Creating my own BABIP-like stat customized to league

[locutusut]

I've been studying BABIP, or batting average on balls in play calculated as (H - HR)/(AB - HR - K + SF), and I was thinking it would be quite possible to create your own metric based on the types of categories in your fantasy league.

For instance, we're going to play head-to-head one win category (e.g., you win as many categories as possible to win the week). Our offensive categories will be:

Runs (R), Hits (H), Doubles (2B), Triples (3B), Home Runs (HR), Runs Batted In (RBI), Stolen Bases (SB), Walks (BB), Strikeouts (K), Batting Average (AVG)

So based on that, I came up with my own metric to consider the categories we'll use in our league as follows:

(R+H+2B+3B+HR+RBI+SB)/(AB-BB+K)

Using this formula, I get the following as the top 25:

Hitter Score BABIP
============================
Albert Pujols 0.809 0.297
Miguel Cabrera 0.796 0.336
Joey Votto 0.741 0.361
Justin Morneau 0.737 0.385
Kevin Youkilis 0.736 0.327
Josh Hamilton 0.726 0.390
Jose Bautista 0.718 0.233
Troy Tulowitzki 0.715 0.327
Joe Mauer 0.696 0.348
Carlos Gonzalez 0.688 0.384
Dustin Pedroia 0.682 0.291
Carl Crawford 0.681 0.342
Nelson Cruz 0.676 0.348
Coco Crisp 0.676 0.307
Robinson Cano 0.671 0.326
Matt Holliday 0.670 0.331
Mags Ordonez 0.668 0.313
Jim Thome 0.666 0.310
Brett Gardner 0.664 0.340
Aubrey Huff 0.660 0.303
Paul Konerko 0.658 0.326
Hanley Ramirez 0.657 0.327
Rafael Furcal 0.651 0.338
Chase Utley 0.648 0.288
Jayson Werth 0.647 0.352
Evan Longoria 0.647 0.336

I'm just wondering what other people think of this approach and feel free to poke flaws in my formula if you think there's a better way to calculate based on these stats.

Thanks!

[A Fleshner Fantasy]

A couple things:

1) BABIP is a stat generally used to measure how lucky a player is, while this seems to try to quantify how valuable a payer is to have on your team, so I don't really see how this is like BABIP. However, it can still be useful, ust not for the same purpose of BABIP.

2) Taking my previous point into account, it's a good start, but you need to take into account the scarcity of some of the statistics. For example, there are far fewer triples hit than doubles. Therefore, you shouldn't have a double be worth the same amount as a triple, which you do in your formula. Perhaps try to add some kind of coefficients to add a weight to certain statistics.

[locutusut]

True...this is more just in the spirit of BABIP, as that's what gave me the idea to create my own statistical indicator.

Yeah, I thought of that too, as you said, where things like Triples would not really represent that much, but in our categories it won't really matter because we're not using points, it's just you either win or lose the category and thus all categories are equal. I'll have to see if I could normalize it somehow.

That said, the same familiar names you see on any draft list seem to roll to the top with a couple of surprises.

[Neato Torpedo]

True...this is more just in the spirit of BABIP, as that's what gave me the idea to create my own statistical indicator.

Yeah, I thought of that too, as you said, where things like Triples would not really represent that much, but in our categories it won't really matter because we're not using points, it's just you either win or lose the category and thus all categories are equal. I'll have to see if I could normalize it somehow.

That said, the same familiar names you see on any draft list seem to roll to the top with a couple of surprises.

Well, categories aren't equal even in roto-style H2H leagues. 15 HR win it for you in a week probably 85% of the time, but 15 hits will almost never beat your opponent. Therefore, 15 HR are worth more than 15 hits.

Also, the variation in triples year to year make it so there's only like 5 players you can reliably draft for triples: Reyes, Crawford, Granderson, Victorino, Span, and Drew. Everyone else that's high up on the triples leaderboard had like 14 one year and like 5 the next. With numbers that are always that low, there's a lot of variation and low predictability, so when drafting and making projections, you shouldn't take triples into account that much except for those top 5 players.

Also, if you go to http://baseballmonster.com, go to "League Settings" up at the top, and change the settings to match your league, it will spit out an accurately weighed rankings list for your league.

[Fenris-77]

Also, if you go to http://baseballmonster.com, go to "League Settings" up at the top, and change the settings to match your league, it will spit out an accurately weighed rankings list for your league.

You're such a killjoy. He was ready to boldly go where no saber-fantasy-freak has gone before and you have to up and pee in his cornflakes with this "there's already a great tool for that" shizz. Yeesh.