A reference post inspired by Jerry's comment a little while back. As a word of advice, percentages should be taken as being +/- 0.5% due to some weird issues with the Hardball Times data.
Hitters:
BA: .270
OBP: .337
SLG: .432
OPS: .769
uBB/K: 0.46
uBB%: 7.7%
K%: 17.0%
XTRA%: 34.5%
RC/G: 5.15
P/PA: 3.76
LD%: 19.6%
BABIP: .305
GB%: 43.6%
HR/FB%: 12.1%
IF/FB%: 10.0%
BA/RISP: .274
Pitchers:
ERA: 4.53
FIP: 4.52
xFIP: 4.70
K%: 17.0%
uBB%: 7.7%
HBP%: 1.0%
HR%: 2.9%
K/uBB: 2.19
K/9: 6.58
uBB/9: 3.00
HR/9: 1.13
P/PA: 3.76
K/G: 6.56
uBB/G: 2.98
HR/G: 1.10
DER: .695
LD%: 19.6%
GB%: 43.6%
FB%: 36.8%
GB/FB: 1.18
IF/FB%: 10.0%
HR/FB%: 12.1%
LOB%: 71.6%
Pretty much any individual stat you're ever going to need can be found at The Hardball Times or Fangraphs.
Might as well make this an open thread for any of you with questions about more advanced metrics.
0 recs | 37 comments
Thanks Jeff
League average is a phrase that gets tossed around a lot, and its nice to have a concrete idea of what that means.The pitching stats must be a bit skewed by including relievers (not just FIP, xFIP etc, but K/9 and BB/9), no?
Also, what is the +/-? is it the range for the 25-75 pecentile or something?
Bearskin Rugburn - December 11, 2006
Yeah, they are a little skewed.
But trying to separate all the relievers from all the starters on the THT stats page is more work than it's worth to me. I don't think the difference would really be that great in any one category, except maybe HR/FB%, where relievers should be ~2% better than starters or so.And the plus-minus represents one standard deviation. In other words, assuming a normal distribution, 68% of pitchers should fall within one standard deviation of the average.
Jeff Sullivan - December 11, 2006
Thanks
That is really useful.Thanks, Jeff.
Jerry - December 11, 2006
As an M's fan
it always startles me to see that the league average OBP is around .340. I guess I'm just too used to our free-swinging ways.Slozbury Stouvre - December 11, 2006
How is it
There is a different rate for BB%, K%, IF/FB and HR/FB betweeen pitchers and hitters?i don't get it?
mariners124m - December 11, 2006
It would even out
if I included all the hitters with < 150 PA's and all the pitchers with < 50 innings.Jeff Sullivan - December 11, 2006
2006 only?
Are these the THT numbers from 2006 only?Trev - December 11, 2006
Yep.
I'm guessing there's little year-to-year deviation, though.Jeff Sullivan - December 11, 2006
The Most Average Player
This is a non-scientific study (I just eye-balled it, but you could use some statistical method to compute sim scores), but here are my picks for most average player based on batting/pitching performance in 2006 (and their salaries too for fun):Batters:
Garret Anderson ($10 million)
Juan Encarnacion ($3.5 million)
A.J. Pierzynski ($4 million)
Brandon Phillips ($350,000)
Pitchers:
Dontrelle Willis ($4.35 million)
Josh Hancock ($355,000)
Victor Santos ($500,000...and selected in the Rule 5 Draft)
Justin Lehr ($337,000)
Trev - December 11, 2006
Yay
we have one of the most average pitchers in baseball.Mariner John - December 11, 2006
The Most Statistically Average Pitcher
This time I actually ran the numbers (I used SQRT(Z-Score[K/G]^2+Z-Score[BB/G]^2+Z-Score[GB%]^2 to compute this), and applied them to all 2006 pitchers in the THT database. Here are the results:Most Average (min 50IP):
- Chad Paronto, ATL
- Vincente Padilla, TEX
- Ryan Madson, PHI
- Josh Beckett, BOS
- Victor Santos, MIL
- Justin Verlander, DET
- Byung-Hyun Kim, COL
- Brandon Lyon, ARI
- Jae Seo, LAD
- Dave Borkowski, HOU
*Most average Mariner, any IP: Cha Seung Baek (he'd be between Lyon and Seo)Least average pitchers (min 50 IP):
Trev - December 11, 2006
I'm surprised
More supremely above average pitchers than below average.Mariner John - December 11, 2006
I was seriously about to do this myself.
You rock.$34m/3yr buys average these days, eh?
Jeff Sullivan - December 11, 2006
The sad part is...
...between park and league adjustments, and assuming relative sobriety, Padilla's probably a better contract than Batista.Trev - December 11, 2006
Most Statistically Average Batters
Because of the way THT displays their data, and the stats I chose to compute "averageness", I went with the fangraphs data instead. Thus, these are statistically average QUALIFIED players.I had a harder time picking out stats to work with compared to pitchers. In the end I went with:
AVG, OBP, SLG, BB%, K%, and BABIP.
