Several weeks ago, ESPN’s Buster Olney wrote an article about using outs to advance baserunners, claiming that it played a larger role in winning games than sabermetricians believed.
He presented precious little data — just enough to support his claim. The statistics he presented, Productive Outs and Productive Out Percentage, were not revealed in their entirety, and their significance could not be tested. I did some independent research on postseason data from 2002 and 2003 that showed Productive Outs to be fairly insignificant, but unless you wanted to delve into the game logs, you didn’t really have any data available to you to do further study.
Until now. ESPN recently began presenting Productive Out data for teams, individual batters, and for some reason, pitchers. Just for fun, I took another look at the data, but first I made some adjustments.
A Productive Out is defined as when the batter advances a runner with the first out, scores a runner with the second out, or when a pitcher sacrifice bunts with one out. Now, while this whole statistic is silly, the last category really creates problems. To say a sacrifice bunt with one out by Mike Hampton is a productive out but one by Neifi Perez is not … that’s just stupid. And it gives National League teams an advantage in this statistic, which is problematic when comparing the two leagues.
For this reason, I removed all the data for pitchers from the team totals, since I was unable to differentiate between the first two categories and the third from the information presented. With that data removed, here are the team totals through Monday’s games:
TEAM PO OPP POP Ari 75 211 .355 Det 83 235 .353 Pit 74 210 .352 KC 72 205 .351 Mon 54 158 .342 StL 76 223 .341 Tex 70 206 .340 Hou 66 199 .332 TB 72 221 .326 Phi 68 209 .325 Col 69 213 .324 Bal 74 229 .323 SD 80 248 .323 Cle 75 234 .321 Ana 71 230 .309 Mil 65 217 .300 Tor 77 258 .298 CWS 62 209 .297 NYY 64 216 .296 Atl 53 179 .296 ChC 60 207 .290 SF 64 225 .284 Fla 57 203 .281 Min 61 220 .277 NYM 67 242 .277 LA 52 200 .260 Sea 60 231 .260 Oak 50 223 .224 Bos 53 249 .213 Cin 42 212 .198 MLB 1966 6522 .301
The teams in bold italics (St. Louis, Texas, Philadelphia, New York, Chicago, Minnesota, Los Angeles and Boston) would be in the playoffs if the season had ended Monday.
Lots of fun stuff here:
– The 15 teams above average in POP have scored 4.77 runs per game, have a .334 POP and a .468 winning percentage.
– The 15 teams below average in POP have scored 4.74 runs per game, have a .270 POP and a .532 winning percentage.
– The top five teams in POP have scored 4.33 runs per game, have a .351 POP and a .392 winning percentage.
– The bottom five teams in POP have scored 4.74 runs per game, have a .230 POP and a .534 winning percentage.
– The Angels and Marlins, with their “diverse offenses” that “use their outs productively,” rank 15th and 23rd in POP, with a .296 combined POP. The same as the Yankees.
– The Yankees’ .296 POP is only 19th in MLB, but while Paul O’Neill and Jim Kaat were quoted in the article as saying how the Yankees are constructed differently than their championship teams, their POP minus pitcher stats in the ’98 postseason was .273, .211 in ’99 and .268 in 2000.
– The overall POP (minus the pitcher PO’s) in the the last two postseasons was .333, and there were 1.1 Productive Outs per game, compared to .97 per game this season — in other words, teams tend to play more one-run strategies in the postseason, which would lower scoring, and make pitching (and getting the most from your baserunners) look more important than it really is.
– Mark Bellhorn, villan of Olney’s article, is 17th in MLB with 11 productive outs; Juan Pierre is 57th with 8. Pierre’s POP is .400 to Bellhorn’s .379, though.
Total Productive Outs to Wins: -.276
Double Plays to Wins: .255
Strikeouts to Wins: .258
Productive Out Percentage to Winning Percentage: -.476
Total Productive Outs to Runs Scored: .121
Total Productive Outs to Runs Scored above Runs Created: .287
Productive Out Percentage to Percentage Exceeding Pythagorean Percentage: -.366
Nothing comes even close to correlating, and while that doesn’t prove there’s no relationship, the fact that as Productive Outs go up, wins tend to go down, may indicate that not only are Productive Outs not important, they may in fact be counterproductive. The positive correlations for double plays hit into and strikeouts are a fun counterpoint, too.
A poster going by the handle “Nod Narb” at Baseball Think Factory did an interesting, quick study that concluded that for a hitter as poor as Adam Everett (.236 GPA) — who happens to have a .512 POP — the only time when a productive out is more valuable than trying to get on base — which includes the risks of not making a productive out — is when there’s a runner on third with one out.
The truth about Productive Outs is that they aren’t something that are good in bulk, but only in limited context. A Productive Out in the ninth inning of a tie game is a good thing, but one in the first inning of any game is merely less bad than other outs. There is a heirarchy of plays in baseball:
Base on Balls/Hit by Pitch
Reach on Strikeout
Reach on Error
Out on Ball in Play
By making a productive out, you prevent any of the outcomes on the bottom third of the list from happening, but it also eliminates the top two thirds of the list. In context, that’s a worthwhile trade, but in general, it’s a terrible, stupid, Brock-for-Broglio-type trade. Counting them in context might give us a tiny bit of useful information — counting them without context gives us little more than noise.
ESPN promoted Olney’s initial article by saying:
Many say it’s a baseball sin to waste any of your allotted 27. But Buster Olney explains why productive outs are invaluable as opposed to the “Moneyball” philosophy of protecting them.
But Olney failed to prove that Productive Outs have any value, let alone being invaluable. The data indicates that they’re fairly worthless, and to use server space to track them is a waste of ESPN.com’s space.
Buster Olney either knows he’s wrong, and doesn’t have the guts to admit it, or he’s a fool. I’d like to believe that it’s the former, but anyone who has followed sabermetrics’ too slow progress in being accepted by the baseball establishment has to know there’s a good chance it could be the latter.