slate column on why coaches punt

[Millhouse]

everything in the column comes from Thinking Fast and Slow (Kahneman) which anyone who is hung up on why coaches punt so much would probably like (even though it's not about football, it explains it)

http://www.slate.com/articles/sport...ain_why_nfl_coaches_punt_on_fourth_down_.html

the thing that he's not talking about in this article is that typically, people who learn about their irrationality, fix it. but most coaches choose to ignore it - I think that has a lot to do with signalling which tribe they belong to - the tough meathead tribe or the sniveling book reading rational twerp tribe. so much of their coaching ascent comes from conforming to the norms of overgrown concussed ogres, they're not willing to abandon that tribe

 

[longtimefan]

everything in the column comes from Thinking Fast and Slow (Kahneman) which anyone who is hung up on why coaches punt so much would probably like (even though it's not about football, it explains it)

http://www.slate.com/articles/sport...ain_why_nfl_coaches_punt_on_fourth_down_.html

the thing that he's not talking about in this article is that typically, people who learn about their irrationality, fix it. but most coaches choose to ignore it - I think that has a lot to do with signalling which tribe they belong to - the tough meathead tribe or the sniveling book reading rational twerp tribe. so much of their coaching ascent comes from conforming to the norms of overgrown concussed ogres, they're not willing to abandon that tribe

You're making the "Money Ball" argument. Why do managers, GM's and scouts behave the way they do? Because they have always done it that way. And no amount of rational thinking or statistical studies will change that.

 

[Scooch]

You're making the "Money Ball" argument. Why do managers, GM's and scouts behave the way they do? Because they have always done it that way. And no amount of rational thinking or statistical studies will change that.

I've had the opportunity to go to the MIT Sloan Sports Analytics Conference the past couple years (or as Bill Simmons affectionately calls it, Dorkapalooza) and it's remarkable to see the advanced statistical work done in baseball and basketball. Football is sooooo painfully far behind the times in that regard.

 

[Millhouse]

You're making the "Money Ball" argument. Why do managers, GM's and scouts behave the way they do? Because they have always done it that way. And no amount of rational thinking or statistical studies will change that.

my argument is a little different. it's partially habit but it's also a way to signify which club you belong to. depending on your team to stop someone is tougher than depending on your team to not be stopped. settling for 3 is the tough guy thing to do . we only need three , here's the ball, now we get to hurt you.

less violent games will have less irrational signalling like that.

the coaches who choose math over being a tough guy usually are pretty bad at teaching people to hit hard. look at the defenses of the teams with the smartest offenses.

marrone wants to throw a lot but he has to show his tough guy bonafides one way or another. stupid punts and 300 lb running backs get you there

 

[OttoinGrotto]

I've had the opportunity to go to the MIT Sloan Sports Analytics Conference the past couple years (or as Bill Simmons affectionately calls it, Dorkapalooza) and it's remarkable to see the advanced statistical work done in baseball and basketball. Football is sooooo painfully far behind the times in that regard.

I'm really jealous.

I despise how a lot of those stats are used for evaluation in baseball, but I'd love to be in attendance at that conference.

 

[Scooch]

I'm really jealous.

I despise how a lot of those stats are used for evaluation in baseball, but I'd love to be in attendance at that conference.

It's a good time, for total geeks like me. It's 90% male and you half-expect people to show up to sessions in Spock outfits.

My absolutely favorite session of all-time was the one where a dude statistically proved how it pays to cheat in college athletics.

 

[Chip]

It's a good time, for total geeks like me. It's 90% male and you half-expect people to show up to sessions in Spock outfits.

My absolutely favorite session of all-time was the one where a dude statistically proved how it pays to cheat in college athletics.

Bump that up to 95% and Millhouse is on the next flight.

 

[KaiserUEO]

Bump that up to 95% and Millhouse is on the next flight.

i figured the other 10% was more of question than female.

 

[Chip]

i figured the other 10% was more of question than female.

Gender/species. Tomayto/tomahyo.

 

[Millhouse]

i figured the other 10% was more of question than female.

chip likes conferences where each individual attendee is 50% male and 50% female.

 

[OttoinGrotto]

It's a good time, for total geeks like me. It's 90% male and you half-expect people to show up to sessions in Spock outfits.

My absolutely favorite session of all-time was the one where a dude statistically proved how it pays to cheat in college athletics.

I don't believe that would be all that hard a case to make actually.

 

[billsin01]

I'm really jealous.

I despise how a lot of those stats are used for evaluation in baseball, but I'd love to be in attendance at that conference.

Yeah, this is how I feel about advanced analytics. They are really interesting and have certainly changed the way most true baseball fans view the game. I'm still not sure I'm entirely sold on the true value of most of the stats, however, and I'm not sure we're that much better off with baseball fans memorizing OPS than batting averages.

 

[Scooch]

Yeah, this is how I feel about advanced analytics. They are really interesting and have certainly changed the way most true baseball fans view the game. I'm still not sure I'm entirely sold on the true value of most of the stats, however, and I'm not sure we're that much better off with baseball fans memorizing OPS than batting averages.

The key is to be balanced. You don't put 100% faith in some 65 year old scout's evaluation of a kid in high school, and you don't put 100% faith in some nerd's evaluation of a spreadsheet. Smart people use all available data and observation to make good decisions.

 

[OttoinGrotto]

The key is to be balanced. You don't put 100% faith in some 65 year old scout's evaluation of a kid in high school, and you don't put 100% faith in some nerd's evaluation of a spreadsheet. Smart people use all available data and observation to make good decisions.

That's reasonable, but not where my beef comes in.

