Iron unkind: The Celtics often struggle to execute down the stretch of close games.
When the Boston Celtics started destroying the league in the last few months of the regular season, their dominance hid what had been one of their biggest problem areas until that point: Clutch offense.
On January 22, the Celtics were 23-24, with a 9-17 record in games within five points in the last five minutes … and a 14-7 record otherwise. That’s when they turned their season around, going 28-7 over their final 35 games to climb from 10th place to 2nd in the Eastern Conference.
The Celtics had both the league’s No. 1 offense and its No. 1 defense over those 11 weeks, outscoring their opponents by an amazing 15.5 points per 100 possessions.
But by beating their opponents so decisively (20 of the 28 wins were by double-digits, 15 were by 20 points or more), the Celtics didn’t really resolve their clutch issues. They were just 4-5 in games that were within five in the last five over those last 11 weeks, scoring just 79 points on 90 clutch offensive possessions (87.8 per 100). Only the New York Knicks had a worse clutch offense over that stretch.
So the Celtics finished the season with a 38-9 record (.809, best in the league) in non-clutch games and a 13-22 record (.371, second-worst) in clutch games. That was the third-biggest differential in that regard in the 26 years for which we have clutch data.
The Celtics (plus-7.4) were 1.9 points per 100 possessions better than the Golden State Warriors (plus-5.5) in the regular season, but the Western Conference champs have home-court advantage in this series because they won two more games. Appropriately, the Warriors won the close regular-season meeting between the two teams, while the Celtics won the blowout.
Things picked up in the first round of the playoffs. The Celtics swept the Brooklyn Nets, but all four games were within five points in the last five minutes. And over those four games, Boston scored 25 points on 21 clutch possessions (119 per 100), with Jayson Tatum’s Game 1 buzzer-beater being the highlight.
Brooklyn, of course, had one of the worst defenses in the playoffs. Since that series, the Celtics’ clutch offense in the postseason has fallen below a point per possession (89.1 per 100 through Sunday). Late-game offensive struggles at the ends of Games 6 and 7 of the Eastern Conference finals almost had them blowing a 3-2 series lead and heading into the offseason with a ton of regret.
Such regret remains a possibility, especially after Boston blew another fourth quarter lead in Game 4 of the NBA Finals on Friday. After taking a five-point lead with 7 1/2 minutes to go, the Celtics scored just twice (six points) on their final 12 possessions, and they went scoreless on all six possessions that qualified as “clutch” (because they came in the last five minutes with the score within five).
It’s not like the Celtics have an identity crisis. They know who they are and how they want to attack offensively, spacing the floor and putting weak defenders in actions against Tatum and Jaylen Brown. But for whatever reason, they’ve often struggled to execute down the stretch of close games.
Here are how things went wrong for the Celtics (and right for the Warriors) late in Game 4.
1. Defending the flare screen
Flare screens (for multiple players) have become a big part of the Celtics’ offense in this series. Early in the fourth quarter, Tatum got an open 3 when Andrew Wiggins got caught under a flare screen and Jordan Poole was late to switch out.
With a chance to build on that five-point lead, midway through the fourth, Robert Williams III set a flare screen for Tatum that flattened Andrew Wiggins. But Tatum didn’t get an open shot, because Kevon Looney was there, switching off of Williams. Tatum drove baseline, Poole provided timely help, and Klay Thompson rotated down to Derrick White, forcing Tatum to back out and reset:
Marcus Smart set a screen to get Stephen Curry defending Tatum, but Curry was able to stay in front of Tatum’s initial drive and force a tough step-back and a shot-clock violation:
It’s possible that Tatum could have attacked quicker after the catch or got off the ball quicker when Poole came to help, but the skip pass to Brown in the opposite corner would have been a tough one to make over Looney.
2. The Thompson stops
Thompson struggled with his shot (10-for-33) through the first two games of the Finals. But he’s been better over the last two, and he had 13 points in the second half on Sunday, while making several big plays on the other end of the floor.
Thompson cut that five-point deficit down to three with a tough turnaround fadeaway in the paint. And he followed it up with a terrific defensive play.
