Should Your Star Come off the Bench?

Don’t worry, we’ll get to that.

In the recent podcast with Josh Weil, we briefly talked about a fun subject we’ve hit lately. Ari Caroline brought up the idea of “floor stretch”, which dealt with the idea of game theory and which set of players you should have on the floor. One of the stories told in the Wages of Wins is that they myth that stars like Jordan make their team mates better is a bit misleading. There’s a very simple issue in basketball. There is only one ball. That means if Michael Jordan, Scottie Pippen and Horace Grant are all great shooters, well only one of them can have it at any given moment.

Some of Arturo’s work has shown that when a good player goes from a bad team to a good team they get a little worse. There are exceptions like the 2008 Boston Celtics, which combined a great set of complimentary talents to make one of the greatest teams ever. And that actually brings me to today’s subject. What if the very idea of how lineups are constructed is wrong?

A team tends to put out their best players and then sub out very inferior players off the bench. A question I had is why is playing a player like Dwyane Wade off the bench a bad idea? If you ensure that both LeBron and Dwyane Wade play 36-40 minutes a game then why does it matter if one of them is on  the bench when the initial tipoff occurs?

I did a fun test. The Thunder lost soundly to the Heat in last year’s finals. Except, it was more like the recent presidential race. The Thunder didn’t get blown out repeatedly. Rather, they had a collection of close losses sink them.

I grabbed the data from Popcorn Machine for how the game went. I then took the players stats from the regular season from the NBA Geek. Then I looked at what the production for each lineup the Thunder put out. Finally I graphed this to show the strength of each lineup the Thunder put out (based on their regular season numbers)

Strength of OKC Thunder Lineups as Game 3 went on.

As we can see the Thunder started with their strongest players. Then their reserves came in. To start the second half, the Thunder put out another strong collection. Then to end the third they put in some very weak reserves. They put back in some stronger players towards the end of the game. Let’s compare this to the actual game flow (chart from Popcorn Machine).

When the line goes up the Thunder were winning. When the line goes down, the Heat were winning. Now this chart doesn’t line up identically with the strength of the lineups the Thunder put out. I’d be surprised if it did. What we can see if that when the Thunder put out their strong lineup in period three they grabbed their biggest lead. Then they replaced these players with their weakest lineup of the game (Derek FisherDaequan Cook, James Harden, Thabo Sefolosha, Kendrick Perkins) and in a very short period of time the game was turned around. It seems like the Thunder were trying to go “small ball” but really they were putting in a weak collection of players.

So, my end question is should teams be aiming to play their best together? Rather than have more peaks and valleys maybe they should instead be more consistent across the game? To be fair, the Thunder had problems by giving players like Fisher and Perkins major minutes. It may be that playing Harden off the bench wasn’t actually a bad idea. Maybe, they should have tried to make sure that every lineup they put out was strong. By putting out players like Fisher, Cook and Perkins together, the Thunder left themselves vulnerable. I’ll have to look at more teams. A single game does not prove a theory. But the idea of if teams should space out there stars just doesn’t seem that crazy to me.


Comments are closed.