Most Average Batters:
- Shawn Green - - -
- Tony Graffanino - - -
- Melvin Mora Orioles
- Hank Blalock Rangers
- Adam Kennedy Angels
- Garret Anderson Angels
- Ramon Hernandez Orioles
- Brandon Phillips Reds
- Nicholas Markakis Orioles
- David Bell - - -
- Adrian Beltre Mariners
Least Average Batters:- Ryan Howard Phillies
- Travis Hafner Indians
- Manny Ramirez Red Sox
- Albert Pujols Cardinals
- Jim Thome White Sox
- Jason Giambi Yankees
- Bobby Abreu - - -
- Adam Dunn Reds
- David Ortiz Red Sox
- Miguel Cabrera Marlins
This one is obviously going to be biased towards good players because players on the extremely good side of non-average would accumulate enough at-bats to qualify. Actually never mind (Clint Barmes was the 17th least average guy, Angel Berrora was the 21st.)Trev - December 11, 2006
I am so indebted to you.
You just saved me like an hour of my night.It's not particularly reassuring to see Adrian Beltre and David Bell occupy consecutive positions on the same list.
Jeff Sullivan - December 11, 2006
You're welcome
You provided the standard deviations, after that it wasn't too hard.I've come around quite a bit on Beltre. If he was a FA, and the M's signed him to a 3/$39M deal this offseason, I probably wouldn't complain too badly.
Then again, I intentionally don't watch many of his at-bats.
Trev - December 11, 2006
cool
thats way awsome.ZeZetheX - December 11, 2006
Most Average or Most Rounded?
I think to call these lot below Most Average is misleading - if they host many of the qualities of a pitcher at average level they're doing pretty damn well. It's the ones who're good in some areas and weak in others who'll get exploited, and the younger a pitcher that fits in that top 10 the greater his potential.Bodes well for Baek, I'd say.
MarkE - December 12, 2006
Well, I agree with your primary points
But not the thing about Baek. Stats lie in small sample sizes.Graham MacAree - December 12, 2006
You know who is remarkably average?
Based on fangraphs stats, Ervin Santana whom everyone has been raving about is almost exactly average in all stats except BABIP, which is out of pitcher control. Its a remarkable correlation.So there. League average gets you 15 wins. Sometimes. Go Baek!
Bearskin Rugburn - December 12, 2006
How is it that
The average ERA+ is 115... do pitchers with less than 50 innings really skew the stats that much?patsfan - December 12, 2006
When I include everybody
the average ERA+ comes out to 105 for some reason. For pitchers with < 50 IP, it's 97.Jeff Sullivan - December 12, 2006
Hey jeff
can you make this a quick link. Like next to the community projections or something for reference so in the future I don't have to go diggin for this page.Scruffy Lefty - December 12, 2006
Already did.
It's on the left, under "References".Jeff Sullivan - December 12, 2006
On the subject of blogkeeping
The Hit Tracker Online link doesn't link to anywhere at the moment.Alex B - December 12, 2006
Whoops.
forgot the 'r' in 'a href'. Thanks for the heads-up.Jeff Sullivan - December 12, 2006
sweet thx
Scruffy Lefty - December 12, 2006
A team full of players
that put up "league average stats" according to your numbers would win significantly more than 81 games. Leaving out the players with fewer than 150 ABs or 50 IP, really skews things.According to my homemade strato-matic baseball simulator 2000, I've just run 10,000 seasons with a fictional team added to the league filled with "statistically average players" (meaning that injured players were replaced with other statistically average players - thus negating the deflating effects of injuries,) and that team averaged 93 wins per year. This allowed the team to make the playoffs 54% of the time, and win the world series 12% of the time.
OK, I just made all that up.
johnbai - December 12, 2006
Someone who really does have...
a homemade strato-matic baseball simulator should try this though, and see what happens. I'm really curious.johnbai - December 12, 2006
Well
perhaps that means that players who last a full season should be held to a higher standard than the "normal" league average that takes everybody into account.Jeff Sullivan - December 12, 2006
DING DING DING WE HAVE A WINNER....
This is something I've been saying for aeons. 180-200 IP of 4.4-ish ERA ball is actually above average.eponymous coward - December 12, 2006
That's not really what I meant.
I mean, you're right, but...well, let's use an example.Overall league average: 4.53 ERA
Min. 50 IP league average: 4.33 ERA
A guy who throws 180 innings of 4.50 ERA ball would look average by the first and below-average by the second. And I think the latter is more meaningful, because what's the good in comparing (say) Horacio Ramirez against a player pool that includes a bunch of rookies and no-names who don't last a month?
Jeff Sullivan - December 12, 2006
Well, because...
Talent distribution as expressed in terms of player performance in MLB =! normal distribution (aka bell curve).It's really a J-shaped curve, with 99% of humanity on the end of the J meaning "you suck at this", .9% at various places in the minors and at replacement level, and .1% being really good ballplayers. (OK, not that, it's more like 99.99999, .0000009, and .0000001, but you get the idea.)
Excluding "replacement level" talent out of your sample of MLB performance levels (which is what you do by tossing out <50 IP and < 150 PA's) is a HUGE bias in evaluating performance.
eponymous coward - December 12, 2006
I'm going to be perfectly honest...
this is retroactive justification. I limited the sample to > 150 PA's and > 50 innings because including everyone led to screwy standard deviations (because for certain stats there was an incredible spread at lower sample sizes).I think I'll probably re-calculate all these numbers tonight, including everyone.
Jeff Sullivan - December 12, 2006
All set.
Everything's re-calculated.Jeff Sullivan - December 12, 2006
Meaningful comparisons
But Jeff,Given that Bavasi has to make choices between players like HoRam and your "bunch of rookies and no-names who don't last a month," I'd say it's meaningful. But I'm starting to feel like a big Bavasi apologist weenie, so I'm just going to stop. ;)
johnbai - December 12, 2006
You must Login with your SB Nation account and be a member of Lookout Landing to post a comment.