Arguing with Lin Fan about regression to the means

[Massdrive1976]

I've practically given up on explaining what regression in basketball means to him. He constantly insists the term means deterioration, and that Lin didn't deteriorate because his 3pt percentage improved. What he didn't understand was that Lin's 3pt volume was very low last year, so any small number of increments would move the percentage up. The difference between the year before last season, and this season's was minute if you look at the hard number.

His 3P M and A were 1.1-3.1 in 2013 and 1.2-3.2 in 2014. He went from .339 to .358. Using percentage alone to grade his improvement was dishonest. His shots from the two years listed were basically identical in averages. They don't have to be exactly alike when you're talking about averages.

The main why I use regression to describe his play last year though is that he had a rubber band effect. He went from starting out really hot, something about 50%< on his 3 pt in the first 8 games. After that he dropped to a low of almost 20%. In the end, it all equaled out to his average from the year before in Houston.

Yet he insists that Lin improved, I've backtracked on my meaning of regression and that regression means deterioration. He doesn't understand that regression to means isn't regression downwards but towards the average whether that is up or down.

If anyone's interested, they can take a look at the argument below under the topic Jilliebeans (me) I wish Jeremy would just improve his game instead of making funny videos all the time. I want to take him a little more seriously at this point.

The argument starts at Michael Jenkins, when he asked, "If he regressed, why did he have a career-year in 3-point percentage?"

https://www.youtube.com/watch?v=SNBIVUeMrvU

I'm Lin fan, but I believe objectivity is most important when discussing basketball.

[RealRapsFan]

here is a question for you.

What is the 'mean' for a player? The average to date regardless of 'age/experience'?

We know, or atleast expect, players will improve with time (up to a point ofcourse where we expect age to have have an inverse effect). So given that Lin is young, has only 4 years experience, and have seen both his volume and % improve with that age/experience - why would we expect that this year to be some sort of outlier and that will fall back to some past average? Isn't it more reasonable to expect that he is either near, or trending towards, what will be his career average - which is higher than the average of previous years?

Lin's 3pt shots were lower last year as a whole (28 less 3pt shots), but higher per game (by 0.1) (and per 36 (0.5) for what its worth), with his % increasing respectively. I also don't notice any notable trend in his 3 pt shooting per month over time - rather he had unpredictably better and worse months (which one should expect).

Did Lin 'regress to the mean', whether up or down? - sure, technically every player constantly is. But I don't see how an argument can be made that his 3 pt shooting didn't somehow improve. Will it fall next year or at some point in the future or over time? Maybe, but the trend clearly shows a trajectory of improvement to date.