Orioles Think Tank

Orioles Coverage for the Information Age

DIPS 3.0

Posted by Mike on February 20, 2006

A couple of days ago, David Gassko (of the Statistically Speaking Blog) published his second article on DIPS 3.0 over at The Hardball Times. Like the original work by Voros McCracken, it’s relatively simple, straight-forward, and incredibly meaningful. These guys have a real knack for making me feel stupid for not thinking of things first.

Years back, McCracken discovered that replacing a pitcher’s BABIP with the league average BABIP and recalculating his ERA based on this was a better indicator of future ERA’s than current ERA. If I lost you in that last sentence, he’s essentially saying that pitchers have, at a minimum, much less control over whether or not a ball in play becomes a hit than previously believed.

David Gassko decided to take this one step further. We already know that pitchers generally have consistent GB and FB rates, and you might as well lob K’s, BB’s, HBP’s, and infield flies onto that pile. HR’s are an interesting case and they are accounted for as a reflection of ((League Ave. LD rate*(Leag. Ave. HR/LD))+ ((FB rate*(Leag. Ave. FB HR/FB)). What does not correlate year to year is the amount of line drives a pitcher allows (as is reflected in Gassko’s HR calculations). So, with some surprisingly simple math, Gassko replaces the line drive rate with the league average and then assigns singles, doubles, etc. according to league average distributions (for instance, a line drive becomes an out 29.3% of the time). Another thing that becomes apparent is that it is silly to differentiate between earned and unearned runs at this point. After all, you’re already separating the pitcher from his defense. This was originally a problem because ground balls more often result in an error, so GB pitchers’ ERA more often overrates their effectiveness.

After normalizing line drives and then redistributing all the other outcomes, you’re left with a number that is a more accurate reflection of performance than ERA, DIPS ERA, and FIP– DIPS 3.0. Without further ado, your 2005 Orioles:

DIPS 3.0 … RA/9 … Diff.

BJ Ryan … 2.92 … 2.56 … +0.36

Chris Ray … 3.90 … 3.32 … +0.58

Aaron Rakers … 4.18 … 3.29 … +0.89

Erik Bedard … 4.36 … 4.19 … +0.17

John Parrish … 4.47 … 3.12 … +1.35

Daniel Cabrera … 4.57 … 5.13 … -0.56

Jorge Julio … 4.59 … 6.28 … -1.69

Tim Byrdak … 4.61 … 4.78 … -0.17

Todd Williams… 4.69 … 4.01 … +0.68

Bruce Chen … 4.71 … 4.29 … +0.42

Jason Grimsley … 4.96 … 6.14 … -1.18

Steve Kline … 5.06 … 5.02 … +0.04

Rodrigo Lopez … 5.08 … 5.42 … -0.34

Rick Bauer … 5.12 … 9.72 … -4.60

Sidney Ponson … 5.17 … 6.70 … -1.53

James Baldwin … 5.24 … 4.47 … +0.76

John Maine … 5.40 … 6.75 … -1.35

Hayden Penn … 5.56 … 7.09 … -1.53

Steve Reed … 5.60 … 6.61 … -1.01

Eric DuBose … 6.21 … 6.44 … -0.23

You can see the results for yourself– Sidney and Jorge weren’t quite that bad, BJ and Chris Ray weren’t quite that good, and so on.

One caveat: This is highly dependant on the subjective art of scorekeeping. At some point, this will have to be addressed and quantified– as far as this can be done. I don’t think this is enough to negate the meaning of the numbers above, but quite frankly, I have little clue as to the variance in the classification of GB’s, FB’s, and line drives among different scorekeepers.

All in all, I think this is some fascinating work. McCracken taught us that pitchers have less control over balls in play than previously thought. And, as Gassko tells us in his article, his research shows us why.

Advertisements

17 Responses to “DIPS 3.0”

  1. Greg said

    I checked out the articles. Gassko is definitely one to read. I used to read THT all the time but it seems watered down these days. This definately made me intrested again.

  2. Mike said

    If anyone out there knows how to insert charts into blogger, please email me at

    Mike@OriolesThinkTank.com

    These unsightly X…X…X charts have got to go.

  3. Anonymous said

    DIPS 3.0= WHAAAAT?

  4. Mike Devereaux said

    Curse you Mike @ Orioles Think Tank… Your latest post has done me the disservice of an uncontrollable anticipation for the release of DIPS 4.0… I’ve had to take 3 cold showers today just to keep from soiling myself. DIIIIIPPPSSSSSSS!!!

  5. Anonymous said

    I was left wondering:

    What does this mean for the 2006 O’s?

    How does this help us analyze where the O’s fit in the AL East?

    So in that regard, it was a waste of time reading the article.. it didn’t really make a point that was helpful to me and the 2006 O’s.

  6. Mike said

    You probably meant to post this at “Where The O’s Fit In The AL East Think Tank” or maybe the “Things That are Helpful to Anonymous And The 2006 O’s Think Tank”.

  7. Mike said

    But seriously, it doesn’t help you analyze the 2006 Orioles to know that Todd Williams’ performance was not as good as his results or that Hayden Penn’s performance exceeded his results?

  8. Local Yokelson said

    Raaar!

  9. Anonymous said

    What those stats show me regarding last years performances tells me nothing about the 2006 Yankees, RedSox, or BlueJays and how we compare to them using those stats and based on that comparison what we might expect to happen THIS season… COMPLETE WASTE OF TIME. Good thing I am bored and have time to waste..

  10. Mike said

    Good thing I went with “DIPS 3.0” and not the original title of “The 2006 Yankees, RedSox, or BlueJays and how we compare to them”

  11. Anonymous said

    Well if decide to post that information you’ll have at least one happy reader.

  12. Anonymous said

    Hey Mike, here is an example of what your post is missing:

    http://www.orioleshangout.com/2006/cook020306.htm

    A thesis, supporting information, and a conclusion. Excellent read.

  13. Anonymous said

    Sorry, left out part2.

    http://www.orioleshangout.com/2006/cook021306.htm

  14. Mike said

    For those interested in the complete data- follow the link to the article and download the spreadsheet at the bottom.

    I’ve actually tinkered with it and added a %Diff column that indicates what percentage DIPS 3.0 is different from RA/9, which David Gassko is going to allow me to post on this site… as soon as I figure out how.

  15. Mike said

    My post is a review of an article on DIPS 3.0

    My conclusion is that I like it.

  16. Anonymous said

    He’s clearly just spreading the word about Dips. What exactly would be accomplished by posting the Yankee numbers anyway? Some would be better, some would be worse, and the O’s still finish in 4th place.

    Personally, I like that Cabrerra andPenn didnt pitch so bad as I thought last year. I figured that out for myself by looking at the numbers.

  17. Greg said

    I followed the link to the original article. David Gassko says that this stat is more of an analysis of performance and not as much a predictive stat. In that context, I get that Mike was just trying to review a new stat that he thought was really cool. The O’s pitchers DIPS is included as more of a demonstration than analysis. On the other hand, I’d like to see if maybe any AL East staff’s over/under-performed as a whole. But to anonymously call this a “waste of time” speaks more to your lack of understanding the point than anything else. I found the work very interesting, if not specifically relevant to the 2006 O’s.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: