Should newspapers employ a statistical analysis person?

[lcjjdnh]

The contortions of those who want sabermetrics to explain everything are far more entertaining than anything emanating from the Murray Chasses of the world who think they explain nothing. I particularly love how Billy Beane and the A's can't be used as evidence of sabermetrics' shortcomings, while Theo Epstein -- who has always had $150 million at his disposal -- can be used as evidence that it's the only good way to run a franchise.

Feel free to use the A's as evidence of sabermetrics' shortcomings. But please do so by making a well-reasoned argument. If all you're arguing is that the A's performance--measured in absolute terms--demonstrates the uselessness of defensive sabermetrics, there is a huge logical gap, as I pointed out. If not and you have a more detailed argument, I'd love to hear it.

They started to use defensive stats to find "undervalued" players and they turned into a 75-win team. Is that not evidence that it didn't work?

Did you even read my response? Do you understand correlation =/=causation?

By your "logic" the following is true: Jim has $15,000 buys a Kia Sorento. Pat has $350,000 and buys a Lamborghini. Jim knows a ton about racing and drives in a way that he feels best optimizes the strengths of his vehicle. Pat knows nothing about racing and, in Jim's mind, drives in a way that does not optimize the strength his vehicle. Pat beat Jim in a race. Therefore, Jim is a bad driver.

 

[LongTimeListener]

The contortions of those who want sabermetrics to explain everything are far more entertaining than anything emanating from the Murray Chasses of the world who think they explain nothing. I particularly love how Billy Beane and the A's can't be used as evidence of sabermetrics' shortcomings, while Theo Epstein -- who has always had $150 million at his disposal -- can be used as evidence that it's the only good way to run a franchise.

Feel free to use the A's as evidence of sabermetrics' shortcomings. But please do so by making a well-reasoned argument. If all you're arguing is that the A's performance--measured in absolute terms--demonstrates the uselessness of defensive sabermetrics, there is a huge logical gap, as I pointed out. If not and you have a more detailed argument, I'd love to hear it.

They started to use defensive stats to find "undervalued" players and they turned into a 75-win team. Is that not evidence that it didn't work?

That is actually not true. They didn't suddenly decide to use defensive metrics. They gradually added a defensive component to the metrics they were already using.

The reason they "turned into" a 75-win team is because they've had a bad run of young players not producing (or not producing before they could trade them, like CarGo) and the guys they have kept have been hurt all the time.

I'm always amused at the generalizations people make about the A's from afar.

Well, I figure I watch the A's as much as anyone on this board who isn't a beat writer, so I don't know that I'm afar. I am not in the press box soaking up the Aura Of Billy, I will grant you that. And there are a lot of reasons they are in the tank, the biggest of which is cheap-ass ownership that pockets baseball's welfare as profit because it's too easy to field a crap team and make money.

But I don't know what the line between sudden and gradual is to you. In 2006 they were a 93-win team using players who were not by and large defensively adept. The following year they were a 76-win team, and since then they have topped out at 81-81, and that has coincided with much writing locally and nationally that defense is the new undervalued commodity Billy is seeking.

 

[Van Lingle Mungo]

If statistics are used by a ballclub to evaluate talent, those statistics are part of the news and should be covered and of course, explained if possible (relatively few of the new numbers are too complex for the explanation to fit in a story).
BUT, just including statistical analysis articles in and of themselves is IMO a blunder. The NY Times, among other papers, runs a lot of those pieces, and most are attention seeking missiles which make exaggerated claims about a subject for the purpose of defending a particular statistic or statistics.
PS: If anyone had invented a dependable fielding metric, it'd be in universal use. It remains the White Whale of the horsehide quants.

But isn't there a third use in there? Shouldn't reporters and columnist being using the best tools available to analyze and convey to readers what's happening on the field. Reporters use stats all the time to assess and report on the performance of players on the teams they're covering. They need to know which ones are legitimate and which ones are not.

Yup. heck, there's no need to even talk about the stats themselves, if you think acronyms will lose readers.

