The Scouting Report, By the Fans, For the Fans - Most Similar Fielders (March 18, 2004)
Another preliminary report. It's a definite time killer.
--posted by TangoTiger at 03:33 PM EDT
Posted 3:41 p.m.,
March 18, 2004
(#1) -
dr feelgood
Sorry to pick nits, but you list only one Alex Gonzalez. When you have time, you might want to tell us which one. (or is it just assumed that if you don't say "Alex S Gonzalez", that it refers to Florida Alex Gonzalez?) Anyway, keep up the good work.
Posted 4:05 p.m.,
March 18, 2004
(#2) -
J Cross
Great Stuff!!
Some random comments:
Craig Biggio top 3 most similar players have moved between a few positions (Cat, McEwing and E. Wilson).
Barry Bonds most similar player is also his country's greatest slugger.
Mike Cameron's are two exc. defensive CF's (Hunter and Shinjo).
Jermaine Dye and Vance Wilson??? (this is one I don't see)
Darin Erstad and Pokey Reese - making up for bad offense group
Huff's comps: Cuddyer, Bellhorn and Scutaro - "we didn't get a chance to play when we deserved one" group.
The fact that Scutaro is McEwing best fielding comp adds insult to injury for Mets fans.
Rey Ordonez and his comps would be a dominant Hacking Mass team.
Mike Piazza the next "pickin' machine?"
I think it's interesting that players tend to comp well with others from the samp position. It looks like this is especially true for the best defensive players at each position.
The gambler in me see the following next step:
1) Regress MGL's UZRs to make 2004 UZR predictions for each player.
2) Let "Scouting Report" pick the over or under and see how it does.
3) If it does well, make 2005 predictions based on regressing UZR's to the scouting report.
Something like that?
Posted 4:50 p.m.,
March 18, 2004
(#3) -
Hatrack Hines
Royce Clayton and Garret Anderson . . . Seperated at birth?
Posted 5:40 p.m.,
March 18, 2004
(#4) -
tangotiger
On a scale of 0 to 100, with 50 as average, here are Garrett and Royce's numbers, in order of the listed traits:
Garr: 63,58,58,64,69,48,66
Royc: 60,56,54,69,61,56,66
Do you think that too many fans evaluated the players with respect to their positions?
Posted 6:21 p.m.,
March 18, 2004
(#5) -
Brian P
I was suprised to see Hidalgo's only comparison to be Kevin Cash. Sort of a strange combination of positions. I guess I would have imagined seeing him more along the Jose Cruz, JR., Jose Guillen types....
Posted 6:31 p.m.,
March 18, 2004
(#6) -
Hatrack Hines
Are Garrett and Royce the most similar pair?
As someone who has a bit of a fascination with how height and weight impact performance, it's interesting to me to note that Garrett and Royce are listed at roughly the same weight: 190 for Garr and 183 for Royc, while Garr is 6'3" and Royce is 6'0". In fact, they've got basically the same BMI (weight in lbs. / height in inches) as listed.
As listed, Bobby Abreu and Derek Jeter are pretty close in BMI too.
Probably, Royce is more like 5'10", and Garrett probably weighs more like 210, and there's no way in Tampa Abreu weighs 160, but it's interesting, nonetheless.
Posted 10:29 p.m.,
March 18, 2004
(#7) -
tangotiger
Jose Cruz Jr is one of the most complete fielders, according to the fans. Another guy the fans love is Cesar Izturis.
Posted 12:25 p.m.,
March 19, 2004
(#8) -
Nick S
Great fun to skim through, Tango, thanks for putting this together.
We now know definitively that Ramon Martinez is the nexus of the fielding universe.
Posted 2:39 p.m.,
March 19, 2004
(#9) -
Most Similar Fielders Bobby Abreu and Derek Jeter
I guess the formula doesn't subtract any points for not standing anywhere near each other on the baseball field?
Posted 3:00 p.m.,
March 19, 2004
(#10) -
tangotiger
That is absolutely correct, and intentional.
I have synthesized the fielding aspects of players into 7 traits. The sim scores are based solely on those traits.
You can argue that 7 is not enough, and that those 7 are not representative enough. However, I'm fairly pleased at how seriously the fans have taken this project, and how many of the comps are good or at least interesting. Mostly when catchers are involved would you think that the sims are not good.
Out of all 8 positions, I think the catcher is the one that does things that are so different from the rest that it might be a good idea to exclude him. You can put Manny Ramirez at 1B, or Nick Johnson in LF, or Todd Helton at SS, and you can already kind of figure how they might do there. Put Ivan Rodriguez at 2B? Ausmus in RF? Their skillsets are not exposed enough at C that you can try to make an informed opinion.
