Saturday, November 07, 2009

Letting Go of Go-Go

See that picture on the right? That's how I think I'll always remember Carlos Gomez's tenure with the Twins. I loved the joy he played with, and was extremely optimistic about his potential. I constantly grumbled that the Twins were going to ruin the kid by not sending him to Rochester to play every day. The joy and athleticism he shows in that picture captures everything good about Gomez for me.

And I'm saying this because I've just reread the story that you're about to read and it sounds like I'm anti-Gomez. I'm not. I liked him.

But he wasn't any good.


---------------------

We bloggers and stats guys have enough to overcome. So how about we quit actively trying to look like morons?

In the joyous wake of the trade which sent Carlos Gomez to the Brewers for JJ Hardy, several otherwise sane and intelligent writers tried to point out that Gomez was undervalued because of his defense. Maybe that's true, and maybe it's not, but when it's carried it to the point where it's stated that Gomez had any present-value at all to a baseball team, it's officially gone too far.

Gomez is a very good, maybe even elite, center field defender. If you're looking for a statistic to try and measure it, Ultimate Zone Rating isn't a bad one, and his UZR states that he saves about 15 runs over an average center fielder every 150 games in his career.

But you can also measure offensive futility using stats, and OPS is a nice short-hand (and intuitive) way to do so. Here are the center fielders in the majors with 300 or more at-bats last year, sorted by OPS. See if you can find Gomez on the list. (Hint, scroll down):

PLAYER TEAM OPS
Carlos Beltran NYM 0.915
Torii Hunter LAA 0.873
Matt Kemp LAD 0.842
Angel Pagan NYM 0.837
Andrew McCutchen PIT 0.836
Marlon Byrd TEX 0.808
Denard Span MIN 0.807
Shane Victorino PHI 0.803
Kosuke Fukudome CHC 0.796
Mike Cameron MIL 0.795
Adam Jones BAL 0.792
Cody Ross FLA 0.790
Grady Sizemore CLE 0.788
Nate McLouth ATL/PIT 0.788
Rajai Davis OAK 0.784
Curtis Granderson DET 0.780
Jacoby Ellsbury BOS 0.770
Dexter Fowler COL 0.770
Franklin Gutierrez SEA 0.764
Willie Harris WAS 0.757
Nyjer Morgan WAS/PIT 0.757
Melky Cabrera NYY 0.752
Josh Hamilton TEX 0.741
Michael Bourn HOU 0.738
Aaron Rowand SFO 0.738
Colby Rasmus STL 0.714
Vernon Wells TOR 0.711
Chris Young ARI 0.711
Gary Matthews Jr. LAA 0.697
Tony Gwynn SDG 0.693
Alex Rios CHW/TOR 0.691
B.J. Upton TAM 0.686
Rick Ankiel STL 0.672
Mitch Maier KAN 0.665
Carlos Gomez MIN 0.623
Willy Taveras CIN 0.559

Gomez wasn't just bad at that plate. He was awful. Eye-watering. Breathtaking. His badness approached "evil".

And how much did his defense make up? Not much. A bump in OPS of 4 to 5 points is generally worth a single run over 500 or so at-bats. So let's add a column which points out just how many runs better or worse each of these guys was over Gomez:

PLAYER TEAM OPS Runs
Carlos Beltran NYM 0.915 +65
Torii Hunter LAA 0.873 +56
Matt Kemp LAD 0.842 +49
Angel Pagan NYM 0.837 +48
Andrew McCutchen PIT 0.836 +47
Marlon Byrd TEX 0.808 +41
Denard Span MIN 0.807 +41
Shane Victorino PHI 0.803 +40
Kosuke Fukudome CHC 0.796 +38
Mike Cameron MIL 0.795 +38
Adam Jones BAL 0.792 +38
Cody Ross FLA 0.790 +37
Grady Sizemore CLE 0.788 +37
Nate McLouth ATL/PIT 0.788 +37
Rajai Davis OAK 0.784 +36
Curtis Granderson DET 0.780 +35
Jacoby Ellsbury BOS 0.770 +33
Dexter Fowler COL 0.770 +33
Franklin Gutierrez SEA 0.764 +31
Willie Harris WAS 0.757 +30
Nyjer Morgan WAS/PIT 0.757 +30
Melky Cabrera NYY 0.752 +29
Josh Hamilton TEX 0.741 +26
Michael Bourn HOU 0.738 +26
Aaron Rowand SFO 0.738 +26
Colby Rasmus STL 0.714 +20
Vernon Wells TOR 0.711 +20
Chris Young ARI 0.711 +20
Gary Matthews Jr. LAA 0.697 +16
Tony Gwynn SDG 0.693 +16
Alex Rios CHW/TOR 0.691 +15
B.J. Upton TAM 0.686 +14
Rick Ankiel STL 0.672 +11
Mitch Maier KAN 0.665 +9
Carlos Gomez MIN 0.623 +0
Willy Taveras CIN 0.559 -14

