Thursday, October 27, 2005

That's Efficiency, Homes

The common template for printing a ballplayer’s stats is .250/.250/.250, and since everyone knows what each value is, I’ll skip the part where I explain that it’s batting average, on base percentage, and slugging percentage (oops).

Yesterday, I opined that the OPS statistic – the addition of the last two .250’s – should be replaced by what I believe is a more efficient statistic. However, the more I think about it, the more it makes sense for it to completely take over for slugging percentage. If a walk is as good as a hit (as one school of new thought in baseball, hence focus on the OBP), shouldn’t walks be averaged into slugging percentage?

Of course, that would make the OBP statistic a bit less relevant, since walks would be accounted for in slugging percentage. But who’s to say that OBP is necessarily the best indictor of efficiency?

I’m relatively new to this sabermetrics thing, but I’m a bare bones kinda guy. I would rather have the parts of a formula and evaluate them in my own way. If I come up with something similar to what someone else came up with, so be it. But I like to reason these things out in my head rather than rely solely on someone else to provide the statistic.

The most fascinating statistic I came across was simply labeled “outs.” What a stripped down concept, I thought. After browsing a sabermetrics glossary for exactly went into that stat, I had to slap myself in the forehead. D’oh! It made perfect sense: double plays and caught stealing count against you (I had figured on sacrifices).

To anyone entangled with sabermetrics, I may seem like a doof for not figuring this out earlier. And I’ll accept that assessment of myself, because like I said before, I’m new to this way of thinking.

ANYWAY. After discovering the outs statistic (if I haven’t seen it, it’s new to me), I scoured the rest of the “special batting” line at baseball-reference.com, but found no statistic that I think makes perfect sense: non-out percentage.

The formula is rather simple. Instead of walks and hits divided by plate appearances, this would be non-outs (plate appearances minus outs) divided by plate appearances. The idea behind it is that if you are rewarded for walking – i.e. you didn’t make an out – then you should be penalized for a caught stealing (since you effectively took yourself off the base paths and created an out) and for grounding into a double play (since you not only made an out, but you took another guy off base).

By the same token, if we want these stats to be as efficient as possible, why wouldn’t stolen bases minus caught stealing be factored into slugging? That’s a free base, and while there obviously was an amount of risk to the action, guys usually steal bases at consistent clips. So why not reward them for turning a single into a double via a little thievery? This may be an argument for another day.

So here we go again, this time the top ten guys in OBP in the AL, followed by their non-out percentage (NOP). One of these days, I’ll learn how to make one of those fancy pantsy easy to read charts:

Jason Giambi -- .440, .428
A-Rod -- .421, .401
Travis Hafner -- .408, .393
David Ortiz -- .397, .379
Vlad -- .394, .364
Jeter -- .389, .359
Man Ram -- .388, .357
B-Rob -- .387, .359
Mike Young -- .385, .355
Marky T -- .379, .355

Sure, everything seems to stay in order, save for B-Rob, whom Billy Beane would be salivating over if not for Mark Ellis. Speaking of Ellis, where the hell is he on this list (that I scoured off ESPN)?? Yes, that deserves a double question mark, because he has a .385 OBP, and has more at bats than Giambi. What, does ESPN think that just because they haven’t heard of him that he doesn’t count? Well, his NOP is .354, if anyone cares.

(Aside: I thought official stats were official once you had a certain amount of at bats. Is it a certain number of plate appearances? Because that’s the only explanation that would make sense for Ellis’s absence from this list – he also wasn’t on mlb.com’s list. He fell short of 500 plate appearances, at 486. Does anyone know?)

If that list proves anything, it’s that a lot of guys fall into that .350-.360 area. This is effective for evaluating players, since it proves that you’re not getting much more efficiency out of Manny than you are out of Mike Young, save for the power numbers. If I had my way (and I never will), these two players would be compared like this:

Young: .331/.355/.552
ManRam: .292/.357/.645

I should clarify what I said in the last paragraph: Manny is negligibly more efficient than Young when it comes to avoiding making an out, but much more efficient in other areas. From just a glance at the numbers, it would appear that Manny walks more, avoids double plays better, and gets caught stealing less (yes, equal, yes).

It is my belief that these statistics, NOP and OBS (which I may rename, considering its similarity in acronymity to OBP and OPS), will help better judge a ballplayer from an efficiency standpoint. These are the numbers that I will use in evaluating the returning Yankees and the players I believe they should pursue this off season.

But it won’t stop there. There are also other stats out there, such as at bats per strikeout and bases on balls to strikeout ratio that are not only telling statistics, but seem to remain more consistent over the years. That is for next week, however, as I would like the dust to settle before I get into anything serious about trades and free agents.

Congrats to the White Sox. I have to say, nothing would be scarier than bringing this team back as is next year.