The Stats Fallacy: #5 In A Series

"Stats don't show a guy's guts. They can't tell you the human side, help you recognize fear, or figure out who is having a good day or a bad day. I look at the numbers all the time, but people who play fantasy games don't have to look a guy in the eye or try to help restore his confidence." --Dusty Baker

This is a familiar refrain from managers, who after all, are paid to make personnel decisions about guys they spend a lot of time with. Is it right or wrong?

Neither. It's a false dichotomy.

The question isn't, and never was, stats versus scouts. It's a fantasy of Michael Lewis that drove tradition-bound observers batty. The truth is and always has been that mathematical analysis and the softer human arts complement each other, and any team relying on just one or the other is doomed to failure. Numbers can enlighten us objectively about players' past results, but only a human eye, brain and heart can make observations about their physical, emotional and psychological health going forward.

In fact, the truthful dichotomy is: stats versus stats. Is Dusty making decisions based on batting average and saves and pitching wins and home runs or is he dialing up stats that actually matter, like on-base percentage, batting average on balls in play (BABIP), slugging average, etc?

If Neifi Perez bats .372 in the first half of the season, has he suddenly transformed into a hitting machine? The human brain tends to try to make sense of the unexpected, so it concocts explanations: Perez is squaring his hips or seeing the ball better or swinging through the zone. Perhaps, but if his BABIP has suddenly skyrocketed, particularly in the absence of a change in the types of balls he's hitting, (i.e., grounders, line drives, flies, etc.) it really tells you he's just been lucky and can expect a big ol' bag of return-to-earth in the second half. The same calculation works for pitchers: if a guy's BABIP-against spikes, the seeing-eye grounders have been getting through and he's primed for "improvement" as the year progresses.

Fans who dismiss sabermetrics "because numbers don't tell us everything" make me laugh. If any analytical system told us everything -- or anywhere near it -- the game would be soulless, stultifying and utterly predictable. The beauty of baseball is that it's the opposite, and that while a good mix of human evaluation and advanced analytics can improve a team's decision-making, it will nonetheless have to endure a great deal of failure. The best teams, after all, still lose 60 games.