Does Anyone Else Here Think Sabermetrics is a Sham?

[DaMook]

i've always thought the best way to view postseason ball is like a tournament which is what it is -- start the tournament over the day after it ends and you probably get different results

Exactly! Thats a perfect way to look at it.

[DaMook]

I wouldn't say its entirely a crapshoot (i think if you replayed this postseason over 1,000 times the Padres would only win about once), but it is definitely true that the best team doesn't always win. I also think that even a full season doesn't necessarily give a good assesment of how good a team is. For example, i don't think the 2001 Mariners were as good as the 1998 Yankees, but they won two more games.

I think the Padres would win more than once. Its such a handful of games that ANYTHING and I mean ANYTHING can happen.

A full season also doesnt paint a perfect picture but its a much better way to measure a teams ability cause the sample is so large. Teams go through upps and downs. Things then to level out.

[Ubiquitous]

I wouldn't go so far as to say that they are a sham, but many people on this forum certainly overvalue them. Their lure is great. Like religion, anything that presents itself as a panacea is wonderful. Advanced metrics can offer answers to player comparisons, even for players whom the evaluators have never seen, but the conclusions are not set it stone and must be interpretted in context.

This first part can be a descriptor for anybody who blindly follows anything including those who blindly follow the RBI and Batting Average.

I have often stated that sabermetrics have no greater predictive value than any other stats because in terms of winning it all you reach the point where you must apply the macro to the micro. SABR focuses on statistical research, correlations that are proven over time. Any 5 or 7 game series can wholly contradict its premises without invalidating the study itself. We must be aware of the limitations of all stats and theories, SABR and classical.

I too are not sure what is classical. The second thing is I think you are confusing different metrics. Practically every stathead worth his salt will tell you that his formulas are not exact and that they don't apply the same to every single situation. Lots of them are based on averages or baselins of that nature. Looking at something like batting runs to see who should pinch hit for your pitcher while facing Randy Johnson is wrong. The metric isn't wrong the person using it in that way is wrong. SABR is much more then just win shares or creating fancy metrics. Its looking at the game in detail. Seeing how often a player can hit a fastball low and away. Where he hits it. How often that particular SS can get to that spot so on and so on. Most of the stuff that gets posted on this site and others is the armchair SABR stuff, because that is the stuff available to us as outsiders. Theres a whole lot more that we only get glimpses of.

One thing I don't like about the advanced metrics is that any yahoo with a Bill James book can look up win shares or some other stat and think that they can add a meaningful contribution to a profound discussion of several players and their relative prowess.

This is the old hatred of the new comer showing itself. Its the clicky nature of people. SABR is new so people who grew up with the old ways of discussing things look at SABR as perhaps a shortcut or not real.

SABR, IMO, is better suited to explain how what already happened, happened and not to predict what will happen in any given situation. SABR devalues the stolen base, and sure overall that's probably true, but ask Dave Roberts what he thinks of that. One instance that defies the theory can prove to be more meaningful than a boatload of instances that support it. When taken too far, SABR creates for its own sake and strays from its purpose. Those who further the research must always ask themselves questions, make sure the findings refer back to the game and understand the differences between discovery and invention.

Practically everything in the world is better at explaining what happened as compared to predicting. If predicting was easy then the future would not be unknown. Weather simulations can show you exactly what happened but it gets dicier when looking ahead. That doesn't invalidate the prediction or the attempt to look forward.

[SABR Matt]

Funny Ubiquitus should mention weather modelling. I'm studying to be a meteorologist...and I can tell you that as much as people love to make weathermen the butt of jokes and claim it's the easiest job on the planet because you can be wrong half the time and keep your job...meteorological models are the most complex, advanced, outstanding accomplishments ever generated by the human scientific community. They represent the van guarde in scientific progress...and they've saved millions of lives in the last thirty years since they began to be used by forecasting offices...particularly in the last 15 years since weather model outputs became widely disseminated and drastically improved in their accuracy in the 0-84 hour period.

In point of fact, we're about 90% accurate on forecasts for 48 hours out in this country now on the whole...that level of progress was not even remotely dreamed of when back in the 1920s when pioneering meteorologists first came up with the idea of numerical modelling (because back then, we didn't have the computers that would make it possible to apply quickly enough to matter).

Compared to how much time we've spent thinking about how to forecast the weather (about 300 years of gradual scientific advancement)...sabermetrics is barely out of its' infancy.

Its' critics here seem truly steeped in their arrogance to assume that what we see today in the field of sabermetrics is all we ever will see. It's akin to the folks who laughed at the idea of numerical weather prediction as late as World War II.

