CHONE fielding projections
We linked to Jeff Zimmerman's UZR projections a few weeks back, but here are CHONE's fielding projections. Even though they use a less granular measure (TotalZone instead of UZR), the thing that makes these unique is that they involve both major and minor league data. That makes them MUCH more reliable for young players with little or no major league experience. They also include catching stats, which I like.
Here are your Reds, with numbers given in runs per season compared to their position:
C Hanigan +7
C Hernandez -1
1B Votto +0
2B Phillips -3
3B Rolen +7
SS Janish +5
SS Cozart +2
SS Valaika -18
CF Stubbs +11
CF Taveras +7
CF Heisey +3
OF Bruce +2
OF Gomes -6
OF Dickerson +9
OF Balentien +5
OF Frazier +1
OF Dorn -2
Only one that seems drastically off the mark is Phillips, but for some reason the CHONE system seems to be consistently very tough on Phillips, both on offense and defense. Interesting ratings on Heisey, Balentien, Valaika, and of course, our man Mr. Stubbs.
Keep in mind that the SS & CF positions have much better fielders than 1B/RF/LF, and so +0 at SS/CF means someone is a better fielder than a +5 fielder at 1B/LF/RF. In fact, here's the currently accepted defensive spectrum over at insidethebook.com, which seems to work very well.
C +12.5
SS +7.5
CF/2B/3B +2.5
LF/RF -7.5
1B -12.5
This is largely based on studies of players who play multiple positions. It works best (i.e. is most predictive) within the infield and within the outfield, but other factors come into play when going infield to outfield and vice versa that affect these numbers (lefties can't play 2B/3B/SS, for example).
Anyway, the way to read this is, for example, to say that someone who is an average-fielding shortstop moving to second base will, on average, be roughly +5 runs above average at his new position. Or, an average left fielder moving to center field will tend to be about 10 runs below average (catchers are their own animal, and are getting a bonus here for not only being a challenging position but also suffering a hitting drop-off specifically because of their position...but no one is saying that Hanigan would be a gold-glove shortstop).
These position differences are the basis for the position adjustments used in the WAR calculations for position players.
0 recs |
18 comments
|
Comments
Valaika's projection
seems to be far below what his minor league performance has been according to http://www.minorleaguesplits.com/cgi-bin/pl.cgi There supposedly using TotalZone, he comes out something like a -3 which is a lot more acceptable than the -18 given here.
I'm guessing that's because as you move up a level
your performance compared to your competitors naturally decreases. These runs are runs above average and previously Valaika was being compared to the average shortstop in his league. Move him up to the majors, he’ll be compared to the average MLB shortstop, which is likely much better defensively than the average in the minors. I’m guessing Rally has found that negative defenders in the minors have even larger drop-offs compared to the league average when they reach the majors simply because of improved competition.
Definitely a good argument.
I thought I responded to this earlier, but apparently it didn't take
But yes, this is my interpretation as well. The average-fielding shortstop in the Pioneer league (to take an extreme example) has been demonstrated to not be as good as an average-fielding shortstop in MLB. Therefore, a below-average minor league infielder should be a well below-average fielder in the majors.
The same is interestingly not true for outfielders. Apparently, the main determinant of fielding quality in the outfield has to do with young, fast legs. Rally’s article on MLE-type work for minor vs. major fielding stats can be found linked from minor league splits when you look at the fielding pages.
That said, -18 still seems extreme. Valaika has questionable stickability at SS anyway, so it’s not surprising to see him rated as below-average. But I would have thought that he’d be closer to -5 or -10 at SS, so that he’d be closer to average at 2B or SS.
-j
I write at:
Beyond the Boxscore | Red Reporter | Basement-Dwellers.com | Twitter: @jinazreds
Given the SS aging curve, I thought there was less of a difference
between the average MLB SS and AAA SS. Interesting. For some reason I thought it was typically the bat rather than the glove which kept young shortstops in the minors.
Maybe for guys who do make the majors as SS's
But there are a lot of guys who are drafted as SS’s and eventually move down the defensive spectrum. I imagine that this must be less of an issue as you move up levels (e.g. Frazier moved off of SS pretty quickly), but there are still questions about whether Valaika can play shortstop in the majors and he’s in AAA.
-j
I write at:
Beyond the Boxscore | Red Reporter | Basement-Dwellers.com | Twitter: @jinazreds
that phillips projection is WAY off
by Dave from Louisville on Dec 6, 2009 10:44 AM EST reply actions
TotalZone doesn't like Phillips
it’s actually right about his typical TZ value.
Whether that is way off is a whole other question.
Definitely a good argument.
