APBRmetrics

schtevie · Joined: 18 Apr 2005 Posts: 97

The pre-season stats are starting to accumulate. Are these meaningful for predicting subsequent seasonal performance - in terms of teams, in terms of individuals?

John Hollinger · Joined: 14 Feb 2005 Posts: 88

Can't remember where I saw th is, it might have been here, but somebody did some kind of study that found a really good presason record -- like 7-0 or 6-1 -- is a mildly positive indicator, and nothing else means jack. That's on a team level; haven't seen anything for individuals but my basic premise is that all the veterans are coasting so it's tough to evaluate.

THWilson · Joined: 19 Jul 2005 Posts: 143 Location: phoenix

It was over at 82games.

I've never seen anything about players.

devin3807 · Joined: 11 Oct 2007 Posts: 5

I actually did a piece on this myself.

I was wondering if I could get some input from some of you.

http://www.devinblack.com/preseason.html

Ben · Joined: 13 Jan 2005 Posts: 211 Location: Iowa City

I think a regression with both preseason and previous season records as independent variables would be a natural thing to look at.

Mountain · Joined: 13 Mar 2007 Posts: 293

Devin's article adds some useful things such as the correlations for good, average and bad teams. The team by team correlations of pre-season and regular season are also interesting though coaching stability or change should temper its use as an indicator now.

Hawks will be an interesting case to watch. I think 42 wins is reachable though I'd guess the under is more likely.

7-1 Toronto highlighted in the fall '06 82 games study paid off big in increased regular season wins.