So those 15 runs that he makes up defensively still leave him towards the bottom of the list. He was nowhere near average. He was bad. As in, not good. Beyond shaky. Poor. Inferior. Deficient. Lacking. Flawed. Substandard. Occasionally dreadful. Frequently appalling. Approaching abysmal.

To be fair, when a lot of people praised Gomez's value, their point wasn't really that Gomez was good. It was really:

1) "Hey, we've got these great defensive metrics that I want to show you!" I'm glad we have these defensive metrics. They're a step in the right direction. But a position player's defensive value is almost always dwarfed by the value of their offensive performance. Just because we can measure it now doesn't make it more important than it was a couple of years ago. Just clearer.

2) "Delmon Young is terrible." Which has absolutely been true. I'd love nothing more than to see the Twins bring in a veteran backup that can show Young how to conduct himself on and (maybe more imporantly) off the field. And, even better, threaten to take Young's job if he doesn't straighten out.

But that's far different than saying Carlos Gomez had any present-day value whatsoever. He was never part of the solution. He was always part of the problem. To argue anything else makes us look stupid.

18 comments:

TT said...

"But a position player's defensive value is almost always dwarfed by the value of their offensive performance. "

I don't think that is even remotely true. Which is why guys like Luke Hughes will never play in the big leagues.

Anonymous said...

Don't suppose you could show us where Punto fits on an OPS scale?

Anonymous said...

I think you nailed it. People who compare Gomez to Young are missing the point. If you want a no-hit defensive player they are a dime a dozen. Stick in Jason Pridie, or someone you've never heard of, and their won't be much of a drop off. They might actually steal a base once in a while, something Gomez couldn't do. Upgrade in left field if you can, but putting a no-hit defensive specialist isn't the answer in left or center.

I hope Gomez has a mind-boggling epiphony and becomes a brilliant hitter. That would only encourage other teams to trade more players for our high potential minor leaguers. But the fact is, even if he learned to hit, in a few years he'd be stuck behind Hicks, Revere, Morales, or Benson anyway. To get a starter for a backup is a very good thing, especially at a position of need like shortstop.

John said...

I don't think that is even remotely true. Which is why guys like Luke Hughes will never play in the big leagues.

Of course it's true. The difference is pitching.

ubelmann said...

One season's worth of numbers are going to overstate the case for anyone who has a season at the extreme ends of the scale. On top of that, looking at his PrOPS this year, we have reason to believe that Gomez was somewhat unlucky given the number of line drives he hit.

Let's take a look at Span and Gomez, since you'd rather shove Delmon under the rug. ZiPS has Span projected for next year to be a .757 OPS player. (I'm sure you'll rail against that prediction, but ZiPS is as good as any projection system out there and you haven't shown that your projections can beat ZiPS.) Gomez is projected to be a .662 OPS player. At your conversion of "4-5 points of OPS per run" that puts Gomez 21 runs behind Span.

Looking at Span's fielding numbers is somewhat tricky. If we just look at his CF numbers, they don't really jive with his LF/RF numbers, but you usually need about 3 seasons of data at a given position to get a good idea of how a player will do defensively in the next year. There's about a 10-run adjustment per 150 games between the corner OF and CF, so we can look at Span's corner OF numbers, take off 10 runs per 150G and consider that his equivalent CF performance.