No need to mention FIP. Just emphasize how a pitcher has a great ERA, but he's walking a lot of guys, which often spells trouble down the road.

No need to mention wOBA or WAR. All you have to do is talk about whether or not a batting average is empty, and it goes a long way towards those ends.

The valuable and generally accepted advanced stats are built off numbers everybody knows -- they just emphasize which elements play a bigger role in scoring or preventing runs.

Understanding what is weighted more heavily with those stats is all that's necessary. Throw in a couple other concepts -- run expectancy, to understand the break-even rate for stealing bases and the value (or lackthereof) of sacrifice bunts, and BABIP for streaks and slumps -- and you're most of the way there.

This amounts to a couple hours of reading. No journalist should pout over doing a couple hours of reading about the field they're covering. None of these things have to change the way anybody writes, but the quick analysis that's already being done (He's the team's best hitter! He's been a victim of bad luck!) will come from a much stronger base.

 

[RickStain]

Sabermetrics failing to account for defense is only one of quite a few hypotheses that could be supported by that evidence.

 

[LongTimeListener]

The contortions of those who want sabermetrics to explain everything are far more entertaining than anything emanating from the Murray Chasses of the world who think they explain nothing. I particularly love how Billy Beane and the A's can't be used as evidence of sabermetrics' shortcomings, while Theo Epstein -- who has always had $150 million at his disposal -- can be used as evidence that it's the only good way to run a franchise.

Feel free to use the A's as evidence of sabermetrics' shortcomings. But please do so by making a well-reasoned argument. If all you're arguing is that the A's performance--measured in absolute terms--demonstrates the uselessness of defensive sabermetrics, there is a huge logical gap, as I pointed out. If not and you have a more detailed argument, I'd love to hear it.

They started to use defensive stats to find "undervalued" players and they turned into a 75-win team. Is that not evidence that it didn't work?

Did you even read my response? Do you understand correlation =/=causation?

By your "logic" the following is true: Jim has $15,000 buys a Kia Sorento. Pat has $350,000 and buys a Lamborghini. Jim knows a ton about racing and drives in a way that he feels best optimizes the strengths of his vehicle. Pat knows nothing about racing and, in Jim's mind, drives in a way that does not optimize the strength his vehicle. Pat beat Jim in a race. Therefore, Jim is a bad driver.

Again, I am amused by your contortions. My point on this page was that defensive sabermetrics are not "better" simply because someone came up with a fancy name for them. And yes, Billy Beane has tried to play the defensive sabermetric game, and his team is not doing well, while other teams with similar payrolls have been very competitive in recent years.

Rick's point, that there isn't enough value in defense to make investing in it worthwhile, is a good one. But BB Bobcat noted that the stats themselves are of highly questionable value.

There is as much dogma from the statheads that any "advanced metric" is good as there is from the supposed dinosaurs that any "advanced metric" is useless.

 

[RickStain]

A huge problem with both measuring defense and building your team around it is consistency of chances.

Even over 162 games, the number of balls hit on the periphery of a defender's ability to make the play can vary quite a bit. You can try to decide that a great defensive CFer is worth $X million, but then he only gets 20 balls hit to center in a spot where he can make the play but another guy couldn't. You've wasted your money. Then the next year he has 60 of those balls and you look like a genius.

 

[BB Bobcat]

But I don't know what the line between sudden and gradual is to you. In 2006 they were a 93-win team using players who were not by and large defensively adept. The following year they were a 76-win team, and since then they have topped out at 81-81, and that has coincided with much writing locally and nationally that defense is the new undervalued commodity Billy is seeking.

Well, we're off on a full-blown threadjack now...

Just which players who magically appeared between 2006 and 2007 are your examples of the A's sudden emphasis on defense? Their starting eight was almost identical in the two years, except that Chavez was hurt a lot more in 2007 (oh wait, that would make the defense worse!)

To me, the biggest imported players who were hailed for their defensive contributions were Mark Kotsay and Jason Kendall, and they were both there in 2006 and 2007.

