During the week I read something interesting by Joe Posnanski during his conversation with Bill James on why walks are under-appreciated:
This year teams that walk zero or one time win about one third of their games. If they walk two to three times it jumps up to about 45 percent. If they walk four or five times they win about 57 percent of the time. And if they walk six times or more their winning percentage is .646. Those numbers have stayed remarkably consistent for decades now.
This leaped out at me after Thursday's game, where we took seven walks off the Giants' starter, but only converted them into one run - though we did win the game, mostly thanks to Max Scherzer. After the jump, I'll take a look at the Diamondbacks' number in the area of free passes (and a couple of others) over this season and last, to see whether we live up to the expectations of Posnanski. I'll also see what effect team strikeouts have on the win percentage for Arizona, and who leads the team in 'unproductive outs', when we combine K's and infield pop-ups.
Over the 223 games Arizona has played from the beginning of last season up until Thursday, here is the breakdown by the number of walks, along with the results of those games:
|0 or 1||13-27||32.5%||7.88|
|2 or 3||29-48||37.7%||7.96|
|4 or 5||36-28||56.2%||8.64|
Two of those numbers are almost dead-on those projected by Posnanski - we win fewer games than expected with 2-3 walks, but more than expected when we walk six or more times. My first thought was that perhaps the increase was partly due to a increased hits [lots of walks = poor pitcher = lots of hits], so I went back and added the last column on the end - the number of hits the Diamondbacks averaged in those games. While there is something of an increase, it only works out to about an extra one and a quarter hits between the low and high categories - which doesn't seem enough to more than double the win percentage.
As a contrast, I thought it might be fun to do the same thing for strikeouts. Mark Reynolds, in particular, has come in for a lot of criticism because of his strikeouts, so is there a correlation between the number of K's in a game and our win percentage? I divided the 223 games up into roughly-equal quarters: five strikeouts or less (42 games); six or seven (59); eight to ten (68); and eleven or more (54). Here's the same table as above, divided up into those sections:
While there is still some correlation betweek low strikeouts and win percentage, it's a lot less pronounced - the effect only really becomes significant, when you get up into double-digit K's or more, even though the hits also start going back up [this may be due to extra-inning games, which I should perhaps have filtered out of the data. Oh, well...]. This perhaps reflects contests against a really dominant pitcher and/or staff, but otherwise, there's little evidence that strikeouts are 'bad'.
This ties in with other research on the topic, such as a great Baseball Prospectus article Just Another Out?, in which Ryan Wilkins looks at the correlation between strikeouts, run production and various other measures of offensive and pitching production. Or, as far as hitters go, the lack thereof, since here's a summary of Ryan's conclusions on the topic. Might be worth holding on to this for the next time someone belabors Reynolds with the strikeout stick.
There is virtually no positive correlation between a team's strikeout totals and its runs-scored totals. When it comes to offense, an out is an out is an out. On an individual level, the evidence against strikeouts as the scourge of the earth only gets more damning... [Oh, let's give this next sentence the prominence it deserves]
No matter how you slice it, it just doesn't appear that strikeouts have much of an effect on a team's — or an individual's — ability to produce runs.
They may be frustrating, but there just doesn't seem to be data to show that the strikeout is worse than any other kind of failure to reach base. While the 'productive out' may not be possible when you don't put the ball in play [dropped strike threes excepted], this is negated by the seriously detrimental effect of grounding into a double-play.
And, of course, there are other unproductive ways of being retired - the infield pop-up would be the most obvious case, and Mark Reynolds has massively improved his numbers there, dropping from 14% in 2008, all the way to 4% [league average this year is 13%]. Someone suggested that we combine the numbers for those and K's, and see what the totals are for "unproductive outs". Note, it's not correct just to add the numbers together, since K% is based on all at-bats while IF% is a percentage of fly-balls. Therefore it's wrong to say that 38% of Reynolds' at-bats are unproductive, and I'll slap anyone who does. But we're talking quick and dirty here, so it's exactly what I'm doing...
So, when you combine these two areas, Mark Reynolds is not actually all that much ahead of league average. No surprise Chris Young tops the chart, but with the recent signs of a revival in his fortune, it'll be interesting revisit this chart in a month or two and see if/how things have changed.