THE BOOK cover
The Unwritten Book is Finally Written!
An in-depth analysis of: The sacrifice bunt, batter/pitcher matchups, the intentional base on balls, optimizing a batting lineup, hot and cold streaks, clutch performance, platooning strategies, and much more.
Read Excerpts & Customer Reviews
If you are a media member and would like a review copy of The Book, please contact Kevin Cuddihy of Potomac Books.

Buy The Book from Amazon

MOST RECENT ARTICLES
MAIL : You ask | We say

Advanced


THE BOOK--Playing The Percentages In Baseball

<< Back to main

Monday, October 02, 2006

Can fans evaluate fielders, without a position bias?

By Tangotiger, 02:49 PM

Darin Erstad switched from CF to 1B back to CF.  There is no more unusual position switch than going between CF and 1B, other than catcher.  How did the fans evaluate Erstad, year-to-year?


Year In Fi Sp Ha Re St Ac
2003 85 90 83 84 59 45 56
2004 84 83 74 86 67 48 64
2005 90 80 67 84 74 53 71
2006 81 79 63 76 56 28 58

If you don’t recognize the headers:
Instincts
FirstStep
Speed
Hands
Release
Strength
Accuracy

Hard to tell when he was a CF and when he was a 1B, right?  2004-5 was as a 1B.  The only anomoly, if you even want to call it that, is his release evaluation was about 10 points higher as a 1B than CF.  (His arm strength this year was way below his norm, suggesting something terribly wrong this year. )

All those who think it’s “impossible” for fans to evaluate a player, without respect to his position, raise your hand.

I’m going to look for more major position switches, and see how the fans did.  My gut feeling is that Fans do a great job.  But, let’s see.

Updated: Oct 3, 10:45
Here’s two high-profile switching teams/positions, Soriano and Nomar.

Soriano
Sori was a 2B from 03-05 (two in NYY, one in Tex), and then of course LF with Nats. Let’s see how the fans compared him.

Soria In Fi Sp Ha Re St Ac
2003 32 53 80 07 21 61 37
2004 63 82 84 17 31 55 32
2005 24 66 81 00 32 58 32
2006 37 60 84 41 50 71 71

His instincts were pretty stable, except for the huge blip in 2004/NYY. His first step again showed a huge blip up in 2004/NYY. However, the 2005/2006 comparison is the one we are most concerned about, and that is pretty stable. His speed is consistent across the board. His Hands shows the biggest change, with horrible hands as a 2B, and below-average as a LF. Same applies with his release/footwork. His arm-strength was always above average, but looks even better when evaluated as a LF. His arm accuracy is through-the-roof. If we take the average of 2003-05, and compare to 2006, we have:

Soria In Fi Sp Ha Re St Ac
0305 40 67 82 08 28 58 34
2006 37 60 84 41 50 71 71

Clearly, from the time he actually reaches the ball, to the time he gets rid of it, we see two different evaluations of Soriano.  Are his coordination numbers (Hands+Footwork) as horrible as the 2003-05 evaluation says, or just simply a bit below average as the Nats fans say?  Does he have an average arm, or a plus arm?  Is this a situation where the fans failed to evaluate him context-free, or a situation where Soriano is two completely different ballplayers when at 2B/LF?  Is it a situation where most players are two completely different ballplayers when at 2B/LF?

If professional scouts knew how well Soriano would have played at LF, they would have moved him there a long time ago, considering how much he struggled at 2B.  I opt for the situation that we’ve got two different Sorianos, that he adapted differently, and few could have expected his transition to have been as smooth as it was.

Nomar

Nomar In Fi Sp Ha Re St Ac
2003 73 76 69 53 59 90 50
2004 60 55 58 50 50 78 35
2005 58 42 50 40 42 71 39
2006 61 59 52 58 52 63 62

Nomar was a SS in 03,04.  In 05 he split his time at SS/3B.  In 06, he was a 1B.  He was in Boston in 03-04, Cubs 04-05, and Dodgers today.  He of course has been beset with injuries.  How did the fans see him?

