Introducing The xWHIP Calculator

If you have the old version of the xWHIP calculator (anything before v.1.4.3), please download the new one (available through the link below), as it will give you the most accurate projection.

Enjoy!


____________________________

I am forever locked in Mortal Kombat for the souls of sports fans everywhere. Statistics are my science and 'the immeasurable character of men' is the obsolete religion of blind faith. My job is to prove that God doesn't exist and that athletes are merely cold, metal machines with no hearts or souls.

Earlier this week, I projected Cliff Lee's prospective rest of season WHIP. The formulaic process of calculating his xWHIP got me thinking and I spent the (entire) morning creating an xWHIP calculator (visually based on the THT xBABIP Quick Calculator). If you would like a copy of the program, you can download the xWHIP calculator by clicking here. The password to utilize the excel sheet is soto18.

Explanation of the xWHIP Calculator

According to Gameday data, circa 2005-2010, BABIP by batted ball type is general broken down as follows:
  • Popups: .008
  • Groundballs: 0.237
  • Outfield Flyballs: 0.269
  • Line Drives: 0.733
This data includes home runs, which is why the Outfield Flyball xBABIP is so high. If you take home runs out of the equation, the xBABIP for Outfield Flyballs and Line Drives fall to .174 and .727, respectively.

I've taken the above numbers and paired them with a formula that normalizes a pitcher's line drive percentage to 19% and spits out remaining balls in play (BIP) data. The calculator also features a defensive adjustment so that you can account for a pitcher's team defense. The defensive adjustment operates under the assumption that all "saved hits" were of the singles variety.

Below is a picture of the xWHIP calculator. The numbers plugged into the model for the picture are those of Tom Gorzelanny through July 17, 2010.



The grey cells are for data you should input. The green cells feature the xBIP data per IP. The blue and orange cells feature xWHIP and xHit calculations. The data cells are pre-formatted to visually round all numbers to keep the sheet clean, though cells will retain the full value of any number entered.

I hope everyone enjoys this. If you have any questions/concerns/comments/criticisms, please email them to gameofinchesblog@gmail.com, with the subject line xWHIP Calculator.

On a final note, I would like to give a special thank you to Derek Carty, who (possibly unknowingly) helped me create this xWHIP calculator.

3 comments:

Brimhack said...

This is pretty similar to some research I had posted online about a month before this:

http://www.fangraphs.com/forums/topic.php?id=3643 , http://brimhack.com/2010/xwhip-expected-whip/

The results of the simplified form of xWHIP yield results quite similar to the formula provided by David.

David "MVP" Eckstein said...

Dear Brimhack,

I can assure you that the xWHIP calc I developed was done completely independent of your research.

-For one, it is based on Game Day data. I am not sure what your regressed BIP data is based on.
-Second, prior to you posting on my blog, I’ve never heard of you or seen your research. I’ve only once ever before visited the Fangraphs comments section and that was to post an advertisement for our site.
-Third, the theory of xWHIP is nothing novel.
-Fourth, my model incorporates park factors and defense. Yours does not
-FIfth, your model incorporates K/9. Mine does not.
-Sixth, xWHIP was not something I created “a few short weeks” after your post went up as you so claim. It’s been a project I’ve been working on since I tried forecasting pitchers in the early season this year and realized I had no specific method laid out. You can see hints of my xWHIP Calc in earlier posts such as one I did about Cliff Lee for The Hardball Times. Other references in it can be found in my xBABIP-formula stripped hitter projections and xBIP articles which highlight research about BIP regression. There are other places I've talked about the theory of xWHIP as well, including email communiques written well before your posting.
-Seventh, I never claimed the formula to be novel and of my own creation. I state in my xWHIP post that the data is from Game Day. All I did was organize the data in to an organized math function. I did spend the hours of time it took to create a working visual version of the xWHIP calculator however.

As a law student, I do not take kindly to allegations of fraud. To think that two people in two places of the world might have similar ideas is nothing insane. You might say that your work is actually “eerily similar” to work that Derek Carty did a few years back on a DIPS WHIP formula.

Anyways, glad to know you are doing good research. I’d appreciate it, however, if you wouldn’t imply I am stealing your ideas. Thank you.
-Jeffrey/David MVP Eckstein

Brimhack said...

As I indicated in my email:

"Sorry if I came across like I was accusing you; my intent wasn't to accuse you of plagiarism. I've been interested in xWHIP for awhile and haven't really found anyone who has seriously entertained the notion. Mainly I'm interested in further developing the xWHIP stat and I wanted to know if you had seen the work I had done before or if this was, as you mentioned concurrent"