Give us time...scientific advancement takes years...decades...CENTURIES...less time now of course since this is the information age...but we've only been seriously as this since the mid 70s and making real progress since the late 80s...now especially...you have got to give us the chance to get closer...

Of course the critics/traditionalists aren't the only arrogant ones...sabermetrician after bloody sabermetrician has come out with a new thing and called it revolutionary...myself included...we want to be recognized for the work we're doing...we're looking for what everyone else is looking for in life...a sense of belonging...but I fear we have allowed ourselves to be too impatient...too willing to settle for less in order to push our names out htere...too unwilling to listen to other sabermetricians...and too unwilling to listen to the traditionalists' real criticisms of our approach...I've found that some of my own best work (now I'm not saying I'm anywhere near finished...so...don't take it that way) hass come from an effort to answer a legitimate complaint of a traditionalist...

[digglahhh]

Yeah, and I don't pay attention to weather forcasts either.

To address some comments:

I am speaking hyperbolically, about people who go overboard with the application of the metrics.

It is interesting that one of you guys stated that sabermetrics was really about analyzing the game in detail. Hmm, now isn't that blatant semantic framing? I mean, people analyzed the game in detail long before anybody knew what VORP was.

When I said ask Dave Roberts, I was referring to his SB against the Yanks last postseason that helped lead to the win that sparked the BoSox comeback.

I didn't direct my comment about anybody with a Bill James book to anybody, although I could see how it could be interpretted as a dig (this is how rappers get shot).

By classical, I was just referring to the traditional theories that are basically proverbial, good pitching beats good hitting, productive outs...

When I said that one instance defying saber could be more important than several that affirm it I just meant that sometimes the defining moment in a game is something that is not really valued by SABR, a stolen base, a game saving web gem by a first baseman, or for that matter a sweep of the Yanks by KC (how sweet was that?).

[SABR Matt]

diglahhh...your ignorance is showing...I'm sorry, but not listening to weather forecasts because sometimes we get it wrong due to the overwhelming complexity of the weather is just plain foolish.

[Sultan_1895-1948]

Baseball is too complex for any one "system" to be counted on for absolute finality. There are several layers to the game, the players, the numbers they put up, who they put them up against, where they put them up, and how those numbers compare with others who put theirs up.

Each successive under-layer contributes to the outcomes of the upcoming layers. Saber probably isn't meant to be the say all, end all, but is just meant to take the top layer results, and create a clearer picture of what they mean in the big picture. I think ?

[SABR Matt]

Well Sultan...you are absolutely correct that the game and its' results are layered...

Any approach you use to understand the game that is not layered will do what you described...it will clarify the surface and define what it really means. It won't have much predictive value.

That's been the problem with sabermetrics...the statheads develop methods that do a good job of explaining what was...and hope that they will be predictive.

It's slow going due to the emmense complexity, but like the first weather models, today's sabermetricians are starting to change how they approach it (myself included)...we're starting to look at the individual game as the first layer and get a better feel for what goes on to create single-game outcomes and how those single game outcomes translate into seasonal results. The model Randy and I are building presently is a completely iterative/layered approach that is based on the idea that EVERYTHING in the game is interrelated...we're refusing to take any statistics at face value...not even the formative ones because they all effect and are effected by the environment in which they occur...we're building systems of non-linear equations that we believe when optimized will give us the best possible idea of how the interrelated pieces of the game (league contexts, parks, team sides, strengths of schedule, teams' unique reactions to parks, strengths of league) fit together to explain in a completely context neutral way what has happened...we're hoping the neutrality will allow us to place everything on the same level and thereby drastically improve our ability to predict into the future.

Of course there will always be limits to how accurate projections can be (freak injuries, extended runs of good or bad luck...the rules of statistics still apply)...but we don't believe we're anywhere near that limit yet.

[digglahhh]

diglahhh...your ignorance is showing...I'm sorry, but not listening to weather forecasts because sometimes we get it wrong due to the overwhelming complexity of the weather is just plain foolish.

Quite the opposite my friend. Not valuing something, or rather two things, that you are personally invested in does not constitute ignorance. Assuming others should share the reverence you have for your own passions- that's ignorance, not to mention arrogance.

I don't pay attention to the weather because I simply do not care. I do not need their assistance. It ain't all that complex to me, its raining/snowing or its not, its cold, its hot? Do I wear shorts or pants, boots or sneakers? I can tell all this from sticking my head out the window. The jet stream, swelling coastal fronts- these things do not interest me.

You can call me many things, ignorant is not one of them.

The revolution wil not be rained out!

