February 5, 2009

Wait 'Til Next Year

Let's Get Efficient!

by Bryan Smith

Printer-friendly

Contact Author

I don't think it's a stretch to say that college baseball is an offense-oriented game: the ping of the ball off of the bat testifies to that. There have been great strides made in reducing the number of games ending in football scores-only five times did a team score in double digits in Omaha-but as I watched the 19-10 slug-fest in Game Two of the College World Series championship, it was a frustrating reminder of how easily aluminum can shrink a ballpark.

Fresno State's national championship, however-and Oregon State's before that-says something about recent college baseball victors. It certainly helps you get to Omaha if you can swing the bat like Miami or Florida State did in 2008, but big offenses have not been what has defined the champions in recent history. Anyone who watched Tommy Mendonca and Justin Wilson last June knows the kind of effect that good pitching and defense can have in a college contest.

It's that latter notion, the defense, that I concerned myself with this offseason. Statistical transparency has thankfully gotten better in college baseball in the past few years, and, with some digging, there is now access to a far wider range of metrics than ever before. It's easy to identify the top pitching staffs; hell, just find a copy of North Carolina's 2008 numbers. To identify the top fielding teams, though... that's not so easy, at least not if you know enough to discredit the idea that fielding percentage is the end-all indicator of a team's defensive abilities. The NCAA lays that out in a nice, accessible table:

Rk   School           FPct
 1.  Duke             .978
 2.  South Carolina   .977
 3.  Creighton        .976
 4.  Western Kentucky .976
 5.  Troy             .976
 6.  Texas Christian  .976
 7.  Pepperdine       .976
 8.  Maryland         .975
 9.  Stanford         .975
10.  Hawaii           .975

That's all well and good, I suppose, but there's a lot of noise in those numbers-one coach told me last year that at certain ballparks there are scorekeepers who refuse to call errors on their home squad-and, to be honest, I can't put much stock in a statistic where eight-thousandths of a percent separate the first-ranked team (Duke) from the 49^th (Auburn). Fans of Bill James know that there is a better way; one of his simplest, yet most substantial contributions to sabermetrics is the concept of Defensive Efficiency. It looks complicated written out as a formula, but it's actually as simple as this: what percentage of balls in play are converted into outs? It's the true test of a defense, and this summer I discovered that the numbers necessary to compute DER for college baseball were available for 283 of the 286 Division I teams.

The results in this survey were overwhelming: defense matters a great deal at the college level, and DER is a good tool for measuring defense. Every team in the College World Series in 2008 finished above the Division I average of .617. Seven of the eight CWS teams, including both finalists Fresno State and Georgia, finished in the top 50. Trust me, there are not a lot of statistics out there that can lay that kind of claim, and while only Stanford of all the CWS teams finished in the top ten in fielding percentage, Georgia was close at 12^th, and most of the other teams weren't far behind.

How's this for a test? Make an internal list of the top five conferences in college baseball last season. Now take a look at how close they come to the top of the list:

Conference    TBF      K%    BB%     DER
Pac-10      21,116   .188   .097   .6407
Big 12      23,642   .186   .088   .6363
SEC         29,660   .195   .090   .6351
ACC         28,760   .194   .094   .6343
C-USA       22,270   .185   .101   .6323
B. East     26,924   .172   .093   .6311
WCC         13,603   .179   .094   .6272
CAA         24,199   .165   .097   .6245
MVC         20,112   .166   .094   .6229
WAC         21,754   .150   .095   .6220
B. South    18,834   .161   .086   .6211
B. West     20,737   .171   .075   .6211
Patriot     11,946   .157   .088   .6207
MAC         26,659   .148   .096   .6189
Southland   27,748   .153   .096   .6149
A. 10       31,005   .154   .096   .6142
Big 10      21,706   .145   .101   .6140
A. East     14,155   .156   .094   .6139
Sun Belt    25,921   .171   .095   .6139
OVC         21,540   .160   .100   .6114
Horizon     15,685   .131   .088   .6100
MAAC        18,137   .142   .103   .6099
SoCon       24,209   .167   .102   .6098
A-Sun       28,104   .157   .092   .6098
MWC         16,645   .153   .088   .6075
NEC         16,160   .152   .096   .6059
Ivy         13,244   .154   .091   .6016
MEAC        15,213   .150   .133   .5741
SWAC        18,205   .139   .122   .5707

(Please note that my calculation of Defensive Efficiency is the same that BP uses in the MLB Defensive Report, though with one exception: I had to use raw errors rather than ROE. This likely escalated the numbers a bit, but not significantly, and any differences should be spread out evenly across the board.)

I surely would have had the Pac-10 at the top of my list, and that notion is validated here. In 2008, it was the deepest conference in college baseball, and every one of its teams had an argument to make for the post-season group of 64. League Defensive Efficiency leader Washington had a strong resumé, as did league trailer USC. What's most reaffirming about the legitimacy of using Defensive Efficiency for college baseball is that the indisputable top five conferences from last season end up as the top five in this regard as well.

Here is a list of the top 25 teams at converting balls in play into outs in college baseball last season, along with their accompanying 2008 win-loss record.