For Instincts, fairly stable.  His first step took a big drop from 03-04, then stabilized.  Same with his speed.  His Hands have been fairly stable, with a dip as a 3B, and raised a bit at 1B.  His Release fairly stable.  His arm strength has been dropping year-by-year.  His accuracy dropped alot, then rose substantially in LA.

What can we see?  The only big concern I have is his arm accuracy with the Dodgers.  Every other evaluation doesn’t seem to have a position-bias (and we’ve got three different teams and three different positions).  It’s possible that with the very short throw from 1B that it’s simply too hard to evaluate arm accuracy at 1B, position-free.

This is the 1B report for 2006:
http://www.tangotiger.net/scouting/pos2006_1B.html

In terms of arm accuracy, Nomar is 5th among all 1B.  His 2003-2005 numbers (SS/3B) are pretty clear that his arm strength is far far superior to his arm accuracy.  As well, of all the 7 traits each year, arm accuracy was always his lowest score, year after year (and arm strength his high.) His 2006 numbers however show them as equals (arm strength still highest of all traits), and that both of those are his strengths.  Either the Fans blew it on this particular trait (just 12 ballots), or Nomar has adapted his throwing approach so well, as to fit into a 1B mold.  My guess is the Fans blew this one trait.  It would be more helpful to see video of him.

I’ll look for more…

Updated: Oct 3, 11:15

Brad Wilkerson

Mostly an OF in 2003 in Montreal.  Split his time 1B/OF in 2004.  Then went to Washington in 2005 as mostly a CF/OF, with some 1B.  Then in 2006, in Texas, as a LF.  Seems like a nice spreading around.  How’d he do?

Wilk In Fi Sp Ha Re St Ac
2003 60 45 48 54 49 55 61
2004 80 73 56 72 89 64 82
2005 66 60 56 61 61 52 57
2006 28 32 38 32 48 35 35

His 2006 numbers, across the board, are abysmal.  Let’s focus on 2003-05 first.  His 2003 first step is out of ... uh… step, with 04-05.  His Speed is stable.  His Hands look much better in 04-05.  His release was far far better in 04.  His arm strength is stable.  His arm accuracy much better in 2004.  This one is pretty strange.  The two seasons that most match are the 2003 and 2005 seasons, one in Montreal as OF, the other in Washington as an OF, with parttime 1B.  His split time in 2004 as 1B/OF seems to have biased his arm evaluation with Montreal fans.

His 2006 numbers however are completely different from the rest.  While we may have some position-bias issues, the 2006/Rangers evaluation shows something completely different.  Either he had some terrible injury, or Fans made a personal attack on him.  I’d like to hear from Ranger fans on this one.

#1    MGL      (see all posts) 2006/10/02 (Mon) @ 16:28

Geez, Tango, you are really digging in your heels on this, and trying to make something that is grey, black and white.  Do you seriously think that some if not many (or even all) of the fans do not know that Erstad played 1B and CF and that they have not followed his career through those movements, and that they would not rate him similarly even if he broke both legs and became the batboy?


#2    MGL      (see all posts) 2006/10/02 (Mon) @ 16:39

Just to be clear, I am not suggesting that that is the case, only that it is a possibility, and a distinct one at that.


#3    Rally Monkey      (see all posts) 2006/10/02 (Mon) @ 20:00

I’m one of the fans who has evaluated Erstad every year of this project.  I’m trying to evaluate on what I see that year, but I’m sure what he’s done in the past creeps in and affects things.

I think I can explain the arm rating - Erstad as a gold glove OF had a so-so throwing arm.  Playing at first base, we probably just rated his arm based on what we remembered from him playing OF.  How often do you ever watch a 1st baseman really throw anyway?  Going back to center, thw 2 year layoff, injuries, age, lack of throwing, whatever seemed to take its toll.

