The Ultimate October Blueprint

October 6, 2015

Our regular readers know that we often cogitate over the roles both skill and luck play in sports and business.  The eve of the baseball postseason feels like a good time to revisit the subject.

Especially since we’ve found new data, even in an admittedly a small sample size.

In The Ultimate October Blueprint, David Schoenfield studies the past 5 post seasons (“when the strike zone started increasing in size and offense began to decline“) and draws a few tentative conclusions:

  • Don’t strike out – even though you’re facing better pitching!
  • Contact trumps power, although “the weird thing is that home runs in the playoffs have matched the frequency of home runs in the regular season.”    But none of the teams who led all playoff teams in HRs – in either league – have won the World Series in the past 5 seasons.
  • Use your bullpen, early and often.  Starting pitchers don’t fare as well as they start going through the lineup a second and third time, so don’t let them lose a game in the middle innings.  “Go to the bullpen. Hope they do the job.”

However… all of the past 10 World Series teams had a starter step up in the postseason:

The thing is, sometimes that starter is a Bumgarner or Verlander or Lincecum, but sometimes it’s an untested rookie like Wacha, a veteran having a so-so season like Lester (he had a 3.75 ERA in 2013) or a mid-rotation starter like Vogelsong. And sometimes it’s Colby Lewis…

The last team to lead the majors in starters’ ERA during the regular season and win the World Series? The 1995 Braves. In other words, having a season’s worth of gems from your rotation guarantees nothing in October — maybe bad news for all five NL playoff teams, who rank 1-5 in the majors in rotation ERA.

But here’s an indicator that may help: In looking for which pitchers may come up big in October, it appears a strong finish is important.

The data also suggest that velocity is overrated.  “Maybe when the chips are down, it’s those crafty veterans throwing in the low 90s and situational relievers who win you World Series.”

Schoenfield summarizes advice for when the contest isn’t long enough for the Moneyball math to work:

[The playoffs are unpredictable but] there are a few strategies that seem to work: Battle pitchers with two strikes and put the ball in play; turn the game over to your bullpen in those middle innings; rely on one starting pitcher if you have to; get some big home runs from unlikely heroes along the way. And maybe hope you have a starting pitcher who can throw five innings of relief on two days’ rest in Game 7 of the World Series.  [As Madison Bumgarner did for last year’s Giants. – ed]

The fan inside us is fascinated by the prospect of advantages gained in the short term, but over the long term our conclusion remains the same:

While big data may help make accurate predictions or guide knotty optimization choices or help avoid common biases, it doesn’t control events and can be undone by cluster luckModels are useful in predicting things we cannot control, but for those in the midst of the game – players or entrepreneurs – the results have to be achieved, not just predicted.

 

Related stories:

© 2017 Ballast Point Ventures. All rights reserved.