Saturday, April 01, 2006
They shoot, they score, we misinterpret
(11) George Mason vs. (3) Florida (CBS, 6:07pm ET)
(4) LSU vs. (2) UCLA (CBS, 8:47pm ET)
In what I didn't realize is an annual tradition until I paused and thought about it, I'm about to poke holes in the recurring Friday Wall Street Journal sports feature known as "By the Numbers," just as I did last year on the weekend of the Final Four.
I actually enjoy "By the Numbers" but when it's about college basketball it seems I find problems. (So maybe I should remember that when it's about pro football and I think it's really good. Anyway....) Last year, for example, the WSJ had elaborate data purporting to show that Louisville would beat Illinois in the Final Four because Rick Pitino had extensive experience coaching in the Final Four and Bruce Weber, then, had nada.
This year (no link--look in the dictionary under "paid site" and you'll find the WSJ) Allen St. John has come up with something he terms "shooting differential," which actually isn't a bad idea as far as it goes. (He even calls it "SD" so you know he's serious.) The only flaw--and it's a big one--is that St. John uses straight FG percentages (FG/FGA) and not effective FG percentages (FG + (0.5 * 3FG)/FGA). An eFG pct. recognizes scoring reality: threes are harder to make but they're worth 50 percent more than twos.
Using straight FG percentages and subtracting defensive FG pct. from each team's own FG pct., here's what St. John finds:
1. Florida (.105)
2. George Mason (.097)
3. LSU (.070)
4. UCLA (.061)
This would be like computing a "hitting differential" using batting averages instead of slugging percentages. If he'd used eFG percentages instead, St. John would have found this:
1. Florida (.119)
2. George Mason (.099)
3. UCLA (.078)
4. LSU (.048)
Markedly similar, no? Maybe. Let's take a look at these numbers....
The first thing to notice is that LSU and UCLA switched places. Why? Because LSU never ever shoots threes, devoting just 21.8 percent of their attempts to shots from beyond the arc. So their FG percentage looks better than UCLA's (Ben Howland's team devotes 34.5 percent of its shots to threes) but, in terms of actual scoring effectiveness, the Bruins get more out of each shot than do the Tigers.
This dynamic works on the other side of the ball, as well. For instance, UCLA's opponents shot fewer threes this year (26.9 percent of attempts) than did the opponents of any other Final Four team. The fewer threes your opponents shoot, the worse your straight FG percentage defense will look. So UCLA gets dinged in St. John's numbers, despite the reality that their FG defense is quite good. (In point of fact, there is much less difference among the four teams in FG defense than there is in shooting the ball. All four are very good at defending their opponents' shots.)
And then there's the larger point: there's more to winning and losing than shooting and defending shots. There's rebounding. There's holding on to the ball and forcing turnovers. And there are free throws.
Take Florida. The Gators are indeed one of the best shooting teams in the country. But they're hardly invincible--most notably they turn the ball over more often than does their opponent today, George Mason (thus vitiating their strong shooting). And they allow a fair number of offensive rebounds to their opponents (thus detracting from their strong FG defense).
Nor does shooting differential tell the whole (or even most of the) story with regard to today's other semifinal. While UCLA shoots much better than LSU (while playing basically the same level of FG defense), the Bruins are also the most turnover-prone team in Indy this weekend. And the Tigers make up for their (relatively) mediocre shooting by attacking the offensive glass--they are the strongest offensive rebounding team in the Final Four.
Suffice it to say: shooting well and defending shots are important, indeed crucial. Success starts there. But it doesn't end there.
In today's less Wonk-ish venues....
From the "great minds think alike" file. Ken Pomeroy lobs some pointed darts at the Coach of the Year voting this morning. As it happens, intrepid blogger Oof the Quick has done the same--and added some historical perspective, as well. Good reads.
Wonk back!
Don't just mutter ineffectually; email me! ¶ 6:51 AM
(11) George Mason vs. (3) Florida (CBS, 6:07pm ET)
(4) LSU vs. (2) UCLA (CBS, 8:47pm ET)
In what I didn't realize is an annual tradition until I paused and thought about it, I'm about to poke holes in the recurring Friday Wall Street Journal sports feature known as "By the Numbers," just as I did last year on the weekend of the Final Four.
I actually enjoy "By the Numbers" but when it's about college basketball it seems I find problems. (So maybe I should remember that when it's about pro football and I think it's really good. Anyway....) Last year, for example, the WSJ had elaborate data purporting to show that Louisville would beat Illinois in the Final Four because Rick Pitino had extensive experience coaching in the Final Four and Bruce Weber, then, had nada.
This year (no link--look in the dictionary under "paid site" and you'll find the WSJ) Allen St. John has come up with something he terms "shooting differential," which actually isn't a bad idea as far as it goes. (He even calls it "SD" so you know he's serious.) The only flaw--and it's a big one--is that St. John uses straight FG percentages (FG/FGA) and not effective FG percentages (FG + (0.5 * 3FG)/FGA). An eFG pct. recognizes scoring reality: threes are harder to make but they're worth 50 percent more than twos.
Using straight FG percentages and subtracting defensive FG pct. from each team's own FG pct., here's what St. John finds:
1. Florida (.105)
2. George Mason (.097)
3. LSU (.070)
4. UCLA (.061)
This would be like computing a "hitting differential" using batting averages instead of slugging percentages. If he'd used eFG percentages instead, St. John would have found this:
1. Florida (.119)
2. George Mason (.099)
3. UCLA (.078)
4. LSU (.048)
Markedly similar, no? Maybe. Let's take a look at these numbers....
The first thing to notice is that LSU and UCLA switched places. Why? Because LSU never ever shoots threes, devoting just 21.8 percent of their attempts to shots from beyond the arc. So their FG percentage looks better than UCLA's (Ben Howland's team devotes 34.5 percent of its shots to threes) but, in terms of actual scoring effectiveness, the Bruins get more out of each shot than do the Tigers.
This dynamic works on the other side of the ball, as well. For instance, UCLA's opponents shot fewer threes this year (26.9 percent of attempts) than did the opponents of any other Final Four team. The fewer threes your opponents shoot, the worse your straight FG percentage defense will look. So UCLA gets dinged in St. John's numbers, despite the reality that their FG defense is quite good. (In point of fact, there is much less difference among the four teams in FG defense than there is in shooting the ball. All four are very good at defending their opponents' shots.)
And then there's the larger point: there's more to winning and losing than shooting and defending shots. There's rebounding. There's holding on to the ball and forcing turnovers. And there are free throws.
Take Florida. The Gators are indeed one of the best shooting teams in the country. But they're hardly invincible--most notably they turn the ball over more often than does their opponent today, George Mason (thus vitiating their strong shooting). And they allow a fair number of offensive rebounds to their opponents (thus detracting from their strong FG defense).
Nor does shooting differential tell the whole (or even most of the) story with regard to today's other semifinal. While UCLA shoots much better than LSU (while playing basically the same level of FG defense), the Bruins are also the most turnover-prone team in Indy this weekend. And the Tigers make up for their (relatively) mediocre shooting by attacking the offensive glass--they are the strongest offensive rebounding team in the Final Four.
Suffice it to say: shooting well and defending shots are important, indeed crucial. Success starts there. But it doesn't end there.
In today's less Wonk-ish venues....
From the "great minds think alike" file. Ken Pomeroy lobs some pointed darts at the Coach of the Year voting this morning. As it happens, intrepid blogger Oof the Quick has done the same--and added some historical perspective, as well. Good reads.
Wonk back!
Don't just mutter ineffectually; email me! ¶ 6:51 AM
