Wednesday, May 29, 2019

The algorithm eats itself.

One of my favorite epithets flung around by the sabremeretricious is, "small sample size."  This is usually unleashed whenever one makes a prediction about any event, with the implication that anything short of a twenty-year chunk of day-in, day-out action is too small to draw any conclusions.

I heard the demons in my head shrieking this last night, when they announced that The General was 0-8 with 5 strikeouts against old friend Kirby Yates.  No, no:  small sample size!

But sure enough, LeMahieu then grounded into the ever thrilling, delayed-video-replay game-ending double-play.

Hey, this is not to bad mouth either The General or Ma Boone.  Of course you had to bat him there.

But what occurred to me was that, if you really make pitchers completely interchangeable, you'll kill the algorithm.

EVERYTHING will be "small sample size."  No pitchers will ever face batters often enough to get a really good grip on how they fare against one other.

Out go the algorithms.

Which is probably exactly what baseball owners want.

Lately, we're seeing serious signs that the old cartel system of professional sports leagues—which originated in the United States, with baseball—is starting to break down.  Our leagues/cartels are always trying to establish parity.  But the harder they try these days, the more that goal seems to elude them.

The draft, as Duque and Will Leitch have pointed out, is more and more an ineffectual anachronism.

Beyond that, the leagues are now just too big for many teams to get even a whiff of a championship for 30, 40, 50 years (Yes, I'm looking at you, New York Knicks!).  There are too many ownership groups just free-riding on the backs of teams that are actually trying (And yes, I'm looking at you, Wilpons!).

Rather than more parity, we see team after team running up huge win totals, while everybody else festers.  Franchises resort to tanking—but with what these owners want to charge, fans just aren't willing to sit through that for 4-5 years at a stretch anymore.

Throw in the diminishment of the pitcher, the increasingly one-dimensional nature of the current game, the constant injuries to superstars (whose accomplishments are often achieved under a cloud of PED suspicions to begin with), the general mistreatment of the fans, and the whole societal shift to entertainment through social media...and you have the reasons for all those empty seats we're seeing this spring, on the road and even in the Bronx.

So what to do?

Seriously reassess how to keep players healthy and playing?  Teach pitchers how to pitch without destroying their bodies and batters how to hit to the opposite field?  Find every way we can to go back to the old ways of playing that everyone prefers?  Maybe even...charge less money???

Nah.

What baseball owners would probably prefer to do is go to the "Rollerball" model:  nobody is above the game.  Nobody is anything more than an interchangeable cog in the machine.  No superstars means no super salaries.

Or put it in medical terms:  faced with a sick patient, the owners aren't interested in curing him, so much as altering his genetic structure.

If you can't achieve parity through everyone trying hard to build the best team they possibly can, you try to achieve it by making each at-bat as random as possible.

It will be more like a crude video game than a traditional fan experience.  With the only problem being that people prefer to actively play video games, and can do that almost for free.  Hmm...








23 comments:

HoraceClarke66 said...

Surrenderer warming up. Need two more runs. Fast.

Anonymous said...

OK, genius. Tell us how pitchers can throw dozens of pitches at that velocity during a game without injuring their arms--you who seem to think that all the problems of baseball can be solved by dialing back the Wayback Machine to 1950. You have zero idea what you're talking about. You seem to have a manic typing disorder accompanied by vacuity of the cranium.

And speaking of small sample size--do you think that's an anomaly or something that cropped up last week? I mean JAYZUS--it's a statistical reality that you need to roll the dice a thousand times or so to establish the actual probabilities of certain numbers coming up. Same with player performance--even Joe D. and Ted W. went 0-for4 occasionally in a single game or maybe even slumped for a week or two. Did those individual games or brief stretches serve as definitive guides to their skills, or do we judge only by the larger samples? I feel like I'm talking to a two-year-old. No wait--a two-year-old would grasp this better than you.

The one small sample size we'll never have to worry about is the volcanic outpouring of your incoherencies and inanities every day on this blog. Cold comfort.

Anonymous said...

"Throw in the diminishment of the pitcher, the increasingly one-dimensional nature of the current game, the constant injuries to superstars (whose accomplishments are often achieved under a cloud of PED suspicions to begin with), the general mistreatment of the fans, and the whole societal shift to entertainment through social media...and you have the reasons for all those empty seats we're seeing this spring, on the road and even in the Bronx."

Evidence cited for this torrent of incoherent conjecture? Zero, as usual. Just the fevered outpourings of advancing dementia.

Anonymous said...

"But what occurred to me was that, if you really make pitchers completely interchangeable, you'll kill the algorithm.

EVERYTHING will be 'small sample size.' No pitchers will ever face batters often enough to get a really good grip on how they fare against one other.

Out go the algorithms.

Which is probably exactly what baseball owners want."

Truly demented paranoia, fueled by the usual blizzard of nonsequiturs. I'm sure that at the annual baseball meetings there's a secret conference of owners every year devoted to shrinking sample sizes--job one, next to figuring out how to win their divisions.

Duque--don't you think it's time that HC66 had his posting privileges curbed, for the sake of the credibility of this blog? I'm sure even you realize that there's a serious problem here.