[SABR Matt]

I don't care if someone doesn't find interest in following the things I do...I labelled you as ignorant because you dismissed the pursuit of a better understanding of the weather and of baseball statistics out of hand. I recognize that most people aren't thrilled by the weather...I've called myself a nerd here many times without fear. But you dismiss the pursuit of sabermetrics as having no value here many times...then when I draw a comparison to another science that relies on similar methods but is much more advanced, and make the case that sabermetrics will improve with time in much the same way, you just say "I don't care about that"...

[digglahhh]

I did no such thing.

I do not dismiss sabermetrics, I question the amount of stock certain people put in them. I question if all those who espouse them understand how they are formed and are educated enough in the advanced mathematics involved to even have a valid opinion on their relevance. I question their applicability in predicting postseason success. This is not dismissing SABR, this is constructive criticism, it is pushing you guys to address and consider diverse perspectives. This reminds me of the conservative political pundits who claim that any time somebody critisizes America is un-American, when in truth critiquing the country is the most American use of free speech possible.

I don't dismiss the study of science, although, much it too is corporate guided and largely devoid of the altruistic and humane spirit that should guide the field. Anyway, I don't need the level of weather analysis that you guys provide.

[SABR Matt]

Several times in the threads here in the Sabermetrics Forum...you have said you believed sabermetric approaches were worthless in making predictions and best if only used to diagnose what has already happened...while there is certainly a reduction in the amount of information you can get once you go into the future, you overstate your case, and then claim you didn't say it.

Most people don't need the level of weather analysis available these days and that's fine. You don't need sabermetrics to watch a baseball game either...but if you're a GM running a baseball team...just as if you're running hundreds of oil derrecks in the Gulf or a large farm in Iowa...you need to see what's coming before it gets there...you need to be better informed...

[Ubiquitous]

It is interesting that one of you guys stated that sabermetrics was really about analyzing the game in detail. Hmm, now isn't that blatant semantic framing? I mean, people analyzed the game in detail long before anybody knew what VORP was.

There was sabremetrics long before Bill coined the word. Sabremetrics wasn't something that got invented overnight, it didn't go poof here I am. Sabremetrics isn't a philosophy, is Biology a philosophy? Is chemistry? Is Algebra?

[ssbguyincognito]

Listen people, as long as there are a substantial number of people (fans, GMs, and owners included) who don't believe in sabermetrics, this just gives sabermetricians like you and me so much extra time.

The human element of looking at the game is so arbitrary and fickle sometimes it amazes me. For example, there are people who say the Red Sox won the world series becaue Dave Roberts stole a base.

Never mind how stupid it is to attribute an entire 14 game, 3 series win streak to one event. Never mind sabermetrics has never said never steal bases, only that stealing bases as a long-term strategy is counter run productive, but there are situations in which it is good (just like the bottom of the 9th inning with a runner on 1st and no out and your fastest guy pinch running and you're down by 1, hmmm, sound familiar?) Never mind that the Sox still had to win 7 more games in a row after that. Never mind Ortiz won game 4 with a HR (not a stolen base) but nobody remembers that. Never mind that Dave Roberts barely played again in the postseason, somehow his presence on the bench must have helped them. Never mind Kevin Millar walked to leadoff the inning so Roberts could pinch run and steal second base (this is my favorite). There are people who will always, endlessly, stubbornly say the Red Sox won because of Dave Roberts.

Another thing that people say that's stupid: the White Sox are a small ball team. I don't know where people got this idea from. The White Sox are a team with a couple of good baserunners (only 1 or 2 actually). They are stacked with a bunch of slow HR hitters. And that's what they do: they hit the HR. They had 7 guys with 15+ HRs in the season, that's not small ball, that's bashing-the-hell-out-of-the-ball. I think they were 4th in the majors in HRs. Even Podsednik, their small-ball guy, hit 2 HRs in the postseason! Amazing how people will still insist the White Sox won because of small ball.

Sabermetrics is not a sham. All it is, is taking objective scientific principles, and applying them to baseball. And it turns out, when you use objective scientific principles on anything, you get a better understanding of it. And if you're a GM, you can use that understanding to try win games. And if there's anything sabermetrically inclined teams do in common, it's win games.

But still, there will always be people who attribute either the inherent randomness in the game or their own subjective element as proof that sabermetrics doesn't work.

[Brian McKenna]

i think the small ball stuff about the white sox came from the media during the postseason - to push their big enlightenment about how great the pitching rotation was and about the utter brilliance of ozzie guillen - they pushed the small ball stuff to excentuate the rotation's value and to magnify guillen's role in the team's destiny