Heading into the 2017-18 season, the Celtics likely expected their most used lineup to include Gordon Hayward, Al Horford, Kyrie Irving and some combination of Marcus Smart, Marcus Morris, Jaylen Brown or Jayson Tatum. Maybe Aron Baynes could have snuck in there as well. Five minutes into opening night, all of Brad Stevens’ well planned strategies were thrown out the window when Hayward was injured. And this was starting the season with Morris already out due to lingering knee soreness.
From that point forward, Stevens has been in test and adjust mode. The Celtics have used a whopping 188 different lineup combinations in their first 22 games. The only Celtics to appear in all 22 games? Tatum, Baynes and Terry Rozier. The most common starting lineup has been Baynes, Horford, Tatum, Brown and Irving. The second most common—but most used when Morris is healthy—is Horford, Morris, Tatum, Brown and Irving.
Boston’s most used lineups have appeared in just 9, 12 and 9 games and for just 85, 85 and 81 minutes respectively, per NBA.com. While those sample sizes are small, there are some telling trends starting to appear.
The Celtics most used lineup features Horford, Morris, Tatum, Brown and Irving. That group has played 85 minutes together over nine games, and all but two games were games where Morris started. This isn’t really a surprise, as few had Tatum pegged as a starter coming into the year. Tatum likely would have been replaced by Hayward, if the entire team was healthy. Surprisingly enough, Tatum is now the lone Celtic to have started all 22 of the team’s games.
An even bigger surprise? The Horford, Morris, Tatum, Brown and Irving group has been the Celtics worst lineup to date. That lineup has been a team worst -22 in terms of plus/minus. The Celtics have a total of 75 different combinations that are negative in plus/minus, but only eight have played as many as 10 minutes together. This means that outside of the grouping mentioned above, Stevens has avoided long stretches of poor play with the same players on the floor. Yet, for some reason, he uses this group a lot, despite it being a net negative.
Now, to be fair, the Celtics are 8-1 when Morris starts. But, in nine those games, Morris is -3 himself. In the nine games where that lineup has appeared, the team is 7-2, but that most used lineup is -22. In addition, in five of Morris’ nine starts, the Celtics have fallen behind by 10 more points. While Boston has established themselves as the “Comeback Kids”, that isn’t a sustainable strategy for winning basketball games.
The Celtics second most started lineup is Baynes, Horford, Tatum, Brown and Irving. That group is +17 in 81 minutes. The vast majority of the minutes this group played together was early in the year when Morris was recovering from knee soreness. Boston is 10-1 when Baynes starts, with the lone loss coming when the Miami Heat snapped the Celtics 16-game winning streak.
That unit is Boston’s best rebounding group by a fairly wide margin, including easily the team’s best offensive rebounding unit. They’ve shoot poorly, but that is compensated for by being one of the team’s best defensive groupings, out of a lot of really good defensive groupings.
Speaking of defense, it should come as no surprise that 10 of the Celtics top 12 groups in plus/minus feature Smart. This includes the team’s best unit of Horford, Tatum, Brown, Smart and Irving, which has becomes this year’s version of the “IT & D” closing lineup. “Kyrie and D” maybe?
Regardless of catchy nickname, that lineup is +31 in 85 minutes together. They also rebound well, pass well and somewhat surprisingly shoot it at an over 50% clip. The foul rate is high, but that is to be expected with a unit full of aggressive defenders, led by Smart.
None of this is to suggest that Smart should start, because he shouldn’t. While Smart is capable as the team’s “Sixth Starter” as Brad Stevens has often called him, he’s at his best when he can be used to change games off the bench. Bringing him in as a reserve also allows Stevens to control his fouls, which can become a problem when he starts games.
So, who should start then? The data and the eye-test supports the Baynes, Horford, Tatum, Brown and Irving lineup. That group gets the team off to good starts, especially defensively. It also creates a balance that the team is missing when Morris starts.
When Morris opens games, the Celtics most commonly go with Smart, Baynes, Terry Rozier and usually Semi Ojeleye as the reserves. On occasion, Daniel Theis and/or Shane Larkin mix in as well. Noticeably lacking in that group are offensive creators. Rozier can get his own shot, but no one else can. More often than not, when Boston goes to four or five reserves, which Stevens does often, the lead dwindles or disappears entirely. This puts the starters, or starters plus Smart, in a position to have to come back on a far too regular basis.
When Morris has come off the bench, albeit just three times so far this year, the entire squad is more balanced. In three games off the bench, Morris has helped stabilized the second unit offense, simply by being another shot creator. He’s able to make plays for both himself and for others. It takes the pressure off Rozier to create everything, or, God forbid, Smart having to make plays off the dribble.
When Morris starts, every possession he uses is one that isn’t used by Irving, Horford, Brown or Tatum, all who are more than capable of making plays. All too often, this leaves Tatum as the fifth option and he ends up just hanging out around the perimeter. This is part of what has led to Tatum scoring just 2.9 points in the first quarter of games. In the fourth quarter, when Tatum bumps up to the third or fourth option (pending how Jaylen Brown has it going that night), his scoring increases to 4.3 points. His true shooting percentage also goes from 56.3 percent to 78.7 percent, which is a sign of Tatum’s ability as a third or fourth option, and his ability to close games.
Back to Morris: he can also reasonably play 3-5 in the Boston system, because of his ability to defend on the perimeter and in the post. He’s also able to function both inside and outside on offense. Baynes doesn’t offer that ability. As more and more teams downsize on their second unit, Baynes often ends up a liability because he’s defending guys who are perimeter players versus post players. By starting Baynes, he can matchup with the opposing center, as most teams still start a somewhat traditional 5. And this allows Morris to fill the versatile backup big role. And when the Celtics run into a team with a bigger backup, they can always call upon Theis to handle that responsibility.
Marcus Morris was acquired in part because the Celtics didn’t want to simply give away a player to clear the cap space to sign Gordon Hayward. The trade was a win-win in that it created cap space, while also delivering the Celtics a useful player. Morris is the type of versatile player that Danny Ainge has given Brad Stevens to play with.
If Hayward were healthy, it may have been a no-brainer that Morris would come off the bench. But even without Hayward, it should be just as simple of a decision. It isn’t often that statistics and the eye-test match, but in this case they do. Marcus Morris should be coming off the bench for the Boston Celtics.