Thursday, March 27, 2008

The Value of FTs

Ken Pomeroy is usually an AP 'must-read', but sometimes his application of (Bill) Jamesian-numerical analysis of college basketball leads him to counter-intuitive conclusions that are also counter-factual.

Take his analysis of the value of FT% (in the context of Memphis' chances to win the national title):

For one thing, the ability to make free throws has very little correlation to winning close games. It needs to be noted that just three teams in the country have a better track record than Tennessee in close games this season. The Volunteers are 8-2 in one-possession games--games decided by three points of less at the end of regulation. Memphis hasn't been in many of those games this season, but has won all three they've come across. That's 11-2 for the two bricklayers combined in close game action.
Such analysis is of no solace to Bob Huggins and the West Virginia Mountaineers.

WVU lost tonight against Xavier, 79-75 in OT, a two-possession game that falls outside of KenPom's definition of a 'close' game (although in fairmess, Pomeroy expands the definition to four points later in his column.) Two trips to the foul line were key for WVU:

* Coming out of a timeout with 0:21 left in regulation, and down by 2 points, WVU ran a terrific disguised screen play to get the ball to their best player -- Joe Alexander -- on the low block. Alexander made the short jumper -- tying the score -- and was fouled. Although a terrific foul shooter (82.1%), Alexander missed the "and-1" and -- after a Xavier miss at the other end -- the game went to OT.

* Later in OT, after Alexander had fouled out, the Mountaineers were clinging to a 4-point lead. In the last 3 minutes of the game, WVU made three trips to the line with Wellington Smith (54.2%) going 0-for-2, and Joe Mazzulla (64.8%) going 2-4 in two trips. While the proximate cause of WVU's loss was the failure to cover Xavier's BJ Raymond on two-straight possessions (including a key one with 0:02 on the shot clock with 0:32 left), the missed free throws -- especially Smith's empty trip with 2:57 left -- were critical.

Pomeroy has a point -- coaches may put relatively poor foul shooters on the floor because their talents in other areas allow their FT% to be overcome elsewhere. But looking at "one-possession" final scores is not a fair test of free throw acumen.

First, relatively few games are truly 'last possession' games. Even games with a close score (i.e., within 3 points) can be decided earlier with a meaningless hoop at the buzzer allowing the score to be not-as-close as it appears to be.

Second, the real test for 'close' games should games that are tied or within one possession in the last five minutes in regulation, or (by definition) any overtime game. Alexander, Smith, and Mazzulla all missed critical free throws. Alexander's miss cost his team the opportunity to win in overtime (although the tie score may have had an influence on the type of shot that Xavier was willing to take in the last seconds). Smith's miss killed WVU's momentum (built in spite of Alexander going to the bench early in the OT). And Mazzulla's misses deflated a team that was clinging to a two-point lead (first trip, with 2:13 left) and all-but-killed the team with 0:17 left (74-78).

Finally, overall FT% is not the same as FT% in the last few minutes of closely-fought, tense game where both teams are playing at a high level -- and the refs are allowing the players to 'decide the game' by only calling indisputable fouls (which includes almost every NCAA Tournament game.)

Is FT% overrated? Perhaps. Can talented teams overcome poor foul shooting? Sure.

But can missed free throws kill a team? Absolutely. And what is harder to measure is the momentum-killing implications of an empty trip to the line at the end of a close game.

Earl Weaver famously said that momentum is "tomorrow's starting pitcher." While that may be true in baseball, a game of discrete (and discretely measurable) interactions, the game of basketball is more complex, and involves more group psychology.

Just ask the Georgetown Hoyas.

1 comment:

Matt Eastwick said...

I would tend to agree with KenPom's FT thought process, but at this point have to question his log5 analysis. UNC with only a 6% chance of winning the tourney before the Sweet 16, which was driven by only a 56.5% chance to beat Wazzu?

Basketball statistical measures are crude when compared to other sports, which leads one to the conclusion that Ken's raw stats are dog food quality meat in the hands of a master butcher.

And what's with his aversion to all #1 seeds advancing to the Final Four? At least we may see some existing basketball next weekend rather that the blowouts of the past four days (excluding Xav/WVa and Dav/Kans). This has the potential to be one of the all-time great Final Fours.