If you do that, you get Span as a -8.5 R/150G center fielder (instead of the -13.8 R/150G CF he is listed at), and you get Gomez at 14.1 R/150G. I'll even give you some regression to the mean on top of that and say that Span is -5 R/150G and Gomez is +10 R/150G. So that puts Gomez 15 runs over Span on defense.

Putting it all together, Span is something like 5-6 runs per season better than Gomez as a center fielder going forward.

Now maybe you'll quibble a bit about their OPS projections or how you do that conversion exactly, but even if you give Span an extra 10-15 runs on offense, it's not like Span's edge over Gomez is the difference between being an obvious part of the solution going forward and not being "any good."

ubelmann said...

Also, this whole business about bloggers and statnerds making themselves look bad is silly. Are pocket protectors part of the uniform in Milwaukee? Obviously they thought they had something to gain here, and it's not as though they are the Royals--they've been just as successful as the Twins over the last three years.

This trade was a pretty even one, with the Twins getting a better player for a shorter period of time and the Brewers willing to make that deal because they save money (which they can invest on improvements elsewhere in their team) and get a cost-controlled player for a longer period of time.

If we want to be simplistic about it, one of these lines is J.J. Hardy in the minors and the other is Carlos Gomez in the minors.

.278/.339/.399
.271/.332/.405

Both players play good defense at up-the-middle positions. Is it really so difficult to believe that the Brewers could end up getting a good deal out of this?

TT said...

"Of course it's true. The difference is pitching."

If it were true, the Twins would play guys like Justin Huber at shortstop. Defense is the primary difference between players, offense is a secondary consideration.

I think that is why people take defense for granted. There are only a handful of truly poor fielders in the big leagues and they all hit like David Ortiz.

Pitching is certainly important, but not even Nolan Ryan struck everyone out.

"If you want a no-hit defensive player they are a dime a dozen."

No, there aren't. Not that play defense like Gomez. Pridie isn't even close and he is well above average for major league outfielders.

TT said...

"Obviously they thought they had something to gain here,"

They had a guy to unload who had nowhere to play for them. The same way the Twins did. I think its an even trade because both of these guys are huge gambles. You could win the lottery or crap out or end up somewhere in between with the first two being the most likely.

JK said...

OPS+ is not the stat to compare Gomez to Span. OPS gives too much credit to SLG and not enough to OBP. This problem is compounded when comparing a high OBP/low SLG player like Span to a low OBP/high SLG like Gomez.

Using Zips projections and 3 year weighted averages of UZR regressed to Tom Tango's fan scouting report I calculate Span at 2.8 WAR and Gomez at 1.2 WAR. This is basically the difference in WAR between 2009 versions of Kubel and Punto.

Anonymous said...

Ah JK, a little help here buddy... Let's assume that not everyone here is up to date on every metric out there (that would be me), could you clarify your position for the rest of us please?

Thanks,
Shannon

jim h said...

I agree with the Geek, Gomez has little present value for the Twins. Trying to assign him value through UZR is not likely to be very accurate. He would have to be the best defensive center fielder, ever, for him to be as valuable as UZR suggests. I like defense and think it is very important to a team, but it is true,(as tt says) you have to be a good defender to play in the majors. Especially at the premier defensive positions like SS and CF. The difference between good major league defenders and elite defenders can't be as great as UZR suggests.

I believe that most chances are major league-routine. Probably 90% are routine(you can substitute a different % if you like, this is for discussion). Likely 90% of the balls that are caught would be caught by any major league CF, likely 90% of the hits to center would be hits no matter who the CF is.

In 2009 Gomez had about 2.8 PO's per 9 innings. If he caught every difficult/potentially catchable ball hit to center, that would likely result in about 50 catches(based on my 90% suggestion). You would have to assume he would catch everyone of those and Span would catch none to get a difference of 50 PO's between the two over 150 games. Now, I doubt that 50 PO's would result in 15 runs, though I suppose it could. But I also doubt if there is that big of difference between Span and Gomez defensively. Span seems to be more fundamentally sound and probably positions himself better, hits the cutoff man better etc. That would tend to reduce the value of Gomez's greater speed.

Elite defense has value in many ways. It reduces the number of pitches thrown in a game and that can keep your starters in the game longer. It certainly saves runs over the course of a season. Still, I think UZR has to way overrate its value when it translates its findings to runs. I also think Gomez will have to be more fundamentally sound to be a truly elite defender.

