Skip to main content
ABC News
Brad Stevens Makes The Celtics Better Than They Have Any Right To Be

Even with home-court advantage, there was little reason to think the Boston Celtics should be favored in their conference semifinal series against the Philadelphia 76ers. As Boston was struggling to grind out a seven-game first-round victory over the Milwaukee Bucks, Philly was resting up, having dispatched the Miami Heat with relative ease. And while the Celtics were down their best scorer (Kyrie Irving), the Sixers had center Joel Embiid back from injury to join all-around terror Ben Simmons as the series’ two headline players. On paper, the Celtics seemed to stand little chance — and the Las Vegas bookmakers agreed.

In the end, though, Boston kept defying the odds like it’s done all season, closing out a surprising five-game series win over the 76ers Wednesday night. Al Horford outplayed Embiid, Terry Rozier continued his unheralded rise to playoff stardom and — perhaps most importantly — head coach Brad Stevens continued his yearlong project of spurring this team to play better than ever seemed possible. That’s been part of Stevens’s reputation since his days at Butler University, and the numbers bear it out, too: Stevens really does stand out in the ways we can measure coaching effectiveness.

First, there are the X’s and O’s of Stevens’s playcalling, which got extra acclaim in the Philly series as the undermanned Celtics relied as much on tactics as talent to engineer the upset. Kevin O’Connor of The Ringer did a great job of detailing them here: How Stevens repositioned his bigs to reduce Embiid’s impact; how he deterred Simmons’s forays to the hoop in transition; how his inbounds plays changed the course of the series. That last skill is one Stevens has been making good use of all season, with Boston ranking fourth in the league in net efficiency on sideline and baseline out-of-bounds plays (according to Second Spectrum):

The best teams on out-of-bounds plays (at both ends)

Net efficiency on sideline or baseline inbound plays, 2017-18 NBA season

Offense Defense Net
Team Rk Pts/100 poss. Rk Pts/100 poss. Rk Pts/100 poss.
Kings 13 109.3 1 70.0 1 +39.3
Jazz 7 123.3 3 87.3 2 +36.0
Rockets 4 132.4 10 97.9 3 +34.6
Celtics 8 122.0 5 88.6 4 +33.4
Nuggets 10 117.0 8 94.7 5 +22.3
Suns 16 105.1 2 84.2 6 +20.9
Pistons 12 110.5 6 89.8 7 +20.7
Raptors 5 131.3 19 111.5 8 +19.7
Spurs 6 123.5 16 106.8 9 +16.7
Bucks 9 121.6 15 105.7 10 +15.9

Data up-to-date through games played May 8.

Source: Second Spectrum

Beyond drawing on the whiteboard, Stevens seems to already be a master of the art of squeezing wins out of a roster via intangible factors such as fit, chemistry and player development. If we use each player’s projected Box Plus/Minus (Basketball-Reference’s estimate for points added per 100 possessions) as a guide,1 and if we plug in how many minutes each player actually logged this season — a big consideration for a team riddled with injuries — we’d have expected Boston to only win about 38 games this season.

The Celtics won 55.

These Celtics produced a lot more value than expected

Actual vs. projected wins added (based on Box Plus/Minus) on offense and defense, for members of the 2017-18 Boston Celtics in the regular season

