Brief Sabermetric Primer: GPA
by redsoxstatscom on Jan 26, 2010 9:54 AM EST
Email
PrintOne of the most fascinating and influential parts of the book Moneyball was the chapter on run production, on-base and slugging percentages. In 1999 Billy Beane hired Paul DePodesta and DePodesta went to work figuring out a better method of combining OBP and SLG than just adding them together to form OPS.
DePodesta concluded that for the Oakland A’s purposes, "an additional point of OBP is worth 3 additional points of SLG." The same year Moneyball was published Sabermetrician Tangotiger penned a series of articles titled OPS: Begone! II in which he concluded that the best-fit multiplier falls somewhere between 1.5 and 2.0, with 1.7 or 1.8 now the accepted norm.
Based on OPS a player with a .325 OBP and .475 SLG is considered to be the same offensive force as a player with a .400 OBP and .400 SLG. This just is not the case.
Aaron Gleeman read the research on this topic and figured out that if you use 1.8 as the multiplier and divide by 4 you get a scale that is very close to batting average. As in, .330 is a superstar, .275 is OK, and .220 is awful.
Gross Production Average was born: ((OBP*1.8) + SLG) / 4
In 2009 the American League had a .258 cumulative GPA while the National League finished at .251.
Albert Pujols lead the majors last year with a .364 mark, followed by Joe Mauer at .347. Yuniesky Betancourt was the worst of qualified hitters at .211.
Kevin Youkilis 5th, Jason Bay 18th, and JD Drew 20th all finished in the Top 20. In his time with the Red Sox, Victor Martinez had a .309 GPA, 2-points better than Bay and Drew.
In the sabermetric community Weighted On Base Average or wOBA is now a popular stat. It is a long formula that uses some stats that are not readily available. Last year it correlated .992 with GPA. In my original sabermetric primer I talked about how you can advance your sabermetric knowledge and still keep it simple. GPA is one of those cases.
34 comments
|
0 recs |
Do you like this story?
Comments
If I'm reading this right
Eight members of the Sox lineup had above-average (albeit only slightly for Ortiz and Ellsbury) GPA’s in 2009, of which at least six will be full-time members of the team in 2010.
Where do the new members of the Sox fall on the GPA spectrum?
Scutaro was at .273 last year, a big jump from what he was as a role player, something Epstein said was hindering him as a player. Beltre was .232 last year, but in the three previous years he was between .261 and .264. Cameron was .267 and .268 over the last two seasons.
by redsoxstatscom on Jan 26, 2010 11:38 AM EST up reply actions
Probably should note that the cumulative GPA of the AL is lower than what we'd expect from an average regular player
since it includes role players, part time guys, replacement players at the end of the year, etc…
And Little Jed Lowrie is at the bottom of the scale
It’s a good thing BS.UF is on hiatus. ;)
Galactus does as he pleases. Because Galactus is drunk.
I have a five-tool player in my pants.
I don't think even BS.UF can say Jed was good last year.
He (and I) just think he’s gonna be fine in the long-term.
USG
by Ben Buchanan on Jan 26, 2010 10:59 AM EST up reply actions
You're probably right.
I’m just amused by his crush.
Galactus does as he pleases. Because Galactus is drunk.
I have a five-tool player in my pants.
I may be on hiatus, but I'm still ninja-lurking.
@bs_uf15bosox9be 12 pieces of bacon, a Red Bull, and go get 'em; Learn to use SB Nation
While I do like GPA a lot more than OPS, it is still not park-adjusted
so like OPS, you need to keep that in mind when using it. On the other hand, because (I believe) OBP is less sensitive to park factors than SLG, GPA should be less sensitive than OPS.
Its best feature is that it’s so easy to calculate on the fly, compared to wOBA.
wOBA
I agree with your assessment of the numbers. Additionally, I’d say GPA’s best feature isn’t that valuable.
It is easier to calculate GPA if all you have access to is the slash lines or the raw counting totals, but frankly I’d still need a calculator to multiply and OBP by 1.8 add it to SLG and then divide by four. But when I have access to a calculator, I almost always have access to the internet, where it’s just as easy to look up either. And because it’s just as easy to get either one of them, I’d recommend using the more accurate wOBA.
It’s true that the step from OPS to GPA is greater than GPA to wOBA, but that’s not a very good argument for using GPA. Judging offense by OPS/GPA/wOBA is like eating cereal with a Fork/Spork/Spoon. Sure you get more from switching from Fork to Spork than you do from switching from Spork to Spoon, but when it’s just as easy to grab any of them on the internet it’s silly use the spork.
Yup.
I never look at GPA. wOBA is better and just as (if not more so) available. Its not like any of us ever actually calculate these stats ourselves.
"Ninety percent [of my salary] I'll spend on good times, women, and Irish whiskey. The other ten percent I'll probably waste."
-Tug McGraw
While I do like GPA a lot more than OPS, it is still not batting order adjusted.
who really cares about slugging from your lead off batter – just get on-base. (steal 2nd with no outs, is like hitting a double).
who really cares about on-base from your clean up batter – just slugging and/or the abilty to move the runners.
i never liked seeing manny or papi or whoever, taking a walk with a runner on 2nd with less than 2 outs – leads into too many double plays.
(there probably is a stat for all of this, however, i’m not a rocket scientist like most of you guys)
Not quite accurate (at all)
1. A double is better than walking. It will score runners on base (leadoff hitters do not always lead-off innings). In every case, a double would be better than “just getting on base” as would a triple or a home run (no matter who hits it).