I think that losing Frank Thomas and Barry Zito (who used to be good) had a much bigger impact on their suckiness from 2007 going forward. I also think the manager change had an impact.

Anyway, I find it hard to believe you've really watched the A's closely and you still think their emphasis on defense is even in the top 10 reasons why they haven't won over the past few years.

 

[Small Town Guy]

Sidebar:

Where would sabermetrics be today if Moneyball hadn't been written?

And it wasn't inevitable that the book was going to be written. Swenk has talked here about how Lewis had to beg his publisher to let him write it, for less money, because no one was interested in the subject.

There had obviously been numerous advanced stats books before Moneyball. And there would have been many more even without Moneyball. But it seems doubtful any of those would have captured the moment like Moneyball did. It had to do with the writer, the narrative and the team.

So if there's no Moneyball, do the majority of teams still utilize the advanced stats? Would it have become as accepted among franchises or would it have taken a bit more time to get ingrained?

And in the media - obviously some don't have much use for them even today. Seems like they'd be even more of a fringe topic if Moneyball - rightly or wrongly - hadn't brought them to such a prominent spot.

And what about other sports? Advanced stats in basketball and football aren't as big of a deal as baseball, yet when those who do use them - or even those who don't like them rip on them - what is referenced? Baseball and Moneyball. No Moneyball, how does someone talk about their advanced stats in hoops? Yes, you could compare it to Bill James' writing. But as great as James is, he'd been writing for nearly 30 years without making the public impact that Moneyball did.

Just with hoops, perhaps no Moneyball would have meant no John Hollinger on ESPN or no Wages of Wins. I could have lived with that.

 

[RickStain]

Honestly, I don't think Moneyball made a lick of difference in the industry. Moneyball just made the rest of us talk about a trend that was already maturing quickly.

The reason you don't see it in other sports is that other sports don't lend themselves to it the way baseball does. It's almost impossible to pull out individual performance from his teammates when they are all working together. Baseball players work mostly alone.

 

[BB Bobcat]

Billy Beane and Theo both said that Moneyball accelerated a process that was underway, mostly because guys who own teams read Moneyball, and they hadn't been reading Bill James or guys like him.

You can also make an argument that Moneyball hurt the "statistical revolution" by making it out to be a revolution, with everyone picking sides, instead of a normal progression of ideas.

 

[LongTimeListener]

No, I think reason #1 is cheap ownership, and I think reason #2 is that Billy is (or was) so dogmatic about walks and not stealing bases that he ran guys like Carlos Gonzalez and Andre Ethier out of the organization. That would be a pretty nice outfield.

I do think it would fit somewhere in the top 10, though.

In 2007, Shannon Stewart was coming off consecutive seasons below .725 OPS -- hardly a Moneyball type of player, but someone who was once considered a good defensive outfielder. They didn't pay him a lot, so that was the primary reason he fit into their strategy, but I think he also is someone where they tried to buy on the cheap for defense. And since then they have continued to invest in players who are no great shakes offensively -- Cliff Pennington to name one -- and who wouldn't have been in their plans eight or nine years ago.

 

[JC]

No, I think reason #1 is cheap ownership, and I think reason #2 is that Billy is (or was) so dogmatic about walks and not stealing bases that he ran guys like Carlos Gonzalez and Andre Ethier out of the organization. That would be a pretty nice outfield.

I do think it would fit somewhere in the top 10, though.

In 2007, Shannon Stewart was coming off consecutive seasons below .725 OPS -- hardly a Moneyball type of player, but someone who was once considered a good defensive outfielder. They didn't pay him a lot, so that was the primary reason he fit into their strategy, but I think he also is someone where they tried to buy on the cheap for defense. And since then they have continued to invest in players who are no great shakes offensively -- Cliff Pennington to name one -- and who wouldn't have been in their plans eight or nine years ago.

When was Stewart ever considered a good outfielder? He may have had the worst arm I ever seen.