Off. Wins Def. Wins
Player Age Pos. Min. Proj. Actual Proj. Actual
Jayson Tatum 19 SF 2438 +0.8 +2.0 +2.3 +4.9
Al Horford 31 C 2277 +3.2 +4.1 +5.3 +6.5
Jaylen Brown 21 SG 2152 -0.2 +1.6 +1.5 +2.9
Terry Rozier 23 PG 2068 -0.3 +3.3 +2.3 +3.6
Kyrie Irving 25 PG 1931 +7.3 +10.2 +0.0 +1.3
Marcus Smart 23 SG 1614 +1.7 +0.2 +2.9 +2.9
Aron Baynes 31 C 1485 -0.6 -1.7 +2.2 +3.5
Marcus Morris 28 PF 1447 +1.3 +0.7 +1.5 +1.1
Semi Ojeleye 23 PF 1150 +0.4 -1.2 +1.1 +1.4
Daniel Theis 25 C 936 +0.3 +0.6 +0.9 +3.0
Shane Larkin 25 PG 775 +0.5 -0.2 +0.5 +1.0
Abdel Nader 24 SF 522 +0.2 -1.3 +0.5 +0.7
Greg Monroe 27 C 496 +0.7 +0.6 +0.9 +1.8
Guerschon Yabusele 22 PF 235 +0.1 +0.1 +0.2 +0.3
Jabari Bird 23 SG 115 +0.0 +0.0 +0.1 +0.1
Kadeem Allen 25 PG 107 +0.0 -0.3 +0.1 +0.1
Jonathan Gibson 30 PG 40 +0.0 +0.2 +0.0 -0.1
Xavier Silas 30 SG 7 +0.0 +0.0 +0.0 +0.0
Jarell Eddie 26 SF 6 +0.0 +0.0 +0.0 +0.0
Gordon Hayward 27 SF 5 +0.0 +0.0 +0.0 +0.0
Total +15.3 +18.8 +22.3 +35.1

Projections use a weighted average of the previous three seasons, regressed to the mean and adjusted for aging patterns.


That’s only the sixth time since the dawn of the 3-point era (1980) that a team won at least 55 games while carrying a projected record under .500.2 Granted, much of it was due to the instant contributions of rookie Jayson Tatum, dramatic improvements from Rozier and NBA sophomore Jaylen Brown, and an MVP-type season by Irving — but that’s also the point. The best coaches, such as San Antonio’s Gregg Popovich, have a repeatable tendency to put up more wins than statistical projections say their roster should. With an average of 4.9 extra wins per season over his first five years on an NBA sideline, Stevens appears to be one of those coaches. And while Stevens’s value was clear to even the most casual NBA observers when Boston kept winning despite the injuries to Gordon Hayward and then Irving, it wasn’t so clear to the other 29 coaches in the league: Stevens received zero Coach of the Year votes from his peers.

He’s taking it up an extra notch this postseason. The Celtics had a schedule-adjusted efficiency differential of +3.4 points per 100 possessions during the regular season (5.7 points/100 better than projected), but they’re up to +4.9 in the playoffs (8.3 points/100 better than projected). Sometimes, teams who overachieve in the regular season struggle to keep beating expectations in the postseason, where raw skill rules and the margin for error is much smaller. But Stevens and the Celtics have only gotten better at overachieving as the playoffs have gone on.

Now Boston will have to face its toughest test yet, in the form of the team that eliminated them from last year’s postseason — the Cleveland Cavaliers. The Cavs did their own overachieving (maybe …) by unexpectedly sweeping the East’s No. 1 seed Toronto Raptors, but that was only after they lost more regular-season games than teams like the Celtics despite superior talent. A month into the playoffs, that talent is beginning to round into form next to LeBron James, who could be playing the best basketball of his entire career.

If that sounds like a problem that even superior game-planning and locker-room management can’t solve, well, it may prove to be. But if anyone is up to the task, it’s probably Brad Stevens.

Check out our latest NBA predictions.


  1. Based on a variant of Basketball-Reference’s simple projection system, which regresses each player (particularly rookies and other low-minute players) toward a below-average, rather than average, mean on offense and defense. (This improves the predictive value of the projections.)

  2. The others were the 1998 San Antonio Spurs (in Tim Duncan’s rookie year), 1993 Houston Rockets, 1990 Spurs (David Robinson’s rookie year), 1989 Phoenix Suns and 1980 Boston Celtics (Larry Bird’s rookie year).

Neil Paine was the acting sports editor at FiveThirtyEight.