Saturday, September 30, 2006

Ichiro Suzuki: Lessons in Personnel, Interpreting Metrics & Getting the Most From Your Talent  

An A.P. story indicated that Seattle Mariners right fielder Ichiro Suzuki was willing to play center field for the team next year. It helps the team in three certain ways & one possible one:

  1. He's a very capable center fielder, a position that requires more skill than right field (and, in fact, he may be a better CF than RF; more on that later).
  2. It protects a secret "weakness" of his other teams are starting to discover and act upon and turns it into a strength.
  3. It enables the team to be flexible in its approach to add a left-handed bat over the off-season (M's DHs were dead last in the AL in OPS and in RBI and in Runs scored).
  4. It overcomes a rumored gnarly shard of office politics that has possibly created a bit of resentment on the roster.

In this entry, I'll elaborate on the first benefits I mentioned and the lesson it offers beyond baseball, and in a future entry, I'll cover the rest.

When Suzuki played in Japan, he was a center- and left-fielder. When Suzuki came to the Mariners, they already had a premier center fielder in Mike Cameron. He really has good range and a better throwing arm than most left fielders, and perhaps he saw playing right field as either a way to showcase his arm or as a new challenge (Suzuki is relentless about setting himself challenges and then vanquishing them).

Suzuki has had a great reputation as a right fielder. Fielding measures are tricky and highly context-sensitive. Most fielding measures don't support Suzuki's reputation.

The readily available measures I look at are Stats Inc.'s Zone Rating (what proportion of balls in the area of the field for which a right fielder is responsible does he turn into outs) and an old standby, Range Factor. They don't always agree, and both are imperfect, though if a player is ranked highly in both or sucks wind at both, it gives me a strong suspicion the player is better than average or not. If you want a fuller explanation of why I take these as dead reckoning number and why I don't accept them fully, I've written about that here.

But Suzuki's reputation in RF and his measures don't align well. Beyond baseball, you see this a lot, where a staffer with a great reputation doesn't really produce superior measured results or overlooked staffers produce beyond the norm and remain overlooked. Most frequently that happens because measures are poorly designed and they're measuring something that has less to do (or even nothing to do) with success.  It happens almost as frequently because the reputation is there and management chooses to not let fact get in the way of a perfectly comfortable opinion. (If you're slapping your forehead and saying, "Jeez, ain't that the truth" and you want some more ammo on that subject, I can heartily recommend Pfeffer & Sutton's wonderful book, Hard Facts, Dangerous Half-Truths And Total Nonsense: Profiting From Evidence-Based Management; it's a subject I've written about many a time, but they have done a better job than I ever have.) And sometimes (this happens in the baseball research community, too) people tweak the way measures are delivered to confirm their pre-examination bias. It's not necessarily a bad thing...if Puckett is your department's perfect picker-packer and you deconstruct what it is that made her so effective, you could build a measure based on that profile and measure in units you call Pucketts. BTW: If you do this, I encourage you to name the units after the person for two reasons. First, it gives recognition and credit to your high performer. Second, it calls to management' attention why they have refined the measure this way -- when circumstances change, and they inevitably will, it gives them better perspective on exactly why they measure that way, making it easier to move on to a new, better adapted measure. But tweaking measures to confirm a bias can be awful, too, a way to protect underperformers by institutionalizing their mediocrity into "standards".

Suzuki's Zone Ratings in RF from about 2002-2005 were not uniformly stellar.

MLB Full-time Right Fielders -- Source = Stats, Inc. via ESPN.Com

Year Suzuki's Rank/
Total Ranked
Suzuki's ZR MLB Starters'
Median ZR
MLB Starters'
High-Low ZR
2002  17th of 18 .845 .891 .929 - .833
2003 4th of 16 .897 .880 .918 - .802
2004 14th of 16 .863 .877 .856 - .924
2005 14th of 19 .869 .878 .819 - .916
2006 4th of 17 .900 .864 .834 - .924

He's fast and he gets a great break on the ball, and goes back and to the sides especially well. Team insiders have leaked to the press that he's more careful about hitting a wall or going down to the ground than most, and while I'm not sure they're absolutely right, my observation of him in a couple of hundred games supports that as a generalization. So as a right fielder, he's "cautious" or "less aggressive" in two directions: back towards the outfield wall and to his left towards the stands in foul territory. He picks his spots, I believe, and is more aggressive in more leveraged situations, but he has been quoted in the press as saying he'd rather preserve his body and play 160 games than do a Jim Edmonds Flying Wallenda Daredevil Triple Axel  Spontaneous Combustion Thang and miss 40+ games a year to injuries linked to such behaviors.

