serial correlation

[ClockMurphy]

Wonder if anyone has done extensive analysis of serial correlation in
their own wagering.

My master hypothesis which has proven true is that I have a much
higher expected value from bets where:

a) I have been recently hot at that particular track and/or
b) I have been successful on earlier races that day at the track in
question

....and, of course, the inverse is true as well. There are some days
when I am hopeless at a given track and I know it. Usually
surrounding wet or drying out tracks.

Are other folks adjusting their asset allocation in the short term
based on being successful/unsuccessful at a particular track?

[Dave]

You know, the baseball stats guys say that there is no correlation
with streaks - that they are pure randomness - but my experience of
sitting in many dugouts and being on the field during many college
games is that streaks are real. There are simply hitters that you know
are having a bad game/series and will not connect THIS time.

Might be interesting to study that in HSH's study system.

[spenc...@earthlink.net]

I do use the demographics to see where I am successful, or not. And
apply filter restrictions acordingly.

Together with my distance and surface restrictions I am very seldom
betting more than three of four races on any single card at a track.

I seldom play anything but fast dirt tracks.

[Dave]

I don't think that is what Clock is saying. I think he is talking
about hot & cold streaks.