I find a lot of sabremetricians, for being such good statisticians, to be really dumb about how they use their stats. There's a difference between using stats as indicators of future performance vs using them to assess past performance.

I hate when I see statements like "Player X isn't having as good a season as his .320 batting average indicates. His BABIP is way too high and is likely to regress as the season continues." Actually, you stathead freak, Player X IS having as good a season as his .320 batting average indicates. Know how I know? Because that's what actually happened. Projected stats shouldn't diminish real achievement. It's a gross misuse of the measure, and what really pisses me off is that it's entering into award recognition.

I understand the value they place on stripping out contextual metrics like RBI, but that also drives me insane because that context frames the entire way that the game is won. If they want to reduce a beautiful game to strat-o-matic baseball, that just misses the point to me.

/End rant

 

[liljoe]

chip likes conferences where each individual attendee is 50% male and 50% female.

I thought chip went 50% chicken and 50% donkey? I feel like there are other names for those animals, but I am drawing a complete blank...

 

[billsin01]

I understand the value they place on stripping out contextual metrics like RBI, but that also drives me insane because that context frames the entire way that the game is won. If they want to reduce a beautiful game to strat-o-matic baseball, that just misses the point to me.

/End rant

First, I agree on BABIP -- it is among the worst of those stats. I don't follow it that closely but I believe a lot of SABR folks agree with that. Really all it does is tell you how well a guy is making contact, which naturally will ebb and flow in a given year or career. BABIP is slightly more useful when context like LD rate and GB rate are factored in, but even then all you're really saying is, "This guy's hitting well b/c he's hitting a ton of line drives." Umm, yes. Thanks for all the hard work. I saw someone saying that Lester's BABIP against on Lester was significantly higher than at any other point in his career and suggested he'd be fine next year and was simply having stuff luck. Anyone who's watched lester pitch can tell you he's giving up more hits b/c he's giving up better contact.

Second, I also don't understand the call for demanding that people pay no attention to RBIs. I get that they are inflated at times simply based on where you hit and what lineup you're in, but there is still value in a guy who can put runs on the board when there are guys on base. There are perhaps better measures of this but I have yet to see a season in which a guy drives in 100 but is somehow not a good player or not helping his team at all.

 

[SmilinBob]

The key is to be balanced. You don't put 100% faith in some 65 year old scout's evaluation of a kid in high school, and you don't put 100% faith in some nerd's evaluation of a spreadsheet. Smart people use all available data and observation to make good decisions.

I agree and that's what good leaders do, they receive all the possible info (plus it has to be good information and correctly implemented) and use it to their benefit. Sometimes I think the only real saber football coaches do is the 2 pt play and when to do it and imo it isn't always right or it has flaws. When you mix brains and a football mentality with the ability to lead you end up with a Lombardi.

 

[Scooch]

First, I agree on BABIP -- it is among the worst of those stats. I don't follow it that closely but I believe a lot of SABR folks agree with that. Really all it does is tell you how well a guy is making contact, which naturally will ebb and flow in a given year or career. BABIP is slightly more useful when context like LD rate and GB rate are factored in, but even then all you're really saying is, "This guy's hitting well b/c he's hitting a ton of line drives." Umm, yes. Thanks for all the hard work. I saw someone saying that Lester's BABIP against on Lester was significantly higher than at any other point in his career and suggested he'd be fine next year and was simply having stuff luck. Anyone who's watched lester pitch can tell you he's giving up more hits b/c he's giving up better contact.

Second, I also don't understand the call for demanding that people pay no attention to RBIs. I get that they are inflated at times simply based on where you hit and what lineup you're in, but there is still value in a guy who can put runs on the board when there are guys on base. There are perhaps better measures of this but I have yet to see a season in which a guy drives in 100 but is somehow not a good player or not helping his team at all.

On RBI I tend to agree that it is a crude metric, but one that does have directional value. But if you want an example of how the old-school way of relying on it (and batting average) to evaluate performance was an epic disaster, I direct you to the 1986 Red Sox and Bill Buckner. John McNamara should have been arrested and jailed for batting him 3rd all year.

 

[SmilinBob]

On RBI I tend to agree that it is a crude metric, but one that does have directional value. But if you want an example of how the old-school way of relying on it (and batting average) to evaluate performance was an epic disaster, I direct you to the 1986 Red Sox and Bill Buckner. John McNamara should have been arrested and jailed for batting him 3rd all year.

Technically if you're really going saber batting orders don't matter. The numbers (in theory) prove that no matter what the lineup is, you're going to have the same results.

 

[Scooch]

Technically if you're really going saber batting orders don't matter. The numbers (in theory) prove that no matter what the lineup is, you're going to have the same results.

Well, they matter in the sense that guys at the top of the order end up getting 100 or so more plate appearances over the course of a season than guys at the bottom of the order.

So in my example any manager willingly wanting Buckner get to the plate more than Don Baylor and Dwight Evans should have his head examined.

 

[Bayside44]

Well, they matter in the sense that guys at the top of the order end up getting 100 or so more plate appearances over the course of a season than guys at the bottom of the order.

So in my example any manager willingly wanting Buckner get to the plate more than Don Baylor and Dwight Evans should have his head examined.

He was there for his glove.

 

[billsin01]

On RBI I tend to agree that it is a crude metric, but one that does have directional value. But if you want an example of how the old-school way of relying on it (and batting average) to evaluate performance was an epic disaster, I direct you to the 1986 Red Sox and Bill Buckner. John McNamara should have been arrested and jailed for batting him 3rd all year.

That is pretty sad. I think there is absolutely merit to new metrics, I just think, as many have posted, that it obviously behooves everyone to have a balanced approach.