Anal-a-mouse said...

"You and others like you drive away potential participants in this blog in droves. Just shut the fuck up already."

Yes, that is you, puckered.

BTW, my screen name is parody of what you do in mommy's basement. But you've already proven that you have no sense of humor. Thus, the reason you don't understand this site. Keep trying though -- eventually an infinite amount of monkeys randomly typing on typewriters will come up with War and Peace. I'd put your odds at understanding below that of tunneling for a large atomic mass object, say a piece of ham.

JM said...

Duque, you really need to do something. I don't know if it's a Russian bot, or a langauage-learning bot, or what, but this jerk is now often monopolizing the comments, like here.

It's sad, but it needs to be done.

One comment on this :

Q "Tell us how pitchers can throw dozens of pitches at that velocity during a game without injuring their arms"

A They simply stop throwing at that velocity so much.

Done.

JM said...

It's not rocket science. Want pitchers to reduce the risk of injury? Teach them how to pitch, not how to throw 100 mph.

What a concept.

Anonymous said...

JM--have you ever thrown anything but your week-old underpants into the laundry hamper? And do you have any data on the effectiveness of pitchers who throw 85 mph vs. 98 mph? Or are you just slinging your usual lazy bullshit?

Anonymous said...

Instead of editing coherent comments, duque, why not consider resuming your career as a newspaper editor and vetting HC66's preposterous incoherent ramblings about conspiracies among owners to limit sample sizes before you allow them to appear on this blog?

HoraceClarke66 said...

Nice win! Good to see Cortes the Conqueror keep his cool in that 9th inning, which could have got ugly.

8 straight series wins—most for Yanks since 1998. Plenty of dingers for the power-mad, and another great stop by Gio (not to mention a home run).

Bring on the BoSox! (Gulp)

Carl J. Weitz said...

I don't think small sample size of hitter vs pitcher was ever relevant. Now, if the batter faced a same-division pitcher regularly over a number of years and had at least 80-100 AB over time you might find some meaningful correlation between the two players.But even then it should be framed by context such as more recent at bats which might be indicative of some kind of adjustment either player has made against each other. And although their might be less AB against the same pitcher there should be enough AB against a certain type of pitcher (vs left-handers, vs sinker-ball pitchers, etc.

One particular stat that I always thought was kind of bogus was how a hitter fared with RISP or with men on base. Other than the perception that the batter "bears down" with runners on base, I don't see how having men on base would influence a batter's skill against the current pitcher he is facing.

But these are just observations of someone who is not the most statistically astute nor am I an actuary.

HoraceClarke66 said...

Sure. These are the eternal debates, Carl. I mean, if you hit a solo shot you're driving in yourself, right?

HoraceClarke66 said...

Oh, and also—very heartening outing from Paxton, of course. We can only hope this continues.

Anonymous said...

Carl J. Weitz makes some good points, especially about so-called clutch stats. Those stats, taken over the course of a career, begin to approach a meaningful sample size, and in almost every case are not appreciably different from the batter's overall offensive stats, thus giving the lie to the cherished myth of "clutch" hitters. Of course, you have to believe in empirical evidence as opposed to fond mythos from childhood to give this much credence.

HoraceClarke66 said...

I agree, and that's one of the nice things the new stats have shown us.

ON THE OTHER HAND...we all know guys we think of as especially clutch. And ballplayers believe in it, too.

Reggie Jackson talked about it as just trying to stay where you normally are in games—not letting the special tension of big moments and big games get to you.

Some guys are better at that than others. Why would they not be?

Anonymous said...

The analysis of "clutch" doesn't really rely on novel stats--it's just taking advantage of computer technology to track performance in specialized situations.

Anonymous said...

And what the analysis shows us is that there really aren't guys who are better in those situations than others--at least not by a significant degree.

Anonymous said...

Reggie Jackson was pretty clutch in the World Series. Unless 1.212 OPS is not clutch. Then he was not clutch. Or if you don't consider the World Series a pressure situation.

HoraceClarke66 said...

You im-becile! (quoting Peter Lorre)

Don't you know that Reggie Jackson had only 116 plate appearances over those 5 World Series??

Small sample size!!!

Anonymous said...

So players with enough plate appearances would be very few, so small sample size?

Did puckered name his johnson "sample size"? Is that why he likes to name it?

Anonymous said...

BTW,

Jackson's seeing 7 pitches from 4 pitchers in game six is one of the most amazing things I ever saw. 4 pitch walk, followed by 3 first pitch HR's.

I didn't like him because he pissed off Thurman Munson, my favorite player. But that was amazing.

John Keck said...
This comment has been removed by a blog administrator.
Isiyku Abdulahi said...


I really want the world to know about this great man who brought back happiness into my life again after my husband left me and the kids 3 years ago for another women online when i contacted Dr Believe he cast a love spell for me within 48 hours my ex husband start calling me and begging for forgiveness for everything that have happened between us. I was so happy to have my family back together with love again here is the email of Dr Believe via believelovespelltemple@gmail.com a man with the great powers you can also call him or add him on Whats-app: +2348156148821
God bless you
I am very grateful for your help in my marriage.