This season as you can see, his Zone Ratings were up from past seasons. Keep in mind that his .900 means a trained observer has used judgement (not perfect micrometer measurement) to determine on each individual chance whose zone the ball was hit to, and in Ichiro's case, he's gotten to 90.0% of them. In his lowest ZR year, he "only" got to 86.3% of them, a 3.7% difference or a a rough difference of a ball every 20-something games. It could be chance in that a few extra balls in his zone are going into the right-center field gap (where he doesn't have a wall to avoid becoming-as-one with), it could be statistical noise. Based on my own observation, I'll suggest he was being a bit of a ball-hog. There was a parade of center-fielders out in Marinerland this year, and none of them were particularly studly-veteran types -- I saw him call off center fielders and then snare flies in the gap more often this year than in the past.

The context changed. Instead of a full-time partner out there, he was playing with a variety of associates he didn't know all that well and chose to take the alpha rĂ´le normally reserved for a center-fielder. Yes, his metrics looked better this year, but it wasn't because he changed his behavior on balls that might involve going to the ground. Within the range of skills he always executes well, he took incrementally more opportunities and, unsurprisingly, succeeded more.

But in-season, he shifted to Center Field. For most right-fielders this would mean lower effectiveness. Most right fielders don't have the foot speed and instinct of a centerfielder. Suzuki does, of course, because he played the position as a professional and it fits his best skills (foot speed, quick start on a hit ball) really well, so instead of exposing potential weaknesses, it harnesses his strengths. As anyone who plays outfield recreationally could have told you in advance, he's done really well in this shift. Here's a table of A.L. starting centerfielders with not-enough-games-to-qualify) Suzuki dropped in so you can see how he's stackin' up at a position he hadn't played regularly since he left Japan in 2000.

Corey Patterson, Bal 132 1073.2 4 .989 2.93 .925
Ichiro Suzuki, Sea 37 320.0 0 .991 3.12 .917
Vernon Wells, Tor 148 1280.1 3 .988 2.33 .904
Grady Sizemore, Cle 158 1361.1 1 .993 2.73 .887
Curtis Granderson, Det 155 1291.0 0 .997 2.67 .884
Johnny Damon, NYY 130 1083.2 1 .990 2.57 .883
Brian Anderson, CWS 132 949.0 1 .994 2.90 .882
Torii Hunter, Min 141 1214.1 4 .989 2.59 .870
Joey Gathright, KC/TB 128 996.2 4 .991 3.08 .864
Mark Kotsay, Oak 125 1036.0 2 .993 2.46 .855
Gary Matthews Jr., Tex 141 1219.0 2 .979 2.46 .847

Here's why Suzuki has rocked out in CF when most RFs would lag a little moving to the tougher position. Because in the context in which Suzuki's skills get applied changed again. Because in center, he has more room to exercise his skill of going back (more room to the wall) and to the sides (in which neither direction has a wall to contend with, ergo he can be laterally as aggressive as he wants to be, and is).

In the next entry, I'll describe a piece of Suzuki's game that' clearly not as good as his reputation, his throwing arm, and explain why his move to center benefits something that was being exposed as a "weakness".

In your own management, it's important to chronically observe not only "performance" as a single overarching way of measuring things, but to observe everyone's little skills that make up how and how well they do the job. This enables you to tweak job descriptions to take advantage of people's strengths and work around their weaknesses And if you observe staff this way, you have the opportunity to learn other things about what one needs to do to achieve high performance -- that is, once you are observing closely, you may discover that aptitudes and skills you ddn't think were relevant to high or low performance are making a difference. And one you know it you can act on that knowledge.

An example: One of my clients had a really nice senior technical guy who had control over internal MIS support along with a flotilla of other responsibilities. Tech Engineer Guy (TEG) really hated tech support, hated being interrupted from his zone (tech support requires that), and his attitude got imbued in his customer service. It wasn't that he was hostile, he just didn't empathize, and the users could easily sense he resented his time helping them. Yes, his level of technical skill made for potentially great technical support, but in this case, it was far outweighed by his attitude and his ability to deflect small needs gracefully. My client observed this enough that he was able to act on it and he redefined TEG's job to exclude that which he hated and wasn't good at and redeployed TEG's talents at the higher-value work he's good at. Win-win-win (Client's organization-TEG-users).

Do you have any contributors who do well when the metrics suggest they don't? Badly when the metrics suggest they do? Any who might benefit if you could deconstruct what it is they do in their jobs so you could re-define their job?

I bet you do.

This page is powered by Blogger. Isn't yours?

free website counter