I seem to remember running across an odd sort of playoff prediction system a few years back that was based on cumulative regular season team stats. I had thought this was an invention of Bill James, though I could be wrong. At the time it didn’t seem all that useful to me for predictive purposes, but as I recall it had an uncannily high rate of success.

It was something along the lines of….

2 points to the team with the most doubles
-1 point to the team with the most triples
3 points to the team with the better ERA

A couple of google searches came up empty for me. Does anyone know what I’m talking about, or have I mis-remembered something, or just made this whole thing up?

Why would a team get -1 for most triples?

Cutter - 17 September 2006 04:49 AM

Why would a team get -1 for most triples?

It’s probably a typo.

That formula still doesn’t make much sense. And I haven’t heard of anything like it.

Bill James did come up with a playoff predictor in one of the Abstracts- I think maybe 1983.  Your memory of the general structure of the predictor is right.  If I recall correctly the predictor did pretty well from 1969-1982, but someone dusted it off and looked at it for recent years and it hadn’t done that great.