[Spiny_Norman]

A couple points on "close games".  First off, I think you are looking at the wrong metric.  What we really want is "close games heading into the final X minutes".  The final score is more or less irrelevant.

Consider the following two scenarios:

1) Stanford is trailing by 12 points heading into the final media time out (4 min).  Though the game is effectively over, Stanford hits a couple shots, gets a couple stops, but ultimately doesn't do quite enough to win.  We are losing 76 - 69 with after our opponent misses the front end of a one-and-one with 5 seconds left, and launch a meaningless desperation three that makes the final score 76 - 72.  Stanford loses another "close game".

2) Stanford is trailing by 2 at the exact same point.  The game is close, but our opponent makes a couple shots, we can't hit the broadside of a barn, and the lead balloons to 6 points with about a minute left.  We foul their guys, who convert, and miss some desperation threes.  Final score 82-70.  Fortunately, this doesn't count in the "close game" metric.

[You can do the same sort of thing with Stanford leading into the final stretch... a four point lead that Stanford stretches out to a comfortable win doesn't count, but a comfortable lead that the opponent whittles down to a couple points at the end counts as a "close win", even though I would prefer taking a close game and blowing it open by the final score than taking a game that should be at hand and stressing over the final minutes].

BC

BC, you are absolutely right.  However since I was not keen on reviewing the box scores of 150+ games to identify which ones were "close" with 2 minutes to play or 5 minutes to play or whatever criteria you want to set to measure late game results, I used "final score" as a surrogate data point.

[Spiny_Norman]

A couple points on "close games".  First off, I think you are looking at the wrong metric.  What we really want is "close games heading into the final X minutes".  The final score is more or less irrelevant.

Consider the following two scenarios:

1) Stanford is trailing by 12 points heading into the final media time out (4 min).  Though the game is effectively over, Stanford hits a couple shots, gets a couple stops, but ultimately doesn't do quite enough to win.  We are losing 76 - 69 with after our opponent misses the front end of a one-and-one with 5 seconds left, and launch a meaningless desperation three that makes the final score 76 - 72.  Stanford loses another "close game".

2) Stanford is trailing by 2 at the exact same point.  The game is close, but our opponent makes a couple shots, we can't hit the broadside of a barn, and the lead balloons to 6 points with about a minute left.  We foul their guys, who convert, and miss some desperation threes.  Final score 82-70.  Fortunately, this doesn't count in the "close game" metric.

[You can do the same sort of thing with Stanford leading into the final stretch... a four point lead that Stanford stretches out to a comfortable win doesn't count, but a comfortable lead that the opponent whittles down to a couple points at the end counts as a "close win", even though I would prefer taking a close game and blowing it open by the final score than taking a game that should be at hand and stressing over the final minutes].

BC

BC, you are absolutely right.  However since I was not keen on reviewing the box scores of 150+ games to identify which ones were "close" with 2 minutes to play or 5 minutes to play or whatever criteria you want to set to measure late game results, I used "final score" as a surrogate data point.