The Westgate Las Vegas SuperBook released its win totals on Sunday for the 2018-19 NBA season, two days after we published our real plus-minus (RPM) projections. As expected, the two measures are quite similar. More than half of the league's teams have a Vegas total within two games of their expected average-wins total from RPM.
Yet inevitably, there are places where RPM diverges from Westgate's efforts to balance the betting on both sides of the total -- starting with the Los Angeles Lakers, whose win total for 48.5 is more than seven wins better than their 41.3-win RPM projection.
What can we take from the differences between Westgate's lines and RPM, and how does it help us locate the betting value?
How RPM has fared relative to totals
We've been publishing RPM projections for the past three seasons. Over that span, just taking the side of the win total favored by RPM has resulted in wins 58 percent of the time. The results get stronger at the extremes. Here are the 10 cases in which team totals exceeded their RPM projections by at least five wins over that span:
Seven of the 10 teams indeed went under, often dramatically so. On the other side of the ledger, teams with RPM projections topping their totals by at least five wins haven't fared quite as well:
While these teams have tended to go way over when they've hit, winning an average of 5.5 more games than their Vegas totals, just 4 of 7 (57 percent) have indeed gone over, slightly worse than the overall average. The potential for lottery teams to fall short of their RPM projections by focusing on draft positioning down the stretch is a risk.
Having considered the track record, let's look at which teams fall into these categories this season.
Total better than RPM projection
Los Angeles Lakers
Westgate total: 48.5 wins
RPM projection: 41.2 wins
Because of their large, vocal fan base located conveniently near Las Vegas, there does appear to be a Lakers tax in the totals. On average since 2006, they've fallen 1.5 wins short of their total. This year's difference is much larger, owing to the excitement surrounding the arrival of LeBron James.
The past two times LeBron has changed teams, that excitement has proven irrational exuberance. The Miami Heat fell 6.5 wins shy of its total in 2010-11, and then the Cleveland Cavaliers were 5.5 wins shy in 2014-15. That seems to reflect what my research has found -- that RPM projections need to be regressed to the mean when players change teams because of an adjustment period and uncertainty about how they'll fit in a different role.
Philadelphia 76ers
Westgate total: 54.5 wins
RPM projection: 47.8 wins
Last season's Sixers had the second-largest difference between total and RPM projection over the past three seasons and ended up hitting the over by nearly 10 wins anyway. That partially stemmed from Ben Simmons' beating expectations as a rookie to play at an all-star level, as well as dramatic improvement from Dario Saric in his second season. Both players have seen their projections improve accordingly.
Nonetheless, the Sixers' forecast still falls short of their total for a couple of reasons. First, the addition of Wilson Chandler looks like it could hurt Philadelphia based on his minus-1.8 projected RPM. Second, the estimate of 50 games played for Joel Embiid remains conservative.
San Antonio Spurs
Westgate total: 43.5 wins
RPM projection: 38.8 wins
Is it ever safe to bet against the Spurs? Last season, when Kawhi Leonard was limited to nine games by injury, was just the third time in the past 13 seasons that San Antonio has fallen short of its total. The Spurs have also tended to beat their RPM projections, doing so six times in the past nine seasons by an average of 4.6 wins in a testament to coaching. That's similar to the gap between San Antonio's total and its RPM projection this time around.
Boston Celtics
Westgate total: 57.5 wins
RPM projection: 53.2 wins
A year ago, the nearly 13-win gap between the Celtics' total and their RPM projection was the largest in the time we've been publishing them. Though Boston went under, that was only after Gordon Hayward was lost for the season in the opening minutes of the first game of the season and Kyrie Irving missed 22 games due to injury. With their returns, expectations are even higher for the Celtics. RPM remains a tad lower on their outlook, largely because of a forecast regression defensively for last year's best defense on a per-possession basis.
RPM projection better than total
Minnesota Timberwolves
Westgate total: 44.5 wins
RPM projection: 49.6 wins
Minnesota's total was the most surprising to me. The Timberwolves won 47 games last season despite star Jimmy Butler sitting out 23 contests due to injury, and they bring back largely the same roster this season. So I think a similar outcome is a reasonable baseline. RPM likes Minnesota to do even better than that, in part because of the departure of guard Jamal Crawford, whose minus-5.4 rating ranked among the league's bottom 10 players.
Utah Jazz
Westgate total: 48.5 wins
RPM projection: 53.4 wins
Utah's total is in line with last season's 48 wins, but there are a few reasons to believe the Jazz could beat that number. The first is potentially better health for center Rudy Gobert, who missed 26 games last season, during which Utah went 11-15. The other is the way the Jazz upgraded their roster at the trade deadline, turning Rodney Hood and Joe Johnson into longtime RPM standout Jae Crowder. Utah went 22-6 after the deadline. Keeping up that 64-win pace is not realistic, but winning 50-plus games certainly is.
Brooklyn Nets
Westgate total: 32.5 wins
RPM projection: 36.8 wins
The difference between the Nets' total and projection might reflect one key philosophical difference between the two measures. RPM projections are designed to best project the mean wins for teams, while totals are trying to balance the odds of the team going over or under (at least what gamblers expect that to be), which would be the team's median wins.
As noted earlier, this distinction is especially important for teams that could prioritize their draft pick over winning games. That might well apply to Brooklyn, which owns its first-round pick outright for the first time since 2013. If the Nets indeed stay in the playoff hunt as RPM forecasts, they've got an excellent chance of going over. If they fall out of the mix and try to sell off impending free agents DeMarre Carroll and Ed Davis at the trade deadline, Brooklyn could fall well short of its RPM projection.
Best bets
Based on this analysis, I think the best values on the board from Westgate's opening lines are the unders on the Lakers and Sixers and the overs for the Jazz and Timberwolves. While using RPM projections is no guarantee of success, doing so wisely has appeared to tilt the odds in the bettor's favor over the past three seasons.