Yet, in our hearts we know there are dangers in that piece of data—especially early in the season when sample sizes are down in the 5s and10s. Still, despite sample size issues, we use these numbers to some extent—even if we pretend to be good at ignoring them. We’re humans, after all, and most of us think we’re more rational than the average bear.
Des Moines GM Ed Murphy’s comment in a Yellow Springs thread—where he said he threw three lefties at Yellow Springs in hopes to take our latest series—got me to wondering about this. I knew for sure that the Yellow Springs lineup is actually pretty tough on lefties, and likely tougher on lefties than righties (*).
I say this because over the last half of the season, I’ve taken to looking at team OPS along with the info screen’s W/L data. Yet, (1) I hadn’t really looked at it very closely, and (2) the Info screen says we’re much stronger against RHP. Taking the risk of giving away the farm here, I asked myself—how often am wrong if I use game’s vsR/vsL data? (you guys may be well ahead of me, but I am soon to be an old man, set in my ways, but this is the first time I really tried to quantify anything here).(*) Note … In our case, it’s particularly interesting, because if you throw a lefty, you neutralize one of our best hitters in Lucas McNeill. In fact, full disclosure here, I’m using a lot of 7-man line-ups, and to be clear, if you throw a particularly tough lefty at us, you take McNeill completely out of the game. This is because in those cases I have often replaced McNeill with catcher Nate Wood, sliding Aaron Stone to DH and Dimiao Laqui to 1B. You’d think this might hurt our offense, but the fact is that Wood is out-OPSing McNeill, so that move seems to make no impact on our offense, and serves to greatly improve our presence behind that plate. Stone is a below mediocre defensive catcher.
Of course, we hit RHP pretty well, too. So, in reality, the YS9 experience is kind of like robbing Peter to pay Paul. Pick your poison.
My first thought was: how do I study this question about handedness?
I started by looking at opponent’s slash lines and homer rates.
Unfortunately (you know me, right?), that got to be complex--or, at least, since OOTP is horrible at giving us valuable information like this, it got to be a lot of work to gather these splits in ways that helped make decisions. So instead, I settled on simply OPS—which seems fair enough. OPS correlates well to runs scored, and are easy to find in-game (just go to league stats and do a split).
For example, using my own Yellow Spring club as fodder, we finished the season with OPS vs. LHP at .862, and OPS vs. RHP at .831. That’s a 31-point split, which is kind of medium for the league. Our is also a but reversed in that most teams hit RHP better than they do LHP. So the bottom line here is that if you believe OPS correlates to runs scored, you’re probably better off on the whole if you throw RHP at Yellow Springs rather than a lefty.
But if you look at our records you’ll see we play .554 ball vs. LHP and .642 vs. RHP.
Interesting, eh? What gives?
The real answer is, I don’t know.
There are lots of reasons why records can fluctuate. Maybe our better pitchers happen to face opponents who throw righthanded pitchers. Maybe teams that threw lefties at us are generally just better teams. Maybe there’s a huge margin between very good lefties and not so good lefties and we’re padding our stats against the poorest. Maybe we faced a righty, then he got replaced with a lefty, who we blew out of the water. This is the thing about low sample sizes. You have to make your own assessment and go. At question, though, is which number you trust more as a systemic measure—Win% of game based on starter handedness, or OPS vs. handedness?
I can’t speak for you, but I can say that I’ve come to look at OPS before I worry about a team’s W/L against handedness.
OKAY…SO WHAT ABOUT YOUR QUESTION?
Oh… yeah…to recall the question at hand, I wanted to get a feel for how often I might be wrong if I look at W/L records. Assuming you’re with me on team OPS (that it’s “more accurate,” whatever that means), we now have a barometer with which to judge that.
To do this, I looked at every team’s OPS vs. R/L and compared them to their W/L records vs. R/L starters as listed in the game. Once I had that data, I tabulated what decision I should make based on each case. Where both indicators said the same thing, I said the use of W/L was “right.” If they differed, I claimed victory for OPS, and declared W/L as “wrong.”
Here is my data:
If you’re with me here, this measure says that in 5 of 30 cases (16.7%) the use of W/L records to determine preferred starter handedness was “wrong.”
A few items of note:
(1) The average OPS split for the league was .017, with the league hitting RHP slightly better than LHP
(2) The vAvg column is the team’s performance based on that average split (Yellow Springs, for example, has a 31-point spilt, which is 14 points bigger than league average)
(3) 11 of 30 teams OPSed vLHP better than they do vRHP
(4) We’ve recently discussed the Huntsville system in one of Vic’s TN. Look at how their system is split. Huntsville OPSed fairly well vs. RHP (4th best in the Johnson), but was absolutely atrocious vs. LHP (worse than all but Brooklyn). Their .102 split was shocking, the biggest of any BBA team.
(5) Edmonton had the smallest platoon gap in the league (4 points). Mexico City is next at 8 points.
There are, of course, issues with this idea of using OPS, too—not the least of which is that the roster a team had in creating these numbers may not be the roster they have now. Teams trade stars, or injuries put them on the shelf. Some might have called up a hot young guy who’s tearing the cover off the ball.
It all adds up to say that there’s no super-simple way to answer this question, nor, of course, should there be. It’s certainly possible that using the W/L records are just as predictive as OPS, just as it’s possible that there are even better and simpler things to do tha use OPS, too. You just don’t know. At least, I don't. Alan probably has us all covered.
Still, it’s fun to try, right?
Fun to think that the way you see things gives your guys the edge they need. And in that case, who cares really if it’s right? The fun is in the chase, after all. As a team without a Landis, that’s my story and I’m sticking to it.
Well, that and OPS, anyway.
