THE BOOK--Playing The Percentages In Baseball

I could, but the trick there is that the zones represent different depths in different ballparks (i.e. Old Tiger Stadium’s “deep center field” is a long way away from Jacobs Fi… er, Progressive Field’s “deep center field.") The important measurement is how far that throw has to go.  But, if you like, I can take a look when I get a chance tonight.

Good point. 

I’ve asked this before to others, but I don’t think I got an answer.  Are the Retro zones “fixed” and simply overlaid to the park, or, are they stretched and skewed so that the zones themselves are fitted to the park in question?

The STATS zones are simply overlaid on the field, implicitly.  Their zones are made up of:
a) slices (22 slices for 90 degrees of the field)
b) distance

The combination of the two gives you zones.  Clearly, there is some combination of slice and distance that will be a HR at Fenway, but an out at Coors, according to the STATS zones.

I do not know about Retrosheet though.  If someone out there has looked at this, can you let us know?  Basically, look at how many balls are hit to the Retro OF zones at Fenway, and compare that to a typical park, like say the Skydome.

So, this goes to what Pizza is saying, as it matters whether we can compare the parks the same.

I can cross-check the retro and STATS zones at the various parks.  I’ll report back later tonight.

I have to think about Pizza’s methodology.  I am not sure you can infer how often a runner would be safe IF we ran from data that only has how often he was safe WHEN he ran.  Pizza?

This kind of stuff (great stuff) gets me thinking about what James said in one of his interviews and then later clarified, I think, and that I (and others) have criticized.

There is still a lot we don’t know about baseball (only because there is a lot to learn and a limited number of people doing the research and the writing), as James says.  When he says that in 100 years, we still won’t know anything (when he says “won’t know anything,” he is being a little hyperbolic of course), he means that we might have answered 1000 old questions, but 1000 new ones will come up.  I think he is right more or less.  So while a lot will have changed in terms of the absolute size of our knowledge and understanding base, not much will have changed percentage-wise (answers to questions ratio).

I agree that every question answered gives rise to a subsequent question.

This isn’t limited to baseball though.  It’s true in technology and any other knowledge-based research.  If every time we learn something, something new is on the horizon, then 100 years from now, we’ll decide that we knew nothing in 2008.

Once we have it all figured out, we’ll be like Joshua in War Games.

I’d like to see Gary Huckabay and James debate that topic (whether or not baseball analysis is dead)

Gary should do an about-face on that one.  He’ll get back the respect of many. 

Baseball Analysis is alive and well! by Fake Gary Huckabay

In my rush to make a bold statement, and in my myopic view of focusing only on MLB management which you guys couldn’t care less for, I made a seemingly silly statement that Baseball Analysis is Dead.

Far from it, it is breathing new life.  PITCHf/x is the gold mine that we’ve been begging for, and will finally bridge whatever gap exists between the scouts and the basement dwellers.  PITCHf/x is what baseball is all about, and we’ll finally get the insight we need as to how the batter/pitcher matchup is controlled.

What we really need to move this to the forefront is to get more analysts.  Baseball Analysis is alive, and we need it to thrive.

Thank you for making me see the light.

MGL, there might be an issue of selective sampling there, but it seems that distance is a pretty good predictor (as these things go) of whether the runner will be safe.  I suppose the only data we have available are to study when the runner went.  It could be that the arm of the outfielder was an influence, which is something that I might try to look at down the road.  More to come, surely.