| Boyd's World-> Breadcrumbs Back to Omaha-> Smart Stats for 2003, Part I: The Hitters | About the author, Boyd Nation |
Publication Date: August 12, 2003
AOPS
This is the first of an annual series of reports I do on performance recognition. This week I'm looking at adjusted OPS, which I introduced a couple of years ago. The stat still holds up fairly well, although I'm working on the next generation of hitting analysis stats, which I'll talk more about after the numbers.
First off, the leaders in raw OPS. Rickie Weeks, the repeat leader, is really suffering from Barry Bonds Syndrome; it's hard to understand just exactly how much an OPS over 1.500 changes the game. Since the relationship between OBP and runs scored is not linear, a .607 OBP (roughly twice average) from one player increases the team's runs scored by around 25% all by itself (that's a back-of-the-envelope guess based on some assumptions about the surrounding players, but it's in the ballpark). We'll see how Weeks does against tougher competition (see his AOPS below, although it's hard to tell how accurate AOPS is with his two extremes of high OPS and low SoS), but he's had a fantastic two years.
Team Player OBP SLG OPS 1 Southern Rickie Weeks 0.607 0.948 1.555 2 New Mexico State Billy Becher 0.482 0.900 1.382 3 Air Force Josh Phifer 0.502 0.804 1.306 4 Texas A&M-Corpus Christi Humberto Aguilar 0.488 0.804 1.292 5 Ball State Brad Snyder 0.522 0.770 1.292 6 Towson Mike Costello 0.500 0.785 1.285 7 Texas-Arlington Ryan Roberts 0.514 0.765 1.279 8 Santa Clara Scott Dierks 0.485 0.794 1.279 9 William and Mary Michael Brown 0.476 0.796 1.272 10 Southeast Missouri State Brian Hopkins 0.465 0.792 1.257 11 Tulane Michael Aubrey 0.505 0.733 1.238 12 North Carolina Jeremy Cleveland 0.510 0.725 1.235 13 Arizona State Jeff Larish 0.528 0.697 1.225 14 Nebraska Matt Hopper 0.505 0.717 1.222 15 Southern Mississippi Clint King 0.449 0.772 1.221 16 James Madison Eddie Kim 0.481 0.740 1.221 17 Toledo Mitch Maier 0.525 0.691 1.216 18 Northern Iowa Adam Boeve 0.471 0.745 1.216 19 Indiana Vasili Spanos 0.513 0.703 1.216 20 College of Charleston Lee Curtis 0.467 0.747 1.214 21 California Conor Jackson 0.538 0.675 1.213 22 Bowling Green State Kelly Hunt 0.500 0.713 1.213 23 Loyola Marymount Josh Whitesell 0.471 0.735 1.206 24 Arizona State Jeremy West 0.513 0.693 1.206 25 Southeastern Louisiana Anthon Garibaldi 0.479 0.715 1.194
Next, the AOPS leaders. Pay special attention to #1 and #3, because they're part of the reason I'm working on the next generation. Also pay special attention to #6, since I don't have a clue why he fell so low in the draft.
Team Player AOPS OPS 1 New Mexico State Billy Becher 1.408 1.382 2 Texas-Arlington Ryan Roberts 1.394 1.279 3 Air Force Josh Phifer 1.373 1.306 4 California Conor Jackson 1.367 1.213 5 Santa Clara Scott Dierks 1.356 1.279 6 Stanford Ryan Garko 1.345 1.172 7 Southern Rickie Weeks 1.340 1.555 8 Arizona State Jeff Larish 1.338 1.225 9 Nebraska Matt Hopper 1.336 1.222 10 Arizona State Jeremy West 1.317 1.206 11 Loyola Marymount Josh Whitesell 1.300 1.206 12 North Carolina Jeremy Cleveland 1.298 1.235 13 Tulane Michael Aubrey 1.297 1.238 14 Arizona Jeff Van Houten 1.295 1.168 15 Stanford Carlos Quentin 1.278 1.113 16 Southern Mississippi Clint King 1.277 1.221 17 Southeastern Louisiana Anthon Garibaldi 1.276 1.194 18 Northern Iowa Adam Boeve 1.255 1.216 19 Texas Dustin Majewski 1.252 1.105 20 Baylor David Murphy 1.250 1.101 21 Alabama Beau Hearod 1.249 1.153 22 Baylor Chris Durbin 1.249 1.100 23 Texas A&M-Corpus Christi Humberto Aguilar 1.247 1.292 24 Washington Bren Lillibridge 1.247 1.150 25 Southeast Missouri State Brian Hopkins 1.244 1.257
Improvements
The most common adjustment made to major league stats is not for strength of schedule (although some analysts are starting to recognize the necessity of some allowance for the unbalanced schedule), it's for park factors. I've only been able to come up with reasonable-looking overall park factors for the college ranks within the last year, and I'm still working on how to incorporate that into an adjustment factor.
The most commonly used method is just to adjust half of the player's stats by the park factor of his home park. That assumes two thing, though: that the player played half of his games at home (few college teams do) and that the player played the road half of the schedule at parks that averaged out to an average park (which is almost never the case in college, since most games are regional in nature, and park factors tend to be more homogeneous within a region). I think it's possible to take this into account, but it will be complicated, and I'm still working through the math. Look for something this winter.
If you're interested in reprinting this or any other Boyd's World material for your publication or Web site, please read the reprint policy and contact me
| Boyd's World-> Breadcrumbs Back to Omaha-> Smart Stats for 2003, Part I: The Hitters | About the author, Boyd Nation |