Sunday, June 14, 2009

Part I: Switch-Pitching: When S.O.P. Kills Innovation, &
Why That Can Be a Good Thing  

In Baseball (and all other endeavours), managers benefit from pattern-recognition, noting what works and what doesn't and generally avoiding repeating efforts that mimic or parallel past failures.

If you know the printer you send your four-color glossy work to and that does a great job of it is also always late with mediocre product when you send them a two-color job, it doesn't serve you well to send them your two-color work again. Though, volume discounts aside, it probably doesn't mean you should stop sending them the four-color jobs at which they excel. No big epiphany, that. We managers survive and thrive because we recognize patterns and act on 'em.

¿But what about innovation that breaks the (not unreasonable) expectations of standard operating procedure? Breaking standards is Innovation's norm (because if it was standard, it probably wouldn't be an innovation).

It's a very tough situation because if you always stick to standards, you cannot exceed standard performance except through luck, and management, as Branch Rickey once said, "Luck is the residue of design", that is, more often than not, luck seems to favor more coherent planners. And if you break standards all the time, you miss out on the highest upsides but also the worst crashes.

So it's with mixed leanings I read the New York Times' Alan Schwarz piece this morning on the A League reliever Pat Venditte, apparently pro baseball's only current switch-pitcher. Venditte labors for the Yankees' affiliate, Mike Veeck's Charleston River Dogs. And when you see that Veeck affiliation, if you haven't looked yet at Venditte's stats, you may think the 24-year-old is another brilliant but satirical Veeck promotion like Vasectomy Day or Enron Night, but you'd be off base.

PAT VENDITTE (from MinorLeagueBaseball.Com)
Opp. Avg.
Reliever 2 1 0.64 25 20 28.0 20 4 2 1 1 38 1.59 .196
vs Left 0.00 10.0 4 2 0 0 1 19 1.75 .121
vs Right 1.00 18.0 16 2 2 1 0 19 1.54 .232

Please note, against lefties he's allowing 5 baserunners in 10 innings (when allowing twice as many runners, 10, would be "very good") and striking out 19 in those 10 innings (when striking out only 10 would be "very good"). Against righties, he's allowing 18 runners in 18 innings, "very good", and striking out 19, "very good" plus a stitch. Feel free to ignore the E.R.A. ... as a reliever it's not particularly indicative. But the single home run allowed over his 28 innings is stingier than average.

He's worth paying attention to, but...

As Schwarz wrote:

The Yankees, whose bullpen is among the worst in the American League, have two arms in Class A ball leading the minor leagues in saves. The left-handed one has kept hitters to a .121 batting average; the right-handed one has not walked anyone in 20 innings. This would all be rather straightforward, except that both arms belong to the same body.

Pat Venditte, the only switch-pitcher in professional baseball, is one of the most dominant — and well-known — players in the minor leagues. National news organizations travel to Charleston, S.C., to revel in his uniqueness. Fans see his statistics and dream of matchup mayhem. But experienced talent evaluators see not just one underwhelming fastball, but two. Sorry, kid.

S.O.P. HATES PAT VENDITTE With stats like that, in an organization hungry for relief pitching, common sense tells you Pat Venditte would be getting groomed for a major league tryout, probably being advanced to AAA. Standard operating procedure, though, treasures a big fast ball, and Venditte simply doesn't have one he can produce with either arm.

As Schwarz wrote:

... other numbers loom larger: his age (he turns 24 in a few weeks, having spent four years at Creighton University) and those mediocre fastball readings. He is dominating much younger hitters with well-located stuff that probably will not survive as he moves higher.

If you have finesse but not overwhelming power , you have nothing to fall back on. You can learn finesse if your power fails, but if your finesse fails, you can't suddenly learn power (not intuitively obvious, but essentially true in the real world). So Nolan Ryan and Randy Johnson can re-start a loss of power with a monster curveball or slider, respectively, when batters started being able to do a little something with any fastball they threw that wasn't their best. Jamie Moyer and Doug Davis don't have that affordance; if their control goes off a little bit, they have no alternative on which to fall back.

In general, prospects with middling fast balls crap out -- the failure proportion of finesse guys like Moyer is a lot higher than the failure proportion of power guys.

Of course, it's somewhat a self-fulfilling prophecy. Because S.O.P. says finesse guys are quicker to crash and burn if something goes wrong, the inevitable something-going-wrong activates the quicker management trigger finger that presumes the standard -- that dude has failed.

BEYOND BASEBALL This quick managerial trigger finger too often cuts short a successful experiment. Experiments are held to a much higher standard than everyday vanilla decisions are. And to a large degree, they should be.

If you make a different reasonable decision every day in response to repetitive identical issues and get comparable results, you're wasting cycles -- it's one of those events that probably doesn't offer much variance in outcomes. If you picked one approach that worked and just made it automatic, you save brain time you could apply to areas on which you could have more effect.

If you could make a decision that consistently outperformed others, even just a little, that is worth your while. But the risk in changing procedure tends to make S.O.P. lovers resist and push back when possible.

I did some pre-microcomputer era work for a retail book place that had two owners, and one just loved S.O.P. Once a year, they had to shut down on a work day during the Xmas rush to do inventory for a state inventory tax. They were just too busy and stores just a little too big to count all the stock and the value of each piece away from work hours. SInce I'd been a clerk in a pair of book stores before, I asked them to experiment beforehand with sampling shelves, picking every third row of the six-row shelves, and and changing the rows they counted each time they moved to the next shelf. We tested the sampling for six shelves against the counting-all-rows sums for each of them, and overall it was within 2%, with the worst difference for an individual shelf being 5%. If they implemented the inventory sampling, they would save about half their labor hours and cost, and be able to be open an extra day between Xmas and New Year's.

I was confident they would embrace this, and one owner did, but her partner would not. The partner had worked for a store that had gotten in trouble because they had fabricated, Bernie Madoff-style, an ersatz inventory and had been caught in an audit. Of course, the inventory tax audits were always samples themselves. But S.O.P. as defined by the tax authority dictated counting everything.

Her full-count approach was safe, but put a ceiling on their Xmas performance they wouldn't have had otherwise.

AND FURTHERMORE... Any experiment to try a sensible/possibly-rewarding innovation goes up in value as the number of competitors willing to follow you goes down (inversely proportional, as Joe Grzenda used to say). So a franchise willing to dedicate a few of their energy bars to refining the work of finesse pitchers, building up a coaching capability & technology and scout pattern recognition lessons (in differentiating the will-certainly-crap-out-at-the-next-level from the needs-only-to-learn-this-to-succeed-at-the-next-level), all optimised for that kind of talent, might find a dense nugget and a set of reproducable successes where competitors find only "a problem waiting to happen". Which y'all already know if you read Paul De Podesta. Again, if everyone sticks to standard operating procedure, you have competitive entropy, a Lucretian Dance of Random Chance, a Whole Lotta Aaron Sele Goin' On.

The chance to gain through innovation in blisteringly zero-sum arenas such as bookselling and Baseball argues the Yanks should be willing to try out Pat Venditte at AAA, and give him a chance against more selective and experienced batters to see what he can do. In my heart, I want them to; I find innovation exciting.

But still stored in my head are additional management reasons they probably shouldn't. In Part II, to follow, I'll lay out the management reasons that argue against taking the risk in Venditte's case.

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

free website counter