Help - Pace-Adjusted Stats

[BuzzerBeaterBry]

Are there any flaws to comparing players like Wilt to players like Lebron using pace-adjusted stats? If so, could you explain?

[Chicago76]

Are there any flaws to comparing players like Wilt to players like Lebron using pace-adjusted stats? If so, could you explain?

Yes. The main problem is that it becomes very difficult to reliably control for other aspects when pace deviates substantially from whatever you are using as your baseline. If you're comparing a guy playing at 100 team poss /48 to one at 88 and normalizing them at 95, it's not really a big deal. When you do the same with a guy at 125 team poss/48, it becomes a much bigger deal. Player touches are distributed differently at really fast pace.

Simple example: assume every possession either ends in a made FG or a missed FG rebounded by the other team (no turnovers or FTAs). Also assume that it takes teams 5.5 seconds to get into their offense within 40 feet of the basket once they get the board or take the ball out after a made opponent FG.

At 90 possessions per game, that means that every team takes 10.5 seconds after they get the ball across and into their O to take a shot. At 130 possessions per game, that 10.5 seconds has shrunk to 4.9 seconds. That will impact a team's ability to get the ball to their most efficient scorer in the best spots on the court. It will hinder efficiency generally.

This also doesn't account for stylistic changes in the game (use of three and impact on spacing, relative distribution of the offense between the frontcourt and perimeter, etc.).

[BuzzerBeaterBry]

Are there any flaws to comparing players like Wilt to players like Lebron using pace-adjusted stats? If so, could you explain?

Yes. The main problem is that it becomes very difficult to reliably control for other aspects when pace deviates substantially from whatever you are using as your baseline. If you're comparing a guy playing at 100 team poss /48 to one at 88 and normalizing them at 95, it's not really a big deal. When you do the same with a guy at 125 team poss/48, it becomes a much bigger deal. Player touches are distributed differently at really fast pace.

Simple example: assume every possession either ends in a made FG or a missed FG rebounded by the other team (no turnovers or FTAs). Also assume that it takes teams 5.5 seconds to get into their offense within 40 feet of the basket once they get the board or take the ball out after a made opponent FG.

At 90 possessions per game, that means that every team takes 10.5 seconds after they get the ball across and into their O to take a shot. At 130 possessions per game, that 10.5 seconds has shrunk to 4.9 seconds. That will impact a team's ability to get the ball to their most efficient scorer in the best spots on the court. It will hinder efficiency generally.

This also doesn't account for stylistic changes in the game (use of three and impact on spacing, relative distribution of the offense between the frontcourt and perimeter, etc.).

I see. So how would you compare players like Lebron to guys like Kareem, Oscar, or West? and how much would you say these players benefited from the pace of the game?

[Blackmill]

Are there any flaws to comparing players like Wilt to players like Lebron using pace-adjusted stats? If so, could you explain?

There are flaws and I have mostly abandoned using pace-adjusted stats.

Let's start by simply stating what a higher pace offense implies. It means field goals are being attempted earlier in the shot clock, and in most scenarios, a higher proportion of possessions end in transition.

Immediately, one should realize this can actually be detrimemental to big men. After all, bigs take some time to set up position, and are unlikely to be be involved in transition, both offensively and defensively. Simply put, a lot of the extra possessions in a high paced offense might not even involve the center.

But if we think about pace adjustement more generally I would argue the same can be said of most any position when played at a superstar level. In making pace adjustements we are assuming that players wait for scoring, playmaking, and defensive opportunities to present themselves, rather than creating their own opportunities. While this may be true for role players my impression is that star players are largely responsible for their opportunities. In this case, pace adjustement makes very little sense.

Also, you may want to read this post.

[Chicago76]

Are there any flaws to comparing players like Wilt to players like Lebron using pace-adjusted stats? If so, could you explain?

Yes. The main problem is that it becomes very difficult to reliably control for other aspects when pace deviates substantially from whatever you are using as your baseline. If you're comparing a guy playing at 100 team poss /48 to one at 88 and normalizing them at 95, it's not really a big deal. When you do the same with a guy at 125 team poss/48, it becomes a much bigger deal. Player touches are distributed differently at really fast pace.

Simple example: assume every possession either ends in a made FG or a missed FG rebounded by the other team (no turnovers or FTAs). Also assume that it takes teams 5.5 seconds to get into their offense within 40 feet of the basket once they get the board or take the ball out after a made opponent FG.

At 90 possessions per game, that means that every team takes 10.5 seconds after they get the ball across and into their O to take a shot. At 130 possessions per game, that 10.5 seconds has shrunk to 4.9 seconds. That will impact a team's ability to get the ball to their most efficient scorer in the best spots on the court. It will hinder efficiency generally.

This also doesn't account for stylistic changes in the game (use of three and impact on spacing, relative distribution of the offense between the frontcourt and perimeter, etc.).

I see. So how would you compare players like Lebron to guys like Kareem, Oscar, or West? and how much would you say these players benefited from the pace of the game?

Way late in responding because I only occasionally lurk, but anyway, I think the thing to consider is how shots, passing responsibility, pts off rebounds, etc are distributed in a really fast paced environment. If you go back and look at the estimated percent of team FGAs stars got in the 60s vs today, it's not even close. This is more true for interior players, but even a West or Robertson type aren't getting as many scoring touches, nor are they getting the same "assisting touches" per possession. When you are playing really quick, the ball is inevitably spread out more. The beneficiaries are often 2nd and especially third option wing types like a Barnett or Twyman. There really isn't a good way to rescale this. I have my own methods, but what you are trying to do is to rescale distribution. The deviation in touches for average players isn't that different. At lower pace, average players would all give up a small portion of their touches. The ends of those distributions for high and low usage players is much more different though. Those little pieces you take from maybe 8 guys would end up in the hands of the stars.

Defensive specialist types would have far fewer touches...unless you're talking about an interior guy who is a great rebounder. At slower pace, those players are more likely to continue to get touched off put-backs.