It mismatches both his UZR scores and his Fan Scouting scores
I think it’s bunk. A similar problem occurs with Erik Aybar, which is on Rally’s own favorite team. :)
-j
I write at:
Beyond the Boxscore | Red Reporter | Basement-Dwellers.com | Twitter: @jinazreds
The important question is, why is it bunk?
I agree that it is wrong, but what about the system makes it wrong? That’s what Rally should be tweaking.
Definitely a good argument.
Baffling
Willy is a plus defender and BP is a minus? Sorry, but any system that ranks Taveras highly is bunk. Any system that ranks BP poorly defensively is double bunk.
"Nate Silver is a genius" .... BK
the system itself isn't bunk
the fact is that individual players can be affected by the way the system is built. For instance, if BP plays more shallow or shaded more in one direction or the other than the typical 2B, it can make his numbers look a little off. Rally should be looking at his system and trying to understand why a clearly superior defender is coming up negative and then trying to find if there is a weakness that can be fixed. And it may not be fixable based on the source data that he uses, which I believe is retrosheet data.
And Willy is not bad defensively. We just remember his defensive mistakes more because he provided no help on offense.
Definitely a good argument.
from Armchairgm
His defense can best be described as inconsistent. His outstanding speed allows him to get to balls most other center fielders won’t even try for, but his routes often leave a lot to be desired. He struggles going straight back for fly balls and will take odd routes on balls hit into the gaps.
That sums up my observations nicely. Willy kind of reminds me of EE defensively…he can make a spectacular play then misplay a fly ball by taking a crazy ass route to the landing spot and playing it off the bounce.
"Nate Silver is a genius" .... BK
Per Fan Scouting Report: willy is average to slightly below-average in CF
…which is a well-above average fielding position.
http://mlb.mlb.com/team/player.jsp?player_id=430838
Seems right to me.
-j
I write at:
Beyond the Boxscore | Red Reporter | Basement-Dwellers.com | Twitter: @jinazreds
does TZ/Chone/etc take arm strength into account?
Or is that inherent in the positional relativism (someone may have the arm for 2b, but not for SS, LF but not RF, etc)?
"And then there was the USAID guy in Kandahar who drove a giant pink Cadillac, which the locals set on fire one day. If you wanted to destroy something symbolic during a riot, you just could not do better than that. Good stuff." - Ghosts of Alexander
Yes.
Pure TotalZone for OF is range on fly balls. But as far as I know (it’d be shocking if he didn’t), Rally includes his outfield arm ratings (as well as his double play turnings) in these projections. They aren’t arm strength, per se, but rather ability to control the running game on balls hit their way. Kenny Lofton had a fabulous arm rating despite a fairly average arm reputation, presumably because he got to the ball so quickly. And post-leg surgery Griffey Jr. had horrific arm ratings despite a decent arm, because he had a hard time getting to balls fast enough to hold runners.
Fan Scouting explicitly considers arm strength, as well as accuracy and footwork. Generally, with some exceptions, the objective arm ratings seem well correlated with fan arm strength/accuracy scores.
-j
I write at:
Beyond the Boxscore | Red Reporter | Basement-Dwellers.com | Twitter: @jinazreds
It may be beyond his control
Could be a data source issue. He’s using play by play data from retrosheet, but it is largely without hit location information. Unless he starts paying for BIS data like fangraphs does, it may be unlikely that he can improve it.
But that’s just a guess, of course. Overall, TotalZone works very well. It just misses on a few players. Same thing happens in other systems, of course. UZR, based on Stats Inc data, rates Ichiro as a below-average CF! Based on BIS data, it rates him as average in CF.
http://www.insidethebook.com/ee/index.php/site/comments/uzr_on_fangraphs_using_bis_on_ichiro/
Frustrating, but based on other sources including the Fan Scouting Report, sUZR just misses on Ichiro. And given the sUZR vs. bUZR difference, it’s a data source issue, not a methodology issue.
…yet another reason to use multiple sources.
-j
I write at:
Beyond the Boxscore | Red Reporter | Basement-Dwellers.com | Twitter: @jinazreds
... as well as PMR and +/-
Haven’t seen the ’09 numbers but Phillips scored high on all the metrics in 2007 and 2008 except for TZ.
Right, though UZR/PMR/+- are all BIS-based stats and thus use the same source data
I’d be more interested to see sUZR. But his ZR (based on STATS Inc data) led qualified NL 2B’s in 2007 and 2008, and was above average in 2009.
http://espn.go.com/mlb/stats/fielding/_/year/2007/position/2b/league/nl/order/true
ZR correlates extremely well with sUZR, so it’s probably good enough. TZ is simply wrong on Phillips’ defense. For most other players, however, it seems to perform well.
-j
I write at:
Beyond the Boxscore | Red Reporter | Basement-Dwellers.com | Twitter: @jinazreds























