Monday, February 12, 2007

This past Saturday, the Texas Tech men's basketball team lost a wild, controversial (regarding some foul calls and non-calls) double-overtime game at Oklahoma State, 93-91. As the resident streakologist, the part of the game that most captured my attention was when Texas Tech opened the second half by making its first seven shots from the floor, including three three-pointers.

The eternal question in hot hand research is, naturally, whether the streak of seven straight made shots is just random fluctuation -- like having several heads (or tails) in a row come up in a long series of coin flips -- or does it represent a deeper tendency of the team to be streaky? One way to answer this question is to compare the observed frequency of such streaks to what would be expected by chance, which is best done over longer stretches than just one game.

More immediately, I have been trying to come up with ways of depicting hot and cold shooting sequences graphically, and the figure below is my latest effort. It shows all of Texas Tech's field-goal attempts, whether each shot was made or missed, roughly how many minutes were left in the period at the point of each shot, and roughly the distance of each shot attempt (grouped into lay-up/dunk, longer two-point attempt, and three-point attempt). The chart does not show other outcomes of possessions, such as turnovers or trips to the free-throw line.

One obvious candidate for another piece of information to add is the identity of the player taking each shot; whether player numbers or initials could be shrunk small enough to fit in the diagram, yet remain big enough to be seen, is a matter for another day. For now, interested readers will have to look at the play-by-play sheet, on which my chart is based, to see who took each shot. Please feel free to share any suggestions for the chart's format in the Comments section below.

At a team level, at least, hot and cold streaks are readily apparent. Other questions, such as whether a team on a hot streak will attempt more ambitious (i.e., long-distance) shots, can also be examined.

As is shown in the chart above, for most of the game's five-minute blocks, Texas Tech's shots consisted of a mixture of made and missed attempts. However, the early minutes of the game featured a stretch of predominantly misses, whereas the second half started off with the aforementioned hot streak. Now, are these two deviant blocks just due to random chance, or part of a larger propensity for streakiness?

1 comment:

JiM Wright said...

Alan picked up on your blog from the Tech basketball forum. I have developed a pretty good formula for evaluting basketball players -- if maybe I could be any use to you. I've even thought that there may be a market for a website to grade basketball players on a daily basis. Here's NBA & BigXII ratings:

JiM Wright