2. The most valuable skill from a batter is simply not getting an out. This is why OBP is valuable.
3. I think what you are actually talking about is not a stat that is “batting order adjusted” but simply one that is context specific. wOBA, OBP, GPA and AVG or all context neutral. They do not adjust for how the play actually effects the game. WPA (win percentage added) shows how each individual event in a game changes the team’s chance of winning. So if a team is down one and Papi draws a walk, it reflects exactly how that changes the game’s outcome.
"Ninety percent [of my salary] I'll spend on good times, women, and Irish whiskey. The other ten percent I'll probably waste."
-Tug McGraw
Leading off the inning
Some Sox 2009 stats (PA = plate appearances, LOI = leading off inning, LOG = leading of game)
Ellsbury 691 PA, 244 LOI, 117 LOG, LOI-LOG = 127, LOI% = 18%
Pedroia 714 PA, 148 LOI, 25 LOG, LOI-LOG = 103, LOI% = 14.4%
Drew 539 PA, 136 LOI, 14 LOG, LOI-LOG = 122, LOI% = 22.6%
Youkilis 588 PA, 114 LOI, 0 LOG, LOI-LOG = 114, LOI% = 19%
This shows that Youkilis and Drew were just as likely to leadoff an inning than Ellsbury when not including the game leadoff plate appearance. So the difference between a leadoff hitter and any else in the lineup is about 100 plate appearances or so
and...
for us teachers in the community if you multiply it by 10 you get a pretty decent scale of actual un-weighted GPA (since the education community has established its GPA well before the baseball stats community). For example – Pujols would have a 3.64 GPA while Casey Kotchman would have a 2.00 and be on academic probation..
But doesn't Pujols deserve a 4.0?
"You know you're having a bad day when the fifth inning rolls around and they drag the warning track." - Mike Flanagan, Baltimore Orioles pitcher, 1992.
No.
He needs to pitch a perfect game and turn an unassisted triple play. Accepting anything less would just be the soft bigotry of low expectations.
"It's just a tiny little nick, but it hurts when I get champagne in there."
- Jason Bay, on getting spiked scoring the winning run in ALDS Game Four.
by 0157H7 on Jan 26, 2010 1:24 PM EST up reply actions 1 recs
What is the basis for 1.8?
And similarly, what is the basis for DePodesta’s statement that in 1999 “an additional point of OBP is worth 3 additional points of SLG,” and was that particular to the make up of that particular team at that specific time, or is it a truism.
I’ve often heard that OPS is the more valuable component of OPS (versus slugging), but I’ve never actually seen the backup data for it.
Manny ain't the only bad man.
Bay is 18th in the AL or MLB?
at any rate, his GPA is higher than Drew, hits for more power than Drew, is MUCH more durable than Drew, but SOMEHOW is not worth as much money as Drew. Do I have that right? Oh wait, am I leaving out the almightly UZR? Sorry, my GPA was only a bit over 3.0.
ignoring UZR
you are completely disregaring defense. Would you claim it is not a part of being a baseball player? Let’s get a team of Adam Dunns! Also, he is making more money than Drew.
His GPA is higher than Bay in a season where Drew is 3 years older.
His GPA in his 2nd to last year will not be so kind.
And, yeah, you’re kind of ignoring, y’know, an entire aspect of the game. Like, 1 of the big 3. There’s offense, pitching, and defense.
It’s that one at the end there.
USG
by Ben Buchanan on Jan 27, 2010 10:33 AM EST up reply actions
Not even close. Drew is more valuable than Bay.
Ignoring projections, just looking at what they did last couple of years:
Offensively, it is a lot closer than you think.
wOBA/wRC+
Player 2008 2009
Bay .387/138 .397/142
Drew .400/145 .389/137
So I don’t know about you, but I’m simply NOT seeing any great offensive edge there for Bay…
Defensively, I’m sorry but I have to include UZR because it is NOT close.
UZR
Player 2008 2009
Bay -18.4 -13.0
Drew 6.4 10.5
Note – those are UZR, not UZR150, so those represent the actual defensive values provided. the UZR150’s are even more in Drew’s favor.
Thus when you add up the WAR values for both years, Drew blows Bay away (8.9 to 6.4). Drew was paid more because he was WORTH more.
As to whether Bay is more durable than Drew? Bay has played 306 games the last two years to only 246 for Drew. But it was much closer this year (151 to 137). And if we include the prior two years:
Bay Drew
2006 159 146
2007 145 140
2008 155 109
2009 151 137
So take away 2008’s injury year and Drew is fairly consistently getting on the field (He’s actually among the top 3 or 4 for the Red Sox in getting on the field since he’s been with them).
And – this is important – in those fewer games, Drew has actually OUT PRODUCED Bay.
LOL - lucky for you ...
… he’s played in 532/648 = 82% of Red Sox games the last 4 years! So you’ve had plenty of chances to watch him play!
Only Youkilis(573), Ortiz (559) and Lowell(barely at 539) have played in more Red Sox games over the last 4 years.
Considering that Ortiz is a DH and Youk has primarily been at 1B while Drew is an outfielder …
Ugh - now I feel stupid - Drew was still in LA in 2006
[ laughs at self ]
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
In Association With
Big League Stew - MLB - Yahoo! Sports
- Five tips for Showtime’s second season of ‘The Franchise,’ featuring the Miami Marlins
- The 10 best things about being a Detroit Tigers fan
- Surprise! A’s finally end the Yoenis Cespedes derby with $36 million deal for Cuban star
- The time Whitney Houston sang to Jim Bowden
- Pitchers and catchers! Mariners get early start in Arizona
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
