Seattle Mariners - an exception to Sabermetrics?

[brettac1]

ARod pretty much proves the point of the sabermetic crowd. Before his last two post seasons he was extremely 'clutch' in the playoffs, the last two he has been horrible. So did he forget how to be clutch or something?

 

Bonds proves it, too, just in reverse. He was awful in the playoffs with the Pirates, then he had possibly the greatest World Series ever in 2002.

 

Or Jeff Suppan for that matter. He got shelled in postseason appearances prior to last year, then he was an ace suddenly.

[GenoSeligPrieb]

"The playoffs are a tiny sample, even a 7 game series is only 30 AB's. Meaningless."

 

This is the brand of absolutist mindset that those of us who question Sabremetrics find the most....intriguing.

 

In the playoffs, you're facing the best teams. Teams who shorten their rotation, bench and pen, so you're most often facing only the finest. The difficulty level is raised significantly right there. When the Braves won their division, they left Wes Helms off the playoff roster. And you won't be facing a gimme like a Jimmy Haynes as the starting pitcher. You can expect to be challenged every inning.

 

Then, add in the fact that, if you lose the series, it's on the biggest stage of your career, and failure is magnified exponentially.

 

How well you do in the face of all this pressure, against this top-quality opposition, goes a long way toward defining you as a player. Your failure/success at the top level of your profession is about as meaningful as you can get.

 

I understand the notion of the small sample size. Believe me, it's been tattooed into my grey matter by now. But when it's playoff time, you can't enjoy the luxury of discounting or dismissing bad performances, while waiting for a larger number of plate appearances to act/decide/assess, because the hitter won't get the chance to right the ship in Game 8 of the series.

 

"Yes, the mathophobes are right, these are games played by people and not computers, and that's exactly why they're wrong when they make stubborn and ignorant proclamations of the uselessness of stats. "

 

I'll never call stats useless, but how come a significant number of these predictive models (Mariners, White Sox, Diamondbacks, Indians? Marlins?...are there more?) are proving inaccurate? I get where you're coming from, mother, stats are a great comparative tool. But I just chafe at the concept of someone posting:

 

Player A is better than Player B, because:

 

A) .278/.354/452

B) .297/.336/.434

 

Case closed!!

 

We've all seen it here. Arm strength, ability to deliver when situations call for it, speed, defensive range, attitude...anything else you can name which helps assess a player's capabilities... None of these factors are calculated into these 3 limited stat lines, but they're repeated as gospel....and even then, sometimes the predictive models can be way off! http://forum.brewerfan.net/images/smilies/wink.gif

[homer]

"but how come a significant number of these predictive models (Mariners, White Sox, Diamondbacks, Indians? Marlins?...are there more?) are proving inaccurate?"

 

Exceptions to the rule do not invalidate said rule.

[Phenomenal Smith]

In the playoffs, you're facing the best teams. Teams who shorten their rotation, bench and pen, so you're most often facing only the finest. The difficulty level is raised significantly right there. When the Braves won their division, they left Wes Helms off the playoff roster. And you won't be facing a gimme like a Jimmy Haynes as the starting pitcher. You can expect to be challenged every inning.

 

Then, add in the fact that, if you lose the series, it's on the biggest stage of your career, and failure is magnified exponentially.

 

How well you do in the face of all this pressure, against this top-quality opposition, goes a long way toward defining you as a player. Your failure/success at the top level of your profession is about as meaningful as you can get.

 

I understand the notion of the small sample size. Believe me, it's been tattooed into my grey matter by now. But when it's playoff time, you can't enjoy the luxury of discounting or dismissing bad performances, while waiting for a larger number of plate appearances to act/decide/assess, because the hitter won't get the chance to right the ship in Game 8 of the series.

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

You can call it failure if you like, but where people go wrong is in attributing the failure to a lack of character. The A-Rod, Bonds, and Jackson examples show the flaw in this. In some series they were brilliant, and in other series they were terrible. Did they have the requisite character all along, or was it only present when they didn't fail? And this is where baseball is different than other sports - you generally only get 4 or 5 plate appearances a game. A basketball player can just keep shooting until he finds his shot. Michael Jordan missed 20 shots in that final championship game against Utah, but nobody remembers that because he finally hit some shots in the 4th quarter.

 

On a side note, the Mariners hitting "close and late" looks like it's largely a factor of an unusually high BABIP in those situations. I doubt they can keep it for the rest of the season, but we shall see what we shall see.

[GenoSeligPrieb]

"Exceptions to the rule do not invalidate said rule. "

 

I didn't and wouldn't ever use the term "invalidate" here, because I agree that the statiistical approach can never be invalidated.