We were like, wow, Erstad’s arm is weak!  It might have been just as week in 2005, but its pretty hard to tell when he’s playing first.


#4    MGL      (see all posts) 2006/10/02 (Mon) @ 20:18

I couldn’t imagine reliably rating an arm “independently” for a first baseman, other than by what you might occasionally hear from an announder, or an occasional throw to second or home.  IOW, I think it is naive to think that an arm rating for a 1B is equivalent to an arm rating for any other position, especially an OF’er.

We somehow know that Juan Pierre (and Damon and Bernie) has a terrible arm in CF.  If these guys were at 2B, I have no idea how their arm would be rated.  I wouldn’t be surprised if their arms were rated as average or even above average.  When I played, I knew IF’ers who had guns for arms for relatively short throws, but couldn’t make a long throw with anything on it if their life depended on it.  I was one of those guys.  The technique for one is completely different from the other for one thing.

So let’s stop pretending that these individual skill ratings are fungible at different positions and that the positions themsleves and other other players at those positions have no influence on the ratings!


#5    Tangotiger      (see all posts) 2006/10/02 (Mon) @ 20:40

MGL, there is a wealth of data out there.  Rather than refusing to believe anything, why not analyze it?  That’s what I intend to do.  It shouldn’t be too hard to find guys who switched positions and teams.

As much as my heels may be digged, yours are right in there!


#6    MGL      (see all posts) 2006/10/02 (Mon) @ 22:14

Honestly, I am just speculating and sort of thinking out loud.  You are making some blanket statetments and reaching some bizzarre conclusions, IMO, and presenting them as such. 

Remember that you started out declaring that a CF was worth 1.5 wins, as compared to the “average” fielder!

I’m really not digging my heels in to any particular construct or paradigm.  Almost everything I have said thus far is along the lines of, “Be careful how you interpret these things...”

For example, you have flat out stated that the raw “skills” for CF as reported by the fans are very similar to those for SS, and therefore one can conclude that those two positions are roughly equivalent in value and perhaps even fungible (that a CF can play SS and a SS can play CF at roughly the same level of play), and I apologize for such a liberal paraphrase.  I merely said, “Hold on a minute partner,” and here is why.  That is NOT digging my heels in.  And of course the expression, “digging your heels in” generally is referred to, “in the face of contrary evidence.”

In fact, I am very open minded about all of this, as I have not worked with the data at all and am merely commenting on what you have been reporting, based on my own judgment and yes, bias. I take that back.  I have worked with the data a little, and my initial conclusion was that trying to come up with true positonal adjustments (what happend when a player moves from one position to another) is going to require a lot of “fudging” and that at best because we are dealing with almost insurmountable selective sampling biases, common sense is probably going to have to prevail. And of course when the data and analytical evidence offer little help, you are going to be left with lots of disagreement among otherwise knowledgeable and intelligent folk.  Sort of like government and politics…


#7    Trev      (see all posts) 2006/10/03 (Tue) @ 02:28

How many entries did you get each year?

If you’re thinking of other position switches, from the last four years, here’s what I’ve got off the top of my head:

Chad Tracy (1B/RF/3B)
Miguel Cabrera (RF/3B)
Chipper Jones (LF/3B)
Jose Reyes (SS/2B/SS)
Kaz Matsui (SS/2B)
Russ Adams (SS/2B)
Aaron Hill (SS/2B)
Michael Young (2B/SS)
A-Rod (SS/3B)
Placido Polanco (3B/2B)
Pedro Feliz (RF/3B)
Brandon Inge (C/3B)

Hope this helps.

I wonder if there are any changes to how a player is viewed after he switches teams.  Does he get a “new start” or is that pretty consistent?


