Drawing purpose from purposeless games

How to evaluate the success of a game when winning or losing doesn't matter

What is a meaningless game? For the Toronto Raptors, you would be forgiven for thinking that they’re all meaningless. Toronto is almost certainly locked into the 2-seed in the Eastern Conference, with a 97 percent certainty per Inpredictable. The Raptors have been locked into their hierarchy in the conference for so long that it feels like every game since the all-star break has been meaningless. As a matter of fact, the last time they weren’t second in the conference was on January 30, and the Milwaukee Bucks haven’t looked back from first place since. The regular season ended long ago for the Toronto Raptors, and like Llewelyn Davis, Toronto can appear to be running in place alone, caught in the doldrums of time.

Of course, sports are never useless, at least not in the sense of why we watch. We watch because we care, and that doesn’t change when the Raptors are stuck in a rut in March. There is purpose in caring.

If you look even smaller, there is purpose in each game that runs far deeper than fandom. The Toronto Raptors find meaning in each game through the space that comes after; each game is a dress rehearsal, and a missed line isn’t a death knell in the present, but perhaps a foreboding toll that can portend problems in future matchups.

So what have we learned, both in the positive and negative?

Toronto’s starters seem set in stone. Specifically, Marc Gasol seems to be the permanent starter. If the more mobile Serge Ibaka didn’t start against the speed demon Kemba Walker, then Ibaka probably won’t start again. Even in losses to the Oklahoma City Thunder and Charlotte Hornets, the starters have performed relatively well. They finished a +15 in 24 minutes against the Thunder and +2 in only 13 minutes against the Hornets. Gasol himself, like most of the Raptors’ starters, boasted a positive plus-minus in both games, while most of the bench players combined for huge negative margins.

So in a sense, both losses have reinforced lessons we already know; Toronto’s bench cannot sustain games on its own. This has been an issue all season long, as Toronto has won games with its starters and lost them with its bench players.

 

Graph from pbp stats, with x axis as number of starters, y axis as percentage of possessions, and colour as net rating.

Put another way, here’s a simple line graph that shows how dramatically Toronto’s net rating dips as starters leave the game. It’s pretty flipping straight.

Toronto knows all of this. It leads the league in minutes played with all five starters on the court. So in a sense, confirming knowledge with a high level of certainty has the same value has a game in March itself: meaningless. But if you dig deeper, there can be meaning gleaned from within individual data points.

Toronto will have to play some bench players at some point in the playoffs. That is a fact. They will not do well in the playoffs if their starters play 48 minutes each a game. But which bench players should play when? For so long, Toronto was treading water with a variety of injuries. They can only learn so much when their ability to combine certain players is limited. But as Sunday night against the Hornets was the first time since the trade deadline that Toronto has had a full complement of players, they finally got a chance to experiment with bench players in different lineup variations.

Toronto learned a variety of lessons with a full roster. For one, OG Anunoby can only play alongside starters. If you take just the Hornets game, Anunoby boasted positive net ratings when paired alongside two starters, neutral net ratings alongside one starters and one bench player, and negative net ratings when alongside two bench players. This is confirmed if you look at Anunoby’s lineup splits over the whole season, with his two worst ratings (alongside players still on the team) coming alongside Jeremy Lin and Norm Powell, and Jeremy Lin and Patrick McCaw.

That’s a clean split that speaks to a lesson we already knew: keep Anunoby alongside more talented players where his role is to shoot open shots and expend all his energy on the defensive end. He should not be asked to create.

 

Here’s an example of Anunoby’s success in a very similar possessions, but it’s alongside more threatening talent, where rotations are later, and he knows what is expected of him.

Another lesson for another bench wing: Powell needs to be tied to players who create for him instead of the other way around. He was excellent against the Hornets, the only bench player other than Fred VanVleet with a positive plus-minus, and it was because Powell played exclusively in lineups featuring initiators VanVleet and Pascal Siakam. Powell never had to carry lineups, and so he finished a +8 against Charlotte. This is confirmed by Powell’s full-season lineup data, where his two best partners are Kyle Lowry and Siakam. His two worst partners are Jeremy Lin and Ibaka. (Powell is much better alongside a passing center like Gasol than finishing centers like Jonas Valanciunas or Ibaka.)

It was to learn these lessons that Toronto staggered its bench across the two halves, playing Lin exclusively in the first half and Powell exclusively in the second. McCaw also played far more in the first half. That’s not a rotation meant for success, but rather one meant to give bench players auditions with the starters. Returns thus far point towards Powell beating out Lin or McCaw for the scraps of playoff minutes a ninth man will receive.

Regular season experimentation is going to lose Toronto some games. But experimentation will continue as long as losses won’t affect seeding. For what else are these games when Toronto cares not for the result? It’s better to do it now than the playoffs. Toronto could easily win all its games remaining; cut the rotation, play the starters 35+ minutes each, and run the plays that they know will succeed. But Toronto doesn’t care if it wins or loses. What matters is knowing what will work in the future. And even though Toronto has already played 74 games, they still don’t know exactly how best to manage a full 48 minutes in a single one.

Games can matter even if their results do not. They can matter in the positive by providing data about lineup combinations. But they can also matter in the negative through bad habits. For one, Toronto has been allowing more 3s to their opponents as of late. That’s a habit that needs kicking, but it comes as a result of experimenting in new pick-and-roll coverages. Toronto’s increased rate of turnovers has similarly been problematic, but what else can you expect from players who know that the results of these game aren’t of the highest priority? Toronto is not limping into the playoffs, but they are playing like a team that knows the biggest test it yet to come. That is a good thing. We’ve seen how regular season consistency can bloom into playoff failure. Teams who always try hard have no extra gear when it matters.

Against Charlotte, no one got hurt. That’s the most important source of meaning for Toronto in these dreary games. Leonard has been praising Toronto’s medical staff while rocking dinos on his custom kicks. By any and all measures, the regular season has been a success, especially as Toronto is at its healthiest state entering the playoffs. Toronto is going to finish second in the regular season, and they’ve won a boatload of games. They may be losing more games recently than expected, but it’s all for a greater cause. Relax. Breath. Don’t panic. The best is yet to come.