OBP/BA question

[GoudaBrew]

I'm not much of a stat guy, but I like casually reading about much of the analysis discussed on this board.

 

My question regards OBP and batting average. I have heard several times that when discussing OPS, 1 point of OBP is three times as valuable as 1 point of SLG. If this is the case, has anyone tried to determine how much value 1 point of BA has in OBP? Obviously, a hit is more valuable than a BB or HBP. How much more "valuable" is a player who hits .300/.370 than a player than hits .250/.370?

 

I think this question came to me when I tried to understand why Adam Dunn's value seems so low for a player who posts near elite OPS #'s (and great OBP #'s) every year.

[logan82]

How much more "valuable" is a player who hits .300/.370 than a player than hits .250/.370?

 

Depends on their SLG/OPS. The guy who hit .250 is likely to have a lower SLG unless he hits for A LOT of power. Dunn vs Ramirez 2008 is a good look at guys with almost equal value and got them in drastically different ways. There are things called linear weights. The one I like is wOBA.

[TooLiveBrew]

I have heard several times that when discussing OPS, 1 point of OBP is three times as valuable as 1 point of SLG.

 

Unfortunately this will just be more hearsay, but what I've seen most often is that OBP is somewhere around 1.6x-1.8x more 'valuable' than SLG. Iirc instead of 'valuable', it's that OBP correlates to runs scored 1.6-1.8x more than SLG.

[rluzinski]

Average run value of each batting event, from 1999-2002:

 

http://www.tangotiger.net/RE9902event.html

 

Look at the last row....

 

Single: .474 runs

Non-intentional Walk: .33 runs

 

So a walk has about 2/3rds the value of a single.

 

If you compare these two lines:

 

A: .225/.350/.450

B: .300/.350/.450

 

Both lines have about the same overall value. Player A has more extra base hits but Player B has more singles over walks. Of course, player B will be the consistent player, from day to day. I don't think there's any real value in that but fans sure do.

[logan82]

Will player B really be more consistent? I would think that a player that draws walks would be more consistent that a player who relies on hitting to drive his stats.

[rluzinski]

They both get on base the same number of times. Player A probably has much more of the least and most valuable batting events, however (walks and HRS). Also, since HRs are relatively rare, there will be times when even a high HR batter won't hit one for an extended period of time, just by pure chance.

[logan82]

Good point. I guess in many peoples' minds player A would appear to go on hot and cold streaks. Hot when he is hitting, cold when he is walking.

[1992casey]

Iirc instead of 'valuable', it's that OBP correlates to runs scored 1.6-1.8x more than SLG.

 

I'd have to find the chart I remember seeing, but I believe OBP and SLG correlate to runs scored almost exactly the same, with SLG getting an ever so slight advantage. OPS would do better than either, with some of the more advanced stats edging it out--again, ever so slightly.

 

I think 1.6-1.8 (or whatever) is a factor that would tweak the accuracy of OPS to make it a bit better.

[kramnoj]

Casey, is this the chart you are talking about?

 

AVG: 0.7244

OBP: 0.8340

SLG: 0.8421

TA-orig (~TB/PA): 0.8739

BOP (~TB/outs): 0.8955

FOES: 0.8990

OPS: 0.9231

ABSO: 0.9250

EqA: 0.9268

OBP*SLG: 0.9310

2OPS (2*obp+slg): 0.9323

GPA ([1.8*obp+slg]/4): 0.9326

RAA/PA: 0.9344

wOBA: 0.9350

R/G: 0.9355

 

from: http://jinaz-reds.blogspo...-do-i-keep-using-ops.html

[1992casey]

It was a different chart that ranked everything in just about the same order. The only difference was that everything had correlations that were just a bit higher. I believe BA was a bit over .8, OBP and SLG were a bit over .9, OPS was around .94 or .95, and the nouveau stats were basically in the .96s.

[rluzinski]

I think we have to be careful with this whole correlation thing. A higher correlation doesn't necessarily mean that a point of SLG scores more runs than a point of OBP. I used multivariable liner regression in a long lost post to try and estimate the relative value of OBP and SLG but was scolded by those who know the stuff better than I do for doing so. There are better ways of estimating this.

 

We know the average run value of reach batting event with some pretty good confidence. The best way to find these values is with a simulator, if you can trust "The Book" (which I do). They are very similar to the linear weights values floating around the web.

[rluzinski]

Here's a post that describes some of the problems with regression analysis:

 

http://www.insidethebook...._linear_weights/#comments