#8    tangotiger      (see all posts) 2006/10/03 (Tue) @ 04:52

"Remember that you started out declaring that a CF was worth 1.5 wins, as compared to the “average” fielder! “

I said NO such thing!  I said the average fielding CF was +6 or +7 runs ahead of the average fielder, and that the SS was equivalent to that.  (If I had said the CF was +15 runs, and I’ve always maintained that CF=SS, that must mean I said that SS was also +15 runs, leaving -30 runs to be distributed among the other 6 positions.  That’s hardly possible.)

That the most we can see the CF drop would be in the 2B/3B territory of +3.

And certainly not, that the CF is an average overall fielder, which is what his average off level was.

***

Trev, in order to remove the “same team” bias as in the Erstad case, it would have to be “new position, new team”.  I’ll take a look at Soriano and Nomar in a couple of hours.


#9    Tangotiger      (see all posts) 2006/10/03 (Tue) @ 07:46

Go to the top of the page, as I have made an Update, that looks at Nomar and Soriano.


#10    Tangotiger      (see all posts) 2006/10/03 (Tue) @ 08:10

Added Wilkerson as well.


#11          (see all posts) 2006/10/04 (Wed) @ 10:59

So in the “How Reliable Can Fans Be” thread, I posted the results of a principle components analysis run on some data that Tango sent me.

I re-ran the analysis on a position-by-position basis. If the PCA gives us different factors for each position, it suggests the fan’s perception of different positions are, well, different. 

For instance, I think that a SSs “accuracy” is a totally different skill than a CFs “accuracy.” A SS has to play a grounder while keeping balance, set his foot, and make a good throw. Whereas a CFs accuracy won’t be affected as much by being off balance, and far more affected by long-range accuracy.

So I predict that SSs and CFs will have very different factors emerge from the factor analysis, reflecting differences in the skills.

Here are the factor rankings:
1B: arm (57%), catching (13%), legs (10%)
LF: arm (42%), legs (21%), catching (13%)
C: arm (57%), legs (16%), catching (9%)
----
3B: turnover (55%), legs (17%), reactions (9%)
SS: turnover (56%), legs (15%), reactions (10%)
----
2B: turnover (47%), legs (21%), strength(11%)
---
CF: catching(52%), arm (19%), legs (11%)
RF: catching(62%), arm (12%), legs (10%)

These factors were composed of the traits as such:
arm: release, strength, accuracy
legs: speed, first step
catching: instincts, hands
turnover: hands, release, accuracy
reactions: instincts, first step
strength: just strength

So for 1B, LF, CF, RF & C, there were distinct “catching” and “arm” factors. That is, fans clumped together instincts and hands into a “catching” factor, and clumped release, strength and accuracy into an “arm” factor.

But for 3B, SS, fans reorganized the traits into different clumps. They tended to see “reactions” and “play turnover” as distinct factors, where reaction was comprised of instincts and first step, and turnover consisted of hands, release and accuracy. This makes a lot of intuitive sense to me, since I would judge a SS on 1) his reaction to a batted ball, and 2) the reliability with which he catches the ball and transitions into a smooth throw to first. I wouldn’t judge a center fielder in that same way.

So when a player rank a SS and a CFs accuracy as equivalent, I’d argue that they are basing the judgment on totally different types of plays, and essentially reflect different underlying skills.


Page 1 of 1 pages


Name (required)
E-Mail (optional)
Website (optional)

<< Back to main


Latest...

COMMENTS

Dec 05 04:40
Sabermetric Moves of the 2009 Pre-Season

Dec 05 05:33
Avery being Avery

Dec 05 05:06
NYC’s 3 1/2 year mandatory jail time sentence for carrying a loaded weapon

Dec 04 23:42
Poll: Would you vote Raines for the Hall?

Dec 04 23:07
How to calculate the area of a baseball field

Dec 04 22:48
Complete Run Expectancy, Retrosheet Years

Dec 04 22:03
Raines for the Hall

Dec 04 15:55
Mailbags on Parade

Dec 04 14:01
What would happen if the shootout period was 10 minutes, not 5?

Dec 04 11:49
Estimating BABIP