Question about Minute to Minute Effectiveness

[JPScott]

Hi,

I was interested in a topic and thought if anyone knew whether it's been looked at before, the people on this board would know.

The topic is how effective players are as a function of their time played. Note I'm not talking about boxscore correlations which look at how well a player did vs. the minutes a player played in a game.

I'm talking about how effective is a player on a minute to minute basis, and in particular at what point is he at his most effective and at what point does this effectiveness start to decrease. In line with these questions is what is the effect of rest on the player. (i.e. how much rest would a player need to come back into the game and be back at his top effectiveness). Note that I'm particular interested in this topic with respect to college players.

The reason I ask is that I'm a University of Kentucky fan and throughout the past season a number of UK fans have had spirited debates discussing the effectiveness of the platoon system etc. vs. more traditional substitution patterns. FWIW I've been a huge proponent of the platoons for that particular team and thought that in the first part of the season before Alex Poythress was injured (and UK was largely adhering to a 5-for-5 substitution pattern) that UK was clearly playing at a historically dominant level. However after his injury, UK started to go more and more away from the platoons and UK's dominance dropped accordingly. (and I think the stats clearly back all this up.)

One positive aspect (of many) of the platoons is that it allows players to go all-out for a relatively short amount of time and then they are assured rest. And it's very simple to assess how the unit is playing as a whole, including on a minute-to-minute basis, because generally it's reflected on the scoreboard, but one can also see it if they know what to look for.

For example, my observation was that at least in college, when the players are playing at a high level (and in particular playing high level defense) that after about 5 minutes they start to visibly show signs of weakness (i.e. not getting out on shooters at the perimeter, starting to get beat off the dribble, not pushing the pace etc.)

One of my arguments for the platoons is that I think overall the effectiveness of the team is better when you shuttle in units under the theory that the effectiveness of players on a second squad playing in say their minutes 1 through 5 is overall higher than the effectiveness of player on a primary squad playing their 6th through 10th minutes. (of course the big question is what is the talent differential between the primary squad and the secondary squad, which in UK's case I don't think there was a huge difference.)

Another aspect of this is the effect of 'rest'. I.e. how much more effective is a player if he plays all-out 5 minutes and then gets a guaranteed rest, and then plays another 5 minutes vs. the same player who plays 10 straight minutes and doesn't know if he's going to get rest or not, and might be expected to play an additional 5 minutes. Again, from what I saw of UK this year under the full platoons the teams never seemed to get tired, whereas later in the year when certain players started to receive extended minutes, there were clear signs of conservation of energy, slowing of pace etc. if not outright fatigue.

In the end this is an optimization problem. I.e. how can you best use your assets to provide the highest overall effectiveness during the course of the game, and in turn put the most pressure on your opponent as well.

One extreme would be to say the five best players should play regardless, so under that scenario one would start the best five players and continue to play them for the entire game (barring they foul out). But that generally never happens at any level, because clearly players' effectiveness will eventually decrease to the point where it makes more sense to bring in a sub and give the starter a rest. (if for no other reason than to allow the starter to regenerate his energy.)

Anyway, I was wondering if anyone has looked into these things before, or perhaps there's been some sort of scientific study or paper expressly looking into a player's effectiveness as a function of the minutes they play, on a minute-to-minute level.

[Crow]

I believe there has been a look at offensive efficiency of lineups minute to minute for stints on court (but I could be thinking of lineup efficiency under different cumulative usage levels). I took a brief look here but didn't find the reference. Perhaps someone else remembers. I am not aware of a player level study of this. I have been interested in the topic though. Perhaps someone here with a play by play database might take it on even for a few sample players.

[NateTG]

I want to look at how the players' peak and average acceleration changes with floor time and with load from the SportsVU data, but that would require access to the raw SportsVU data.