Rk   School               DER     W-L
 1.  Duke               .6725    37-18
 2.  Tulane             .6711    39-22
 3.  Washington         .6702    33-22
 4.  Creighton          .6639    37-21
 5.  W. Michigan        .6632    29-23
 6.  TCU                .6630    44-19
 7.  Canisius           .6620    41-13
 8.  South Carolina     .6616    40-23
 9.  NC State           .6585    42-22
10.  Coastal Carolina   .6575    50-14
11.  Pepperdine         .6570    38-21
12.  Georgia            .6565    45-25
13.  UCLA               .6557    33-27
14.  Texas              .6556    39-22
15.  Oklahoma State     .6555    44-18
16.  Georgia State      .6548    33-23
17.  Seton Hall         .6543    31-25
18.  Stanford           .6531    41-24
19.  St. John's         .6526    42-16
20.  Lafayette          .6520    25-23
21.  Oregon State       .6520    28-24
22.  Michigan           .6519    46-14
23.  Cal-Santa Barbara  .6509    35-21
24.  East Carolina      .6507    42-21
25.  Lamar              .6499    35-23

The top 25 defensive teams in '08 averaged 38 wins versus only 21 losses. Clearly, defense is something that can elevate a team at every level of college play, pushing schools from Canisius to TCU to Michigan ahead of the other members of their respective conferences. It's also a statistic that turns squads that would be decidedly mediocre based only on evaluating their hitting and pitching-Duke and Lafayette, I'm looking at you-and turns them into above-.500 teams.

To speak specifically about the teams on this list, I can personally attest to Georgia's defensive dominance. Gordon Beckham made huge strides in his junior season, and the team opted to forego offense at second base in favor of defense. Add in a third baseman at first base in Rich Poythress, and a trio of rangy outfielders, and their defense pushes them to the next level. South Carolina is another interesting team, as three of their infielders from last year were high draft picks. It's no news that Justin Smoak is a defensive ace, but South Carolina's high ranking also speaks to Reese Havens' and James Darnell's ability to turn grounders to the left side into outs. And remember that with Beckham and the three Gamecock infielders all having since moved on to the pros, there's a lot more than offense that these schools need to worry about replacing.

There are some other notables further down the list:

 Rk   School              DER     W-L
 27.  Louisiana State   .6484    49-19
 29.  UNC-Wilmington    .6476    44-17
 31.  Florida State     .6471    54-14
 32.  North Carolina    .6460    54-14
 33.  Miami             .6455    53-11
 48.  Fresno State      .6405    47-31
 50.  Cal St Fullerton  .6400    41-22
131.  Rice              .6207    47-15
147.  Kent State        .6180    36-21
160.  Southern Cal      .6147    28-28
165.  Missouri          .6137    39-21
199.  New Orleans       .6055    43-21
213.  Mississippi St.   .6016    23-33
216.  Coll. Charleston  .6007    39-20

The rest of the College World Series crew quickly flies off the board, with five more placing between 27^th and 33^rd, and Fresno State just making the top 50, leaving only one laggard (Rice). I just wonder how much better Rice might have done had they not been the worst team in Omaha (and seventh in the C-USA) at catching the baseball. This is a team that struck out 23 percent of batters faced in 2008. Who knows what moving their Defensive Efficiency from 62 to 65 percent might have done for that pitching staff, not to mention Cole St. Clair's draft stock.

This was perhaps the most frustrating aspect of this research, being faced with cases of what could have been. Missouri could have been a CWS contender with a better defense, but they ranked last in the Big 12 at Defensive Efficiency. I also included USC and Mississippi State on this list, because they were at the bottom of their conferences in Defensive Efficiency, finishing with a combined 51-61 record. And whether it's College of Charleston, New Orleans, or Kent State, there are a host of teams that were a good defense away from that last step up.

The most frustrating team in the nation may have been the New Orleans Privateers. What a run they had in the Sun Belt Conference last year, hitting .321/.409/.528 as a team with six players clubbing at least 10 home runs. Ultimately, the season ended at the hands of Southern Miss in a series that the Privateers should have made more competitive, and on the surface, it looks as if they were dragged down by a pitching staff with a 5.45 ERA, and by players like Justin Garcia, one of the weekend starters (5.77 ERA).

It was Garcia, however, that was the one being victimized. He has a good arm; good enough to earn a spot in professional baseball, and even though the Royals sent him to the hitter-friendly Pioneer League in his first summer in pro ball, he responded with a 3.00 ERA, emerging as a closer with 42 innings of fantastic work. That, mind you, was with an Idaho Falls team that turned balls into outs at a 61.6 percent clip (according to Sean Forman), a percentage point higher than Garcia's college team.

These are just some of the less obvious factors that are worth considering before we dig into the 2009 college landscape. As I move towards previewing the 2009 teams, I need to think about how Georgia will replace Gordon Beckham and Ryan Peisel on their left side, or whether Kent State will defend well enough to take the next step forward. As the tournament nears, I'll need to think about which defenses might step up in the way that Fresno State's did last June. And as the draft nears, we should begin to use this statistic to analyze which pitchers are being undervalued by their surroundings, like Justin Garcia was a year ago.

The 2009 baseball season is just two weeks away, and I'm not talking about pitchers and catchers. We'll preview the NCAA in each of those weeks, never losing sight that there's a new analytical tool at our disposal.

Bryan Smith is an author of Baseball Prospectus. 
Click here to see Bryan's other articles. You can contact Bryan by clicking here