 

But what if the formulas predict W/L records for all 30 teams, and more than half of them is off by 3-4 games or more? That suggests they CAN be hit-or-miss at times. Soil your pants enough times, and you can't wear 'em out in public anymore. They're OK to wash the poodle in, or to weed your garden, but you now have to think about the situation you plan to utilize them in, and their usage becomes limited. Same goes for these 98% absolute formulaic equations, when applied to baseball. Enough errors in them suggests they're not so absolute, and shouldn't therefore be quoted with such a degree of certainty. 70%, 80% accurate, sure. But 98%? C'mon!

 

Which, once again, is why I'll never embrace the 10 runs created = 1 win formula, no matter how many times it's spelled out here by my friends on the other side of the ideological aisle.

 

We've all read argument posts like this one before:

 

"Why should we spend $5,000,000 a year to add 1.1 wins at SS, when we can bring in someone at 1/2 the price for nearly the same production, since those 1.1 wins aren't worth it!!"

 

*shudder* That was tough to even type!

[GenoSeligPrieb]

"The A-Rod, Bonds, and Jackson examples show the flaw in this. In some series they were brilliant, and in other series they were terrible. Did they have the requisite character all along, or was it only present when they didn't fail?"

 

Interesting point, Phenomenal. You're proving to be a solid debating opponent!

 

First of all, I unequivocally discount the statistics of EVERY version of Barry Bonds, after 1999, which of course includes the 2002 Rally Monkey World Series version 2.0. Artificial chemicals were so ingrained into his DNA by then, that a sizable chunk of him should be rebadged as "laboratory experiement." It was like sending Steve Austin, the $6,000,000 Man, or at least Hymie, Maxwell Smart's robot partner, out there to compete against his enemies.

 

To address your question, though, the Rodriguez and Reggie Jackson examples are spot-on.

 

While Alex Rodriguez's career is still being written, I believe history will remember good ol' #44 much more kindly than good ol' # 13, and your legacy speaks volumes more about what you did than your WARP ever will. If Rodriguez's career ended after 2007, with the Yankees failing to make the playoffs this season, and never winning a title with him at 3B, he'll be largely viewed as a guy who put up Hall of Fame-caliber numbers in the regular season (with an inordinate weighting of Devil Ray and Oriole pitching thrown in), but when it came to facing the very best, for all the marbles, with the pressure on, his career-defining moments waiting to be written, he kinda failed.

 

Remember the highlight of Steve Young when he finally won a Super Bowl, as the seconds wound down? He literally can be seen laughing, and "lifting" a huge weighty gorilla off his back. While I did play ball (but not surprisingly, not even at the collegaite level!), of course I can't speak for pro athletes. But wouldn't you agree that ballplayers feel their talents, their careers, are vindicated when they succeed in the playoffs?

 

John Elway, without a legit running threat until Terrell Davis came around after all those years, was perceived as another A-Rod. But when he won that first title, he flashed his big horse teeth, sighed with relief and exhaled.

 

Success at the highest levels, m'friend. That's what drives 'em.... I think!

[brettac1]

I think if you dominate over the course of 162 games that is more of an accomplishment than David Eckstein going 8 for 22 over a five game series. I think most baseball players would agree with that. There are amazing players who never even make the playoffs.

[Ennder]

I'm sure your right Geno, I still have arguments about football QB's that end with people just judging them by how many super bowl wins they had, like that has anything to do with how good a QB was.

 

The average fan is always going to judge players by their perceptions and not by reality, not much you can do about it.

 

but how come a significant number of these predictive models (Mariners, White Sox, Diamondbacks, Indians? Marlins?...are there more?) are proving inaccurate

 

Because they are estimators not predictors. If someone wanted to really predict what a teams record should be they would have to break the stats down much more than these simplistic estimators do. However they are close enough that if you are extremely far off your expected pace there is probably some luck involved.

[homer]

I dunno, Geno.....Ernie Banks never won jack and he's treated like royalty around these parts......granted, the Cubs themselves have never won jack either. http://forum.brewerfan.net/images/smilies/smile.gif

 

As far as Pythagorean formula goes (W/L predictor) it's like a weather forecast. If I get up in the morning and read there's a 70% chance of rain, I'm packing an umbrella. If it's dry all day I don't stop reading the weather page in the morning.

[GenoSeligPrieb]

"I think if you dominate over the course of 162 games that is more of an accomplishment than David Eckstein going 8 for 22 over a five game series. I think most baseball players would agree with that. There are amazing players who never even make the playoffs. "

 

Oh, no doubt. Ernie Banks would agree.

 

I'm not an absolutist. No one will ever get me to declare Eckstein a bigger success than Alex Rodriguez.

