WAR and BAB

over 2 years ago Bobby Baseball 12 comments 0 recs |

Comments

I meant to title this WAR and BABIP. And for some reason the description got canceled out. But anyway I was going to ask, can someone please explain the WAR and BABIP stats to me? I see them a lot on this site but I do not understand them. Thank you.

by Bobby Baseball on Nov 5, 2009 5:06 PM EST reply actions

BABIP

stands for “batting average on balls in play”. It’s a player’s batting average if you disregard strikeouts and home runs. I think you would calculate as (H – HR) / (AB – K – HR). (Come to think of it, the calculation might also adjust for sacrifice bunts and flies, I’m not completely sure.) It’s a way of evaluating how a player has done on balls that are handled by the opposing defense (as opposed to strikeouts and home runs, which are, other than the rare case of inside-the-park home runs, basically only a function of the pitcher and the hitter).

Generally speaking, a players BABIP can be expected to be something along the lines of his line drive rate + 0.120, so if player X hits line drives 25% of the time (i.e., a line drive rate of 0.250), then you would expect his BABIP to be .370, and if it were significantly higher or lower then the player might be said to be lucky or unlucky with regard to opposing defenses.

I’ll let someone else talk about WAR, since it’s significantly more complicated and can be determined a number of different ways.

by JoshNY on Nov 5, 2009 5:25 PM EST reply actions

Beat me to it ;)

I usually calculate BABIP by doing (H-HR) / (PA – SO – BB – HBP – HR), that way you don’t have to futz around with bunts and sac flies. Especially for pitchers, BABIP used as a way of determining the “luck” factor, that is, how elements a pitcher couldn’t control hurt his perceived performance. Most pitchers should be expected to see their BABIPs regress to near .300, no matter where they are. It can be used as a way of predicting of struggling or surging players might soon rebound or come back down to earth.

WAR is meant to be an estimated win value for a particular player over a hypothetical “replacement level”. The best non-technical way I’ve heard replacement level described is that a replacement level player is the quality of player where teams will no longer feel the need to compete for his services, a replacement level player can simply be picked up by any team at just about any time for minimal cost. So, for example, in 2009, David Wright was worth 3.4 WAR (according to Fangraphs, scroll all the way down for win values section). That means, we can assume that if David Wright had to be replaced with a player other teams weren’t competing for, that player would have been worth 0 WAR, and David Wright would have been worth 3.4 wins more than that player over the 144 games he played this season. A 25-man roster composed of Replacement Level players would be expected to win about 48 games in a 162 game season. A player whose 1-2 WAR can be considered about “average,” 3-4 WAR “above average” and 5+ WAR is a star.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Nov 5, 2009 5:55 PM EST up reply actions

and 8+

is Pujolsian

"[The Giants] beat us down. We were beat by a grown-man team, a team we want to be like one day. They came in here and took it to us. Out-manned us, out-gunned us. ... It wasn't even close." - Raheem Morris, 9/27/09

by cjmulrain on Nov 5, 2009 8:36 PM EST up reply actions

Thanks for explaining things. BAB is Base Attack Bonus, in my mind.

That website is handy. I’m just starting to learn all this advanced statistical analysis stuff…

"Blinding ignorance does mislead us. O! Wretched mortals, open your eyes!"

by Brooklyn Dodgers Mets Fan on Nov 5, 2009 8:27 PM EST reply actions

It's funny

Someone asked me to define WAR and it’s hard to do succinctly. A short definition seems to contain a set of terms that, in turn, require explanation, such as, “replacement level” or “win value.” It’s not a simple equation of hitting stats.

Try searching for a brief definition of WAR on Google, it just doesn’t exist!

by TheBigStapler on Nov 6, 2009 10:08 AM EST reply actions

WAR assesses defense as well as offense

which is one of its many strengths in my opinion. It’s the closest thing statisticians have to a Unified Theory of Everything for baseball.

by TheBigStapler on Nov 6, 2009 10:13 AM EST reply actions

I still have a hard time fully trusting FIP though

I understand balls in play are a tangled web which is hard to unravel, but because they’re completely left out, I still think of FIP as a rough first approximation.

by DoghouseBlues on Nov 6, 2009 10:42 AM EST reply actions

FIP isn't perfect

but it’s better than ERA. Home run rates are really variable for pitchers from year-to-year so xFIP is better, or tRA which factors batted balls into its calculation. All 3 are > ERA.

by James Kannengieser on Nov 6, 2009 6:59 PM EST up reply actions

OK thanks everyone

Just one more question: How do you calculate WAR?

by Bobby Baseball on Nov 6, 2009 5:48 PM EST reply actions

If you have the time/inclination

Check out Fangraphs’ glossary page – scroll down to the bottom and you’ll see 2 7-part series about WAR (like I said, if you have the time), 1 for position players, 1 for pitchers. The concepts are not easy to grasp but it’s worth the effort, this stuff is really interesting. Hope this helps.

When asked to give a simple explanation of WAR, which is really hard to do at TBS noted above, I say it’s a stat that factors offense, defense, position and playing time to place a value on a player. I don’t bother trying with pitchers because defense independent pitching stats require too long an explanation.

by James Kannengieser on Nov 6, 2009 6:58 PM EST up reply actions

Or

let Fire Joe Morganexplan. VORP is included cuz its similar and explains "replacement player:

VORP (Value Over Replacement Player)
An offensive stat only, VORP attempts to calculate the number of runs a player is contributing above what a replacement-level player at the same position would if given the same percentage of team plate appearances. VORP is a counting stat, not a percentage stat – so, for example, as of July 22, Andruw Jones has a VORP of 31.0. That means that he has created 31 more runs for his team than the average AAA call-up guy would have by this point in the season. It also turns out that every ten runs a player creates is worth roughly one win, so Andruw’s offense alone has earned the Braves three wins. (There are other stats, like Fielding Runs Above Average [FRAA] that do the same thing as VORP, for defense.) See WARP below for more.

Old Baseball Men, this is another good one to bandy about if you’re interested in tearing down a nerd’s argument. Because it sounds funny. VORP. Please. What’s that doing in baseball? Forget VORP, let’s come up with a stat for the size of a guy’s heart, am I right, people? We’ll call it the Eckstein Quotient. No, wait, that sounds too nerdy. Eckstein Number. Nope. Still too smart. Eckstein Thing. How about just Thing? The highest Thing in the majors? You guessed it: David Eckstein. That’s why they almost named it after him.

Once again, Albert Pujols was your VORP leader in 2008 at 98.7. …

WARP (Wins Above Replacement Player)
Sort of like VORP, but with a defensive component, as well. And it’s calculated in terms of wins. It uses VORP and FRAA and all of those things to figure out how many wins a player is worth to his team, by himself, from all phases of his game. There are also WARP-2 and WARP-3, which adjust for various historical factors and stuff like that.

by Jsz on Nov 7, 2009 10:31 AM EST up reply actions