Bwog’s sports editor and everyone’s favorite, Ross Chapman is here again, reviewing this weekend’s games against Princeton and Penn.
If you’ve seen the Columbia Lions Men’s Basketball team play this season, you would have gotten the impression that the team is very, very bad at defending against the three-point shot. In this weekend’s games against the “Killer P’s” of Princeton and Penn, that crucial weakness of the team was on full display. Of the points the Lions allowed in a 70-62 loss at Penn and a 61-59 almost-comeback against Princeton, 69 of those points (52.7%) came from beyond the arc. This is a lot. I’ve written a few times now about how Columbia’s offense ticks, but what makes them fare so poorly in three-point defense?
Two statistics describes how effectively teams use the three-point shot. The most common statistic to consider is three-point field goal percentage (3P%), which divides three-pointers made by three-pointers attempted. On any given shot, how likely is it that the team will make a basket? The average team shoots 35.2% from beyond the arc, with Ivies varying from 37.8% (Yale) to 32.4% (Cornell). But there’s a simple reason why that statistic doesn’t totally describe the situation – you miss 100% of the shots you don’t take. Looking at a team’s three-point field goal attempt percentage (3PA%), also called shot distribution or three-point rate, tells you how often any given shot a team takes will be from three-point range. The national median is 36%, but the Ivies are very three-point happy. Cases in point: Penn and Princeton combine to take about 45% of their shots from long range.
When Columbia went to Princeton on Saturday, they knew that they were going up against a profoundly threatening three-point team. Ranking 75th nationally in 3P% and 13th in 3PA%, the Tigers employ the trey as their primary weapon. Steven Cook led an efficient first half assault against the Lions, converting on 7 of 10 three-pointers in the frame. The Lions weren’t cold tonight, even when they were down 41-26 at halftime. They made their shots at an average clip, and they didn’t fall very behind the Tigers in the turnover margin, a part of the game where Princeton usually excels. However, Princeton slowed the game down – the Lions’ 53 shots were the fewest they’ve attempted all Ivy season. When the Tigers had the ball, they were just hot as hell in the first half (65.2 FG%), with Myles Stephens leading the charge towards the basket. Princeton regressed towards their season mean in terms of their shooting percentages during the second half, but one telling stat stayed true throughout the game – the three-point rate.
Princeton took 26 of their 50 shots from beyond the arc. According to statistician Ken Pomeroy, a team’s three-point rate during any given game is much more easily predictable than its three-point percentage. “The number of three-pointers each team will take is about as predictable as the pace each team will play at. Pace and shot distribution are part of a game’s DNA. Conversely, the r [correlation coefficient] for three-point accuracy is 0.15, which is the lowest of anything I’ll be looking at [in an essay series].” While three-point success percentage depends significantly on skill and daily luck, looking at the three-point rate shows us how teams try to play against Columbia.
In general, teams are more likely to keep their tried-and-true offensive plans than to make a drastic change against a specific opponent. If you look nationally at three-point field goal attempt percentage, the standard deviation of three-point rate is about 35% larger than the standard deviation for three-point rate defense. In other words, a team that shoots “a lot” of threes will exhibit more difference from the “average team” than will a team which allows “a lot” of threes to be shot. Three-point rate is, therefore, mostly determined by the will of the offense.
But Columbia is a special case. Teams that go up against Columbia are making huge adjustments in their three-point rate. Out of 351 NCAA D-1 teams, the Lions allow the 14th most three-point attempts. And as compared to the national 36% three-point rate median, Ivy teams have shot an astonishing 51.5% of their shots from beyond the arc. (If the Ivy-Season-Lions were a team, they would be allowing a higher three-point rate than any other team in NCAA D-1.) This was most exaggerated when the Crimson came to New York, and they made 62.9% of their attempts from beyond the arc, compared to their usual 41.2% rate against other teams. Princeton and Penn made similar adjustments, shooting at a higher-than-their-average three-point rate in their wins against the Lions.
We’ve been mired in numbers – what does this all look like in the lived experience of a game? Teams aren’t just taking more three-pointers than normal because the Lions are amorphously bad at defending against them. The Lions rank very close to the average in three-point field goal percentage defense. But as Pomeroy remarks, defenses have relatively little control over the success rate of an open three-point attempt. What Columbia does wrong is that they allow too many three-pointers. By playing a zone defense, the Lions don’t get as close to shooters on the perimeter. Additionally, the Lions have a tendency to double-team opponents on the baseline to force a steal. When this fails, the opposing team often rotates the ball around to a guard who is open for a three-point shot. This is the crux of Columbia’s failure against the three-pointer. The Lions have done nothing to stop other teams from finding those shots, and could very easily have another loss on their record if not for the inexplicably low three-point success percentage Harvard put up two weekends ago (Harvard took 39 three-point attempts and made only 10).
In eight Ivy games, only once has an opponent shot at a lower three-point rate than their average when they squared up against the Lions. Teams keep trying to exploit the Lions on this weakness, and the Ivy League is a terrible place to be weak to the three-point jumper. Penn and Princeton, three-happy teams in general, turned it up to 11 against Columbia, and I don’t just say that because Princeton made 11 threes against the Lions. Columbia gives up open threes, and in every loss, the three-point rate is one of the first stats I look at. (Interestingly, Columbia is also worst in the Ivies at allowing rebounds after the opponents shoot a three, presumably because they’re woefully out of position after allowing an open three. Penn’s offensive rebounds nearly single-handedly won the Quakers their first-half advantage on Friday.) If they don’t find some way to patch up this problem, the Lions will be at a severe disadvantage over the next three weeks.
Goin’ up on a Twosday via Columbia University Athletics/Mike McLaughlin