Thursday, May 28, 2009

A Better Metric for Strikeouts

Even with sports analytics impacting baseball more every season, opportunities still exist to upgrade some mainstream statistics.

High strikeout totals grab media and fan attention, despite research showing that strikeouts are only slightly less damaging to offensive production than other outs. Nonetheless, since they get perceived as a negative, we should at least evaluate them in the proper context.

Ryan Howard (54), David Wright (50), Adam Dunn (50) and Prince Fielder (48) rank third through sixth in NL strikeouts (through May 27). But their strikeout rates – the percentage of plate appearances that end in a strikeout – rank much better. Howard drops from third in total strikeouts to fifth in strikeout rate. Wright falls from fourth to ninth, and Dunn goes from a tie for fourth with Wright to 10th. Fielder falls all the way from sixth to 13th. All these players get lots of plate appearances, which makes them look worse at making contact than is actually the case.

Strikeout rate also makes a great stat for showing which players shine at making contact. Miguel Tejada leads the NL with a 5.6 percent strikeout rate. Carlos Lee (7.9) and Albert Pujols (7.9) place fifth and sixth. By the way, this measure is superior to the more commonly used at-bats per strikeouts, which penalizes hitters who draw more walks than other players.

Strikeout rate also works as a better tool for evaluating pitchers than the more commonly used strikeouts per nine innings. For example, Zack Greinke (9.7) and Joba Chamberlain (9.1) have comparable strikeouts per nine innings figures. But Greinke (28.6) owns a far superior strikeout rate than Chamberlain (22.8). Why? Chamberlain has faced more batters per inning because he has allowed hits and walks more often than Greinke. This gives him more chances to log strikeouts each inning. But why should Greinke’s pitching effectiveness hurt him in this statistical category? If strikeout rate gets used, this isn’t a problem.

Friday, May 22, 2009

The LeBron Stoppers

In the coming years, Eastern Conference foes will be searching for ways to slow down the Cavs and LeBron James. Defenders effective against them will have tremendous value. Which five-man player combinations excelled at doing this in 2008-09?

The units in the chart had the best defensive efficiency against the Cavaliers with James on the court this season. Defensive efficiency is the number of points allowed divided by defensive possessions times 100.

This list includes the five-man combinations that posted a defensive efficiency below 95 while on the court versus Cleveland and James. To put that figure in perspective, the league had an average defensive efficiency near 108. The average against the Cavs with James playing was 115.1. Among the 71 units that faced them for at least 10 minutes this season, only these 16 posted a defensive efficiency under 95.

The red ink shows the player likely to have guarded James from each unit. In some cases, teams may have zoned or rotated defenders on James.

Units (Team) Def. Eff.
1 Brown-Hamilton-McDyess-Prince-Stuckey (Detroit) 79.4
2 Artest-Battier-Brooks-Yao-Scola (Houston) 80.5
3 Ford-Granger-Hibbert-Jack-Murphy (Indiana) 82.2
4 Deng-Hinrich-Nocioni-Rose-Tyrus Thomas (Chicago) 82.4
5 Brand-Iguodala-Miller-Williams-Young (Philadelphia) 82.6
6 Blake-Fernandez-Outlaw-Przybilla-Roy (Portland) 84.0
7 Iverson-AJohnson-Prince-Stuckey-Wallace (Detroit) 84.8
8 Hawes-Jackson-Martin-Nocioni-Thompson (Sacramento) 86.2
9 Howard-Lee-Lewis-Nelson-Turkoglu (Orlando) 87.1
10 Anthony-Chalmers-Haslem-Marion-Wade (Miami) 88.0
11 Brown-Hamilton-Iverson-Prince-Wallace (Detroit) 88.9
12 Azubuike-Biedrins-Ellis-Jackson-Morrow (Golden State) 89.3
13 Alston-Howard-Lee-Lewis-Turkoglu (Orlando) 90.5
14 Bibby-Horford-Johnson-Smith-Williams (Atlanta) 91.4
15 Deng-Gooden-Rose-Sefolosha-Tyrus Thomas (Chicago) 92.6
16 Dalembert-Evans-Green-Iguodala-Miller (Philadelphia) 93.3

Detroit had three of the top 11 defensive units versus James and the Cavs. The only player in common to all three groups was Tayshaun Prince, who guarded James himself. The Magic had the ninth and 13th place units, which could loom large in the Eastern Conference Finals. The only different between their two groups was at point guard with Rafer Alston taking over for the injured Jameer Nelson to join their other four starters. The Sixers and Bulls also had two units apiece on the list.

While these combinations excelled defensively against Cleveland with James on the court, they may not have limited his scoring. But they slowed down the other four players enough to succeed.

Thursday, May 14, 2009

Opportunity and the NBA

When the Hawks managed a stop in their series versus the Hawks, there was Anderson Varejao battling to keep the ball alive any way possible. He didn’t always grab the board, but it deflated Atlanta when he did. All that work on defense, only to have to start with a fresh shot clock.

Watching players like Varejao and Chris Andersen impact their teams makes you wonder how many players like this are out there looking for opportunities: guys that rebound, defend, hustle and manage to score some without getting plays called for them.

One thing our research has revealed is that teams will always find top scorers. That’s not always true with rebounders. According to the numbers, getting an opportunity is much tougher for these players.

This season, 128 of the 329 NBA players that saw 500 minutes of action averaged 20 points per 48 minutes. Meanwhile, 105 of those 329 averaged 10 rebounds per 48 minutes. What’s interesting is the scorers averaged 2188 minutes played, 567 more minutes than the rebounders (1621). While just 15 of these scorers saw less than 20 minutes per game, 42 of the rebounders failed to get that much court time.

This effect gets even more extreme if we make the groups more selective. The league had 45 players that saw 500 or more minutes average 25 points per 48 minutes. A near equal number of players – 49 of them – had 13 boards per 48 minutes. This group of scorers averaged 2378 minutes played compared to 1628 for the rebounders. In other words, the extra rebounds did nothing to generate more playing time. They averaged only 7 more minutes per contest than the previous 10-rebound group. But the new scorers group averaged 190 more minutes.

Here’s the amazing thing: none of the 25 points per 48 minutes scorers failed to see less than 20 minutes per game. Lou Williams had the lowest figure at 23.7 minutes per game. The rebounders group had 15 players that saw less than 20 minutes per game.

So the message is clear: If you score – even as a bench player – some team will get you playing time. Rebounders have a tougher time. While quality scorers certainly impact the game more than quality board men, the gap isn’t large enough to explain this type of discrepancy.

A look back at the 2003-04 season produced comparable results. This time, the list of underutilized rebounders (13 rebounds per 48 minutes in less than 20 minutes per game) included two interesting names. The first, David West, later became an All-Star. Chris Andersen was the other. Not all the other players emerged – some have since left the game – but there are more Andersens and Varejaos out there. Is there a team that will give them a chance?