 

If I were to broadly state my case here, I'd surmise that, if you've participated in the playoffs a few times, your regular season stats should be weighted as approximately (VERY approximately, by the way) 70-75% (?) of your overall set of accomplishments, your legacy, if you will. Of course, if you reached the playoffs only once, in a 15-20 season career, then of course, you can bump the 162-game stats to perhaps a 90%+ mix.

 

But if you've been out there, in 2, 3, or more series, as Tim McCarver blathers on about you at 12:01 AM, in October games, and you consistently shrink like a wool sweater in the dryer, then your fellow players, managers, coaches, GMs, scouts, sportswriters, even screwballs like me, will start doubting whether you can thrive when the chips are down.

 

It's like an entirely different skill level, and I have to grudgingly admit that there ARE some guys like Lonnie Smith, David Eckstein, Craig Counsell...whose legacy IS improved when they keep rising to the occasion in the postseason. Is this enough to make them Hall of Famers? Hell no. Not even close. But it DOES make you cringe a bit when you have to face them in a win-or-go home situation, and that's a big feather in their caps.

 

Tell ya what, brett. I'd rather fill out my hypothetical playoff roster with a really good player, an occasional All-Star, whose stats show that he tends to just rake in the playoffs, (a Garrett Anderson in his prime?), over a guy with superior regular season stats and a case of the yips in the playoffs, like Alex Rodriguez.

 

All I'm sayin'...

[Ennder]

Hmm, I think anyone would be utterly insane to stick Garrett Anderson on their playoff roster over Alex Rodriguez, that one I simply cannot agree with, especially since before 2006 ARod had a better OPS in the playoffs than in the regular season! I can't throw a guy under the bus because of 29 bad AB's, just doesn't make any sense to me.

[GenoSeligPrieb]

Migod, at today's pace, I'm going to catch Russ with 13,000 posts by 9 PM!

 

But that's because we're having an intelligent, polite discussion...

 

"I'm sure your right Geno, I still have arguments about football QB's that end with people just judging them by how many super bowl wins they had, like that has anything to do with how good a QB was. "

 

It also helps a LOT if you were a QB from New York!

 

Hey, Joe Namath will be remembered, revered, as he walked toward the tunnel with that damned index finger of his up in the air, after winning 16-7. His career stats? Meh. Classic case of this theorem being overused. Good call.

 

"However they are close enough that if you are extremely far off your expected pace there is probably some luck involved.. "

 

Perhaps, but just HOW accurate? Al keeps saying 98%, which, I'm sorry, but looking at the Diamondbacks, White Sox, Mariners, the 2006 Brewers...that's enough to raise much more than a 2% scintilla of doubt. If you were to tell me you embrace OPS to an 80-90% extent, we've come pretty far!

 

"As far as Pythagorean formula goes (W/L predictor) it's like a weather forecast. If I get up in the morning and read there's a 70% chance of rain, I'm packing an umbrella. If it's dry all day I don't stop reading the weather page in the morning. "

 

Nice analogy, homer.

 

My assertion here is that the Pythagorean formula, 10 runs created = 1 win, is quoted as 98% accurate, which is a MUCH higher success rate than even Tom Skilling enjoys on WGN. Sure, you keep looking every day, but that grain of salt you take it with, is more than a 2% chunk! http://forum.brewerfan.net/images/smilies/wink.gif

 

I'm out! It's been fun...

 

Got to go home and have my wife rub some Bendryl on me in some strategic amd not-so-strategic places, after these damned itch mites we have this week in Chicago, turned my skin into a relief map of Marathon and Menomonee Counties. Ouch!

[Phenomenal Smith]

I know Geno is gone, but this is an interesting topic to me. With that said, here's how one player did in first 27 post-season games:

 

Batting average .188 OBP .248 SLG .327

 

Now here's how he did in his next 48 games:

Avg .329, OBP .425, SLG .529

 

The quotable player in question played for the Yankees, but he's not A-Rod. He was a teammate of Phil Rizzuto's.

[brettac1]

Gotta guess Mantle.

[dadofandrew]

On PS's example, anything can happen in 27 AB's, or even 100, about what you'd get in 27 games...that's why its meaningless.

[rluzinski]

Another writer making his ignorance of Sabermetrics painfully obvious while trying to disparage it. If any sport was 100% predictable they would cease to exist very quickly. And even the odds of winning every game were 100% predictable (which isn't possible), over the course of a season, a .500 team has about a 40% chance of winning over 86 games or under 76 games. Of course, guys who write stories like the above have no idea how that would even be possible.

 

Also, stat guys believe in clutch hitting but generally not clutch hitters. Certainly not clutch teams. Can a guy choke in certain situations, though? Sure.