Re: [APBR_analysis] Re: Jump Shots
- On Sun, 24 Feb 2002, HoopStudies wrote:
> No. Tversky looked at both foul shots and field goals. I've got 3But when looking at field goals, surely we'd have to EXPECT to see
> papers that he and Gilovich wrote. I suppose I could scan them in
differences in shooting percentages. Last year the Knicks held their
opponents to 41.7% shooting whereas the Warriors opponents shot 47.2%.
Not only would we expect players to have better shooting nights against
the Warriors, those stats show that they DID have better shooting nights.
This doesn't address the issue of hot streaks inside a game (maybe teams
shoot 37.2% against the Warriors in some quarters and 57.2% against them
in other quarters), but those inside-the-game hot streaks (assuming
they exist) are only one type of hot streak. Indeed, I think rather than
identifying hot streaks that turn on and off during a game, most fans
think of hot streaks along the lines of "well Reggie was certainly on
fire last night wasn't he" i.e. big games.
> Is this true? I did a quick ttest of Van Exel's performances againstWell the relatively small sample sizes involving head-to-head matchups are
> Seattle vs the rest of his career and there was no significant
> differences. Van Exel has had 8 single digit scoring nights against
> Seattle, 5 games of 25 pts or more. I've definitely never noticed an
> effect or even heard of it until it was raised here a little while
going to make it hard to reject a number of null hypotheses. How do
the means compare? Van Exel's averaged about 15 points per game overall.
Against the Sonics, he's fallen short of that average by 5 points or more
eight times; he's exceeded that average by 10 points or more five times.
Even just on those numbers, it sounds like he tends to score more against
the Sonics than he does on average.
> > "hot streak".To me a better test of hot streaks or perhaps more accurately hot games is
> I'd just be curious to see how well you all predict it.
seeing how players' shooting percentages vary from game to game. Actually
not just their percentage, one would have to take into account the number
of times they shot too. Needless to say, some games a player will shoot
80% and some games he'll shot 20%. The question is, does this variation
exceed that which we would expect to see from, say, the binomial model?
My expectation is yes, simply on the basis that sometimes the opponent is
the Warriors and sometimes the Knicks.
Now as for free throws, I don't have any particular expectation that a hot
hand exists. But field goal streaks could be created by crummy defense,
and stopped by good defense. And teams and players do differ on the
quality of the defense they play.
I guess I'll have to look at those Tversky articles. Lack of hot hand for
free throws I can believe. Lack of hot hand for field goals within a
game I could also believe, although if we had detailed information on who
the defenders at each point in time were, we might find something. But a
hot hand for field goals between games is something that I would expect to