With Wiggins anticipating a Tatum curl off a Smart cross-screen, Tatum cut back door, leaving his defender in the dust. But Thompson saw what was happening, switched off of Brown in the corner, and contested Tatum at the rim without fouling:
Two possessions later (clutch possession No. 1 for the Celtics), Brown rejected a screen, looking to beat Thompson off the dribble. But Thompson stayed with him and forced a tough runner that rimmed out:
Later in the fourth (clutch possession No. 5), Thompson completely shut down a Brown isolation, keeping him in front and not giving him enough space for a step-back jumper. Brown kicked the ball out to Horford, but his 3-pointer was a little rushed with Looney closing out:
With the Celtics down six (not a clutch possession), Horford made a corner 3 the following trip down the floor. And after a Looney layup (when the Celtics got the ball out of Curry’s hands), the Celtics had their sixth and final clutch possession of Game 4.
Again, Thompson was most responsible for getting the stop. Brown cut back door and Smart was able to squeeze a pass through, but Thompson stayed with him and (with some help from Wiggins) forced a turnover:
There were more vulnerable defenders — Curry and (sometimes) Poole — on the floor when Thompson got those stops, and it seems the Celtics chose the wrong guy to attack.
3. Help, recover and contest
When Curry hit a runner from the elbow to put the Warriors up three with a little less than four minutes to go, the Celtics called timeout and drew up a play (clutch possession No. 3) to get Curry in an action against Tatum.
But as he took a handoff from Derrick White, Tatum paused for a split second. Wiggins got around White, and Draymond Green pinched off of Brown to prevent a Tatum drive. Wiggins then helped on a Brown drive and quickly recovered to contest Tatum’s 3-pointer:
The following possession (clutch possession No. 4), it was Green doing the same thing. Thompson (one possession prior to the iso above) stayed in front of a Brown isolation, but Green helped in the paint. Brown kicked out to Smart and Green made a huge effort to contest the shot as best he could:
Brown got the offensive rebound, but Wiggins stayed in front of a Tatum spin move and Smart missed another 3 against another Green contest.
4. Sometimes slow and stagnant
The Celtics’ clutch issues on Friday weren’t as self-inflicted as their issues in Game 6 of the conference finals. As noted above, there were good defensive plays/actions from all six Warriors (even Poole!) who played in clutch time.
But on their first clutch possession (Brown rejecting the screen and missing the runner), note that there were only 11 seconds left on the shot clock before the Celtics ran anything designed to gain an advantage, because they were so slow to get into their offense:
Curry was incredible in Game 4, but the difference in these games has been almost entirely about the Boston offense, which has scored 125.5 points per 100 possessions in its two wins and just 95.4 in its two losses. Turnovers and rebounds have been a part of that, but shooting is the most important aspect of this game.
Interestingly, the Celtics have shot only slightly more effectively in the first 12 seconds of the shot clock in their wins (effective field goal percentage of 66.9%) than they have in their losses (63.7%). The much bigger difference in their shooting has been in the last 12 seconds of the shot clock: 54.1% in wins, 35.7% in losses.
Late-clock execution matters, and it’s also important to avoid those late-clock situations. The Celtics also had multiple possessions earlier in Game 4 where they just weren’t very purposeful and settled for bad shots.
5. More clutch time coming?
Through four games, the Warriors have outscored the Celtics by a single point. It’s now a three-game series and chances are that at least one more game will be close down the stretch.
Amazingly, the Celtics are two wins from a championship having yet to really solve their late-game offensive issues. Of course, the Milwaukee Bucks had similar issues last season and then went 3-0 in games that were within five in the last five, scoring 25 points on 17 clutch possessions (147 per 100), in the Finals. And that was against the team — Phoenix — that’s otherwise 65-23 (.739) in clutch games over the last two years.
Even more than that series, it feels like anything can happen over the next seven days. Game 5 is Monday (9 p.m. ET, ABC) in San Francisco.
John Schuhmann is a senior stats analyst for NBA.com. You can e-mail him here, find his archive here and follow him on Twitter.
The views on this page do not necessarily reflect the views of the NBA, its clubs or Turner Broadcasting.