Posted 10:11 a.m.,
March 22, 2004
(#11) -
mommy
tango, i think it's almost impossible to completely divorce our evaluations of players from their positions. for instance, i think it can be hard to translate how an infield arm compares to an OF arm. i know furcal has a rifle, but i don't know how that rifle would look in the OF. i know it seems superior to other arms i see making throws from SS, but it's hard to accurately say how a furcal throw from RF to 3B would look. i can assume it would look like a great arm anywhere on the field, but for me it's hard to know for sure how to compare it to guerrerro's arm.
i think the same could hold true to some extent for other traits as well, except perhaps for Hands.
Posted 10:33 a.m.,
March 22, 2004
(#12) -
tangotiger
mommy, you may think it's "almost impossible", but it shouldn't be (especially if you are an experienced scout). I think a scout should be able to see a player make 100 plays at one position (say like the throws from SS to 1B, or relays from short OF to home), and be able to give it one number, like 95 out of 100.
Change "almost impossible" to "difficult if you're not looking for it", then I will agree with your statement.
The one thing that I noticed in the player evaluations is that catchers are of such different breed, that I should remove them altogether. For the other 7 fielders, we are looking at how they react on batted balls in play, and trying to get the runners out from that BIP. For catchers, they are starting from such a different positions on BIP, that it's really almost impossible to compare the first 2 categories.
There is a quality to catchers (experience) that is far more important than all the other categories I listed. When I compared the average experienced catcher to the average no experienced catcher in things like PB, WP, etc, there is such a huge difference, that I have not captured it in this study.
Since the study is focused on BIP, I don't think I can come to good conclusions with catchers.
Posted 2:27 p.m.,
March 24, 2004
(#13) -
tangotiger
This is something I've been doing 5 different ways: how to compare fielders at different positions. You can check out the various articles I have on this in the Primate Studies index.
The idea goes: every player is compared against the average at that position. Every position's average will end up being 0. So, we need some way to adjust the avg SS upwards and the avg 1B downwards to properly balance them. The traditional way has been to look at their offense, and take the opposite. So, if the avg SS is -10 hitting runs, then they must be +10 fielding (relative to all fielders), so that overall they are zero. The assumption here is that all positions, off+def, are equals.
As noted, I have many different ways to get to that answer.
Here's another. I took Mark Ellis, and looked at the Fans' Scouting Report. I looked for the SS that is most similar to him. I repeated this for all 2B. What I end up with is the average 2B profile, and the average profile of their most similar SS. In my group of 2B, their UZR was +3.5 runs. Their SS equivalents had a UZR of -0.7 runs.
While I can't move my 2B to SS and see how they do, I can instead look at the SS who share the most similar traits to my 2B, and assume that if I moved my 2B to SS, that they would do as well as those SS! Pretty cool, right?
Well, the answer here is 4 runs. When comparing a SS and a 2B, you need to make a 4 run adjustment. What's really interesting about that number is that you get something similar when you do it the many different ways I've been doing it, as well as the "traditional" way.
What I will be doing is making a few more of these types of comparisons. The tricky ones will be the IF-OF comps. We'll see what that gives us.
Posted 2:47 p.m.,
March 24, 2004
(#14) -
tangotiger
Just looking at LF-RF. Wow, what a difference. Based on the Fans' evaluation, there are many more good fielding RF than in LF. And the number of bad fielders are abundant in LF and missing much in RF. Among my sample, the adjustment is a whopping +6 runs. An average fielder in RF should be +6 in LF.
This is in direct conflict with:
http://www.tangotiger.net/UZRmultiple.html
where I show that the avg LF (from 1999-2003) was 3 runs BETTER than the avg RF. In this case, I looked at the players who actually did play in both positions.
What we have here are:
- sample size issues with the Fans' scouting comparison
- Fans' evaluation may be positionally-biased
[an error occurred while processing this directive]
Posted 10:03 a.m.,
March 25, 2004
(#16) -
tangotiger
Good point.
Here are the average scores for the LF (1st line) and RF in the first 4 categories:
48 48 51 47
50 52 59 50
In each case, the RF is ahead of the LF. He's way ahead of the LF in speed.
In the throwing categories, it's no contest, as we'd expect:
45 44 46
59 69 59
The top speedsters in LF/RF are:
Ichiro
Crawford (LF)
Drew
Cruz Jr
Jacques (LF)
Stewart (LF)
Vlad
Sanders
Abreu
Mondesi
Tucker
At the bottom of the pile:
Burrell
Salmon (RF)
Alou
Pujols
Ramirez
Berkman
Dye (RF)
Floyd
Giles
Ibanez
Green (RF)
The fans sure think that the better fielder is in RF.