonics||56-26||+4.2||Lost 2nd Round||1639||1614||-25|
|1978||Trail Blazers||58-24||+5.9||Lost 1st Round||1558||1551||-7|
Forget looking at defending champions, though. The more important variable, as far as sports bettors are probably concerned, is LeBron. Between his experience, his toughness, and his ability to thrive in crunch-time situations, he has a game well tailored to the playoffs. And that shows up in the data:
|ELO RATING IN PLAYOFFS|
|2010||Cavaliers||61-21||+6.5||Lost 2nd Round||1701||1646||-55|
|2009||Cavaliers||66-16||+8.9||Lost Conf. Finals||1725||1742||+17|
|2008||Cavaliers||45-37||-0.3||Lost 2nd Round||1506||1562||+56|
|2006||Cavaliers||50-32||+2.2||Lost 2nd Round||1562||1564||+2|
James’ teams have made the playoffs 11 times prior to this season. And they’ve played really well, both in absolute terms and relative to their regular-season performance. James and the Cavs did have a disastrous postseason in 2010 — when, as the No. 1 overall seed, they lost to the Celtics in the second round — but that’s pretty much the only exception. On average, they’ve gained 34 Elo points from the start of the playoffs to the end. And over James’s past six postseasons, they’ve outperformed their regular-season ending Elo rating by an average of 55 points.
So let’s say that Elo has the Cavs’ underrated by somewhere in the neighborhood of 50 points. Call that a “LeBron clutch factor” or whatever else you like. I asked my colleague Jay Boice to add 50 Elo points to the Cavs’ Elo rating and rerun our playoff simulations. Their championship odds rose … but only to 6 percent.
Instead, you have to add about 150 points of Elo rating to get the Cavs’ odds in the same vicinity as Vegas has them.4 That’s a lot. Elo sees the Cavs’ current level of performance as equivalent to a 48-34 regular-season record. Add 150 Elo points to that total, and they’d project to a 62-20 regular season record. That’s a 14-win gain — about what you’d get from adding someone like Kawhi Leonard or Anthony Davis to the roster.
Have no doubt: I’d love to plunk some money down on the Cavs at the odds our forecast and the other computer models give them. Playoff basketball is a pretty different specimen from regular-season basketball, and our model isn’t doing anything to account for that. This is something for us to examine for future iterations of the model, even if the Cavs get bounced in the first round.
But I also wonder if the bookies aren’t going too far in the other direction. There are plenty of defending champions — and James-led teams — that underwhelmed in the regular season before going on to win a title or at least reach the finals. But few of them underperformed as much as the Cavs have. They also tended to benefit from down periods in the league, as the 1994-95 Rockets and 2000-01 Lakers did. This year, the Cavs will have to get past the Warriors, who might be even better than last year’s 73-9 version5, or, failing that, probably the Spurs.
Nor will the Cavs’ enter the postseason with much rest. Instead, as the East’s No. 1 overall seed has been up in the air between the Cavs and the Celtics, James has averaged 43 minutes per over the team’s last five games. Kyrie Irving has gotten only two days off since the All-Star break. Kevin Love has played heavy minutes despite missing time in February and March due to knee surgery.
James has beaten expectations so many times in the playoffs that transcendent things are almost expected from him. If he leads the Cavaliers to another title this year it really might be his greatest accomplishment yet.