So getting to The Geek's original point, this is a very good trade for the Twins as long Hardy is above average defensively. It is even better if he turns out to be above average offensively, since at this point Gomez is way below average offensively.

John said...

Ubelmann,

I appreciate the generosity, and you're right about almost all the places I would question the numbers. ZIPS might be good for overall projections, but I'll absolutely question a 757 OPS for a guy who is 25 and has a 811 OPS over 1100 at-bats.

But rather than go through a bunch of detailed points, I'll just agree that Gomez is probably 10-15 runs better defensively than Span, but I'd put Span as 30-40 runs better than Gomez offensively. Ultimately, I'd say we're in the ballpark of a 20 run difference.

Is that significant? Is it the difference between part of the solution and part of the problem?

I think the answer is yes. Punto was part of the problem this year, but part of the solution last year, and that was about a 100 OPS shift. If Gomez was posting a .260/.330/.390 (OPS 720), instead of the 623 OPS (or the 638 career OPS) he would have been part of the solution too.

And I'm puzzled by the second comment. Surely you can't believe that Gomez will post a 800 OPS over two 600 AB seasons? That can't be your point.

If people are underrating anything, it isn't the masses underrating Gomez defense. It's the sabremetricians underrating just how bad Gomez was offensively. Maybe we can't get our head around a performance this extreme. He wasn't just bad. We're talking about a .292 career OBP over 1102 plate appearances so far.

JK said...

WAR, stands for wins above replacement. It is calculated by using the offensive runs above average + defensive runs above average + positional adjustment + replacement level. Here is the explanation in more detail. http://www.insidethebook.com/ee/index.php/site/comments/how_to_calculate_war/

0 WAR is replacement level. 1 WAR is a below average regular. 2 WAR is average. 3 WAR is good. 4 WAR is an all star. 8 WAR is Joe Mauer.

I like it as a stat because it allows comparisons of players with different skill sets. Great hitter, terrible fielder = terrible hitter, great fielder = average hitter, average fielder.

My point was that Span is much better than Gomez right now. The difference is roughly 16 runs or 1.6 wins. Gomez is more valuable than Young (1.2 to 0). But Hardy is better than Casilla/Tolbert (3.3 to 0). The Twins added 2 WAR in the trade, so while I think Gomez has potential and should have played more than Delmon, it is good to see the front office improve the team.

Anonymous said...

Ubellman, that's kind of a disingenuous use of metrics. You don't need a lot of advanced stats to figure out the difference between Span and Gomez.

Span can hit and gets on base. Gomez can't hit and doesn't get on base. Mix in some serious issues with makeup (think of O.Cabrera's quote "what, you can remember all these handshakes but you can't remember the signs?") compared to Span's plus makeup, and it's not really a comparison at all.

ES16 said...

Wow, you guys are geeks! All these obscure metrics are giving me a headache and don't really tell the story abut Go-Go. Here's the real scoop: Go-Go showed no ability to take batting instruction and apply it in his at-bats. He made the same mistakes over and over, and showed he was incapable of coming to the plate with a plan and executing it. On the base paths, he had significant trouble reading pitchers. In addition, even though he is a great OF, he had major problems there, as well. He frequently took poor routes to the ball and threw to the wrong place. These are all red flags to a coaching staff and GM, and none of them can really be measured by statistics. The Twins did well to deal Go-Go, while he still had some value.

BeefMaster said...

Wow, you guys are geeks!

Agreed - I never would've expected something like that at TwinsGeek.com.

I loved watching Gomez play when he was playing well, but I never ever wanted him coming to bat in an important spot (not that I am at all willing to give back his division-winning run). Hopefully he can learn to hit and have a decent career with the Brewers while Hardy rediscovers his batting stroke at Target Field.

Anonymous said...

I know that the defensive metrics focus mostly on range. How effectively do they capture throwing to the wrong base (or inaccurately)? I'll be excited to watch Span develop as a CF.

John said...

ES16,

You come here with a handle of "ES16" and call US geeks?!? :-) And sadly, you have only seen the tip of a very sad, pathetic iceberg.

Sincerely,
"He's a Twins fan. He's a geek. It's kind of sad, really.: