Friday, May 14, 2010

Counting stats can be fun...until they're not

Maybe it's too much to say that counting stats don't matter. But increasingly, we hear them described as too random, too dependant on team dynamics or situations, or too damn flukey.

It's possible that there is some truth to that, although it depends on just how seriously someone is reading (or reading into) the counting stats. If someone is going to say a dude with 100 RsBI is clearly superior to the guy with 98, then yeah, that's probably putting a little too much into them. But if you're willing to use them as a rough guide, then you might be able to find a little something worth holding onto.

This is especially true of our new favorite old counting stat: Total bases.

Our admiration for total base tallies crept up on us, until we realized how often we were using it to get a gague of who was performing well for the Jays' minor league teams. After sorting on the old reliable stat columns, including OBP and SLG, we find ourselves scratching our head to figure out what, if anything, they meant, especially given the varying sample sizes.

But if we wanted to sort quickly to get a sense of who was most productive on a team, we'd look under the TB column, and whoomp, there it is. That's who's touching as many bags as they can as often as they can, and in baseball, that's really the name of the game, isn't it. That's what we thought that we really liked about total bases: that it captures the totality of a player's offensive output all in one number.

Sadly, we've only now come to realize a few flaws in total bases. Our notion was that the number included walks (it doesn't), hit-by-pitch (nope) and stolen bases (nuh-uh) as well as hits and possibly bases taken on hits and on sacrifices (pshaw!), to capture the entirety of, or the whole of, or, what's the word we're looking for here? Oh yeah: we thought that TOTAL bases would account for the TOTAL number of bases that a player touched throughout the year.

But sadly, no. And as a result, we'll spare you our whole defense of Alex Gonzalez, which was predicated on using his 82 total bases thus far (fourth best in MLB so far, just behind Vernon's 84).

But you'd have to admit that our make believe version of total bases would be a pretty great number to have, wouldn't it? And if you were to divide it by the number of plate appearances, you could get a pretty holistic sense of what a player is accomplishing on the field.

The is what we get for trying to find profound answers through simplicity. It's a fool's heartbreak for us once again.

Engaged elsewhere
We're about to head away to spend some time elsewhere the next week, so we'll be out of the loop for a while. (And the last time we did that, the Jays announced that they were going to trade Halladay, so expect big things while I'm gone.)

In the interim, we leave you in the very capable hands of The Ack: A man who, we came to understand last week, can grill a steak with the best of 'em, and never lets a man's glass go dry. Truly, a prince amongst men.

Have fun, stay positive, and be good to one another. We'll see you soon.


eyebleaf said...

Happy trails, Tao.

Christopher Jones said...

The Tao for Ack. I make that trade any day of the week.

George Rekers said...

If by bases touched you meant bags tagged, I was the Albert Pujols of my high school. Until recently, nobody knew the underlying cause of that impulse, but well...just Google me.

Peter D said...

That is why OPS is such a good (quick) stat, it includes the total base average (SLG) and factors in walks and hit by pitches(OBP). Unfortunately you will have to look at the SB column to get a sense of a players speed.

The Ack said...

With Tao outta the way for a while, I'm going to take this thing straight lowbrow.

First post: "Lyle O-fer-bay does not hit enough homers for a first baseman". Then maybe I'll deride Ricciardi's five year plan and bitch about Vern's contract.

(I kid!)

Peter D said...

This is an interesting article.

It talks about how some teams scout umpires and mentions this is definitely something the Blue Jays do. I've always thought that this would be a smart thing to do.

NoisyFlowers said...

I also have dreamed of the stat you describe. OPS is a useful shorthand when evaluating a player but it's one of those stats like QB rating that doesn't really measure anything tangible. With OPS, you just take two numbers, one calculated on a scale of one and one calculated on a scale of four and combine them in the crudest way possible by simply adding them up. The number itself is just an arbitrary construct. It's a stat that somehow works in spite of the fact that there's no real logic to it.

I just can't understand that with all the math nerds in baseball, why no one would want to just create a stat that add up all the bases a player contributes and divides them by plate appearance. As a stat it would be so much more elegant and I dare say so much more eloquent than OPS.

Holden Ballfield said...

I want a stat that shows how many women a player has bedded. You could parse the stats according to the freakiness of the situation.

One man + one woman = .250
Two men + one woman = .300
One man + two women = .400
A-rod + Madonna + Kate Hudson + Vanilla Ice = 5000
Dustin Braden + A-rod = OVER 9000!!!!!

Anonymous said...

Interesting total base idea you got there... although you'd have to add this - if a stolen base is going to count as a total base, then a caught stealing would have to eliminate the total base(s) that the player tallied on that particular trip on the bases. Otherwise, pretty nifty

Steve G. said...

I just can't understand that with all the math nerds in baseball, why no one would want to just create a stat that add up all the bases a player contributes and divides them by plate appearance. As a stat it would be so much more elegant and I dare say so much more eloquent than OPS.

Well, you'd need something to also account for the difference between singles and other one-base events, since singles are marginally more valuable.

I imagine that no one has created the stat you crave because two others work better on either basis. VORP is AB-dependent, so it is a good gauge of what a player has actually done. And EQA is a good rate stat that encapsulates a lot of a player's offensive value. Both are probably more relevant than the hypothetical "bases earned per plate appearance" metric, which would probably give a distorted picture of the value of power hitters.

Gil Fisher said...

I like the all inclusive TB proposal. And despite its shortcomings, I will now use TB to sort my MILB stat pages - thanks for tip.

autolycus said...

Thomas Boswell of the Washington Post devised this very stat almost 30 years ago - total bases (including BB, HBP, and SB) divided by Plate Appearances (plus CS). He called it Total Average, and every year did a column on the league leaders.

NoisyFlowers said...

How is EQA calculated? Is it as stat the casual fan could understand and relate to or is it just something they'd have to take on faith? Total Average might not be perfect but at least it actually measures something tangible.

William said...

Wonder why Boswell's Total Average never caught on. It is indeed tangible and it seems, useful.

infieldfly said...

I went for years thinking that total bases meant TOTAL bases. When I realized it didn't, I started using OPS as the gauge for players.

I'm with you though. I'd be curious to see what a REAL total bases stat would show us.