THE BOOK--Playing The Percentages In Baseball

Saturday, January 30, 2010

Moneyball and Walks

In this article, Sky Andrecheck tests the hypothesis that after the publication of Moneyball, there was more of an emphasis on drawing walks in MLB, such that we might see an increase in the number of walks per PA since around 2003 when the book was published.

He says:

If teams were now coaching players to draw more walks, we would likely see a rising trend in the number of base on balls since the publication of Lewis’ book. Do we?

He then shows a graph of the walk rate over time from 1982 to 2009.

After the graph, he states:

In fact, the above chart of walks per plate appearance below doesn’t support that line of thinking. While the amount of walks has fluctuated over the past 30 years, there’s certainly no evidence that walks are overrunning the game. Walk rates peaked in 2000—three years before Moneyball hit the shelves. Last season, seven years after the book’s debut, players walked no more than they did during parts of the mid-1980’s, when nobody had ever heard of OPS and Bill James was just an egghead who couldn’t know a thing about the game because he had never played it.

Now, I am not disputing his conclusion (it may or may not be correct), that Moneyball had no effect on walk rates in MLB (and in the minors, as he also discusses), but…

What he fails to look at or discuss, were other influences in baseball that might effect the walk rate and might overshadow any changes in approach by management and the players themselves.  I am talking mainly about the strike zone, although there may be other factors.

If you look at his chart, you will see a huge valley (low walk total) in 2001, a slight decline until 2005 and then a clear rise through to the present.

Well, the valley in 2001 was almost certainly caused by the “new” strike zone.  In case you don’t remember, Sandy Alderson, in response to the offensive explosion since 1994, and the supposed “incredible shrinking strike zone,” told the umpires (and even held “clinics” for them in ST) to raise the strike zone to 8 inches or so above the belt.  Before that, the de facto top of the zone was the belt buckle.  The effect was a much higher but narrower zone, although overall it was larger with more K (and HR due to the higher zone).  The reason that the walk rate continued to decline (presumably, and in my opinion) until around 2005 was that in 2001 when Alderson gave the edict, not all of the umpires were on board, but that slowly they came around and some of the old-timers retired of course.  The likely reason that it started going up in 2005, at least in part, was that Alderson was no longer involved in MLB management and no one took the reigns from him in terms of enforcing the new strike zone, such that it started slipping back to the way it was prior to 2001.  If you watch a game now, you can see that the current zone is somewhere in between the small zone pre-2001 and the “new” zone in 2001-2005.  Anyway, that’s my take on it.

Bottom line is that Sky ignored the huge influence that the strike zone has on walk rate and because of the always changing strike zone over the years, there is simply no way that you can infer anything about the players without somehow at least adjusting for the changing zone.  At the very least, he should have also looked at K rates or called strike rates, since if there were indeed an increase in walk rates due to an emphasis among teams and players on drawing the walk, there is no particular reason why K’s would increase or decrease along with a change in walk rates.  However, if you find that walk rates decrease AND K rates increase, there is a good chance that that is due to a change in the K zone and not because of any change in approach by the players.