Sabermetrics

[jim]

Baseball is a pretty simple game really. The irony is, the anti-Sabermetric crowd claims that the other guys are making the game to complicated. I view it the other way, fancy names and math and all they make the game simple again. Outs and runs production, pretty much everything boils down to that.

[Radbird]

I'm probably the best person here both at understanding and applying even the most difficult statistics.

It's a curse.

And you demonstrate it with every post you make, son. I'm so proud of you. The more stuff that's out there, the more we can pick and choose to make whatever case we're trying to make.

Seriously, I look at the stats and try to absorb what they (and the people putting them out there) are saying, recognizing that there is can be bias in the presentation and interpretation just like anything else.

[jim]

I'm probably the best person here both at understanding and applying even the most difficult statistics.

It's a curse.

And you demonstrate it with every post you make, son. I'm so proud of you. The more stuff that's out there, the more we can pick and choose to make whatever case we're trying to make.

Seriously, I look at the stats and try to absorb what they (and the people putting them out there) are saying, recognizing that there is can be bias in the presentation and interpretation just like anything else.

There can, in fact Bill James recognized the misuse of stats back when he started getting some attention. "So and so hit's .400 off of lefties on cool April nights with full moons etc...". Which sounds alot like descriptions I hear about the inner workings of the game given by an anti-Sabr person: "You gotta hit this guy 2nd and this guy 4th so that your 3rd guy can drive your 1st guy in after an infield single and then your 8th guy needs to do this and that so your 7th guy can twizzle bomb it for the 4th guy who really sets up your 3rd guy or otherwise it just all falls apart on you. Simple. And then we bring in our left out only guy to face the left in the 7th but not the 8th 'cuz everyone knows he can''t pitch in the 8th 'cuz that's what we hired the 8th inning guy for so he can get setup by our lefty loogy to get to our 9th inning guy. And then when the 9th inning guy comes in after we have our offense moving with our 7th guy twizzle bombing our 9th guy which sets up our 4th guy and showcasing our 3rd, then he can go out and get outs 25-27 which we all know are harder to get than the 3rd or 8th or 15th out. Right? I mean I know baseball is complicated but if you've studied it like I have all these years you know these things and if you don't know these things then write some of them there notecards up and that helps so you don't get your 6th inning one out guy to get the 8th place right hander out in the 5th or 7th instead of the 6th.".

Got it?


The SABR guy is more likely to just say to go get some good hitters and let them score lots of runs, and then get some good pitchers and prevent them.

[omaha_red]

i started this thread and now i have no interest in it

[Leroy]

Yeah, like in the EQh thread or whatever.

[docellis]

math hurts

[clement]

Tony said recently that Albert's numbers this year are even more impressive than they look given the kinds of pitches he sees. That's not just a protection argument; it's also about pitchers and managers realizing how tough Pujols is and being extra stingy with him regardless of who is batting behind him. But it's just common sense that if you have runners on base before him, and serious threats behind him, pitchers will be more constrained in their ability to pitch around Albert. Traditionalists call it protection, sabremetricians call it putting as many good hitters in a lineup as possible. Neither is wrong. While someone like Jeff Kent has a point about sample sizes, when the results of a small sample produces the expected outcome, there is nothing wrong with pointing it out.

I'd also say that both of the examples given where Tony Larussa is "wrong" are debatable, and therefore not wrong.

[jim]

Tony said recently that Albert's numbers this year are even more impressive than they look given the kinds of pitches he sees. That's not just a protection argument; it's also about pitchers and managers realizing how tough Pujols is and being extra stingy with him regardless of who is batting behind him. But it's just common sense that if you have runners on base before him, and serious threats behind him, pitchers will be more constrained in their ability to pitch around Albert. Traditionalists call it protection, sabremetricians call it putting as many good hitters in a lineup as possible. Neither is wrong. While someone like Jeff Kent has a point about sample sizes, when the results of a small sample produces the expected outcome, there is nothing wrong with pointing it out.

I'd also say that both of the examples given where Tony Larussa is "wrong" are debatable, and therefore not wrong.

I still say if you get the right guy in the 2 hole to protect AP in the 3 hole, and then get some guy with some thump in the 4 hole and get some little speed daemon in the 1 hole, and make sure that guy in the 2 hole has good bat control and doesn't K much, and back that up with a real slugger in the 5 spot who can protect the guy in the 4 spot who protects AP who drives in the guys in the 1 and 2 hole unless we decide to bunt with the 2 guy which doesn't make sense because the guy in the 1 hole has the speed to take the base on his own in which case we would want the 2 guy to swing away plus we really don't want our 3 hitter up with a base open so they can pitch around us and completely nullify what we are trying to do, then we hit the pitcher 8th so we get an extra leadoff hitter so that our 1 guy who really isn't fast could bunt him over so our 2 guy could drive him home.... oh man he better drive him home because if he doesn't we got our 3 guy coming up with a base open and oh my god they might walk him which causes the wheels to fall right off the cart, and then if we can figure out by looking at 10 AB's or so off the opposing pitchers and maybe shuffle our order up a little bit to take advantage of this guy owning that guy or vice a versa... well then I think we have a ballclub that can score some runs.

I'm old school though.

[clement]

Tony said recently that Albert's numbers this year are even more impressive than they look given the kinds of pitches he sees. That's not just a protection argument; it's also about pitchers and managers realizing how tough Pujols is and being extra stingy with him regardless of who is batting behind him. But it's just common sense that if you have runners on base before him, and serious threats behind him, pitchers will be more constrained in their ability to pitch around Albert. Traditionalists call it protection, sabremetricians call it putting as many good hitters in a lineup as possible. Neither is wrong. While someone like Jeff Kent has a point about sample sizes, when the results of a small sample produces the expected outcome, there is nothing wrong with pointing it out.

I'd also say that both of the examples given where Tony Larussa is "wrong" are debatable, and therefore not wrong.

I still say if you get the right guy in the 2 hole to protect AP in the 3 hole, and then get some guy with some thump in the 4 hole and get some little speed daemon in the 1 hole, and make sure that guy in the 2 hole has good bat control and doesn't K much, and back that up with a real slugger in the 5 spot who can protect the guy in the 4 spot who protects AP who drives in the guys in the 1 and 2 hole unless we decide to bunt with the 2 guy which doesn't make sense because the guy in the 1 hole has the speed to take the base on his own in which case we would want the 2 guy to swing away plus we really don't want our 3 hitter up with a base open so they can pitch around us and completely nullify what we are trying to do, then we hit the pitcher 8th so we get an extra leadoff hitter so that our 1 guy who really isn't fast could bunt him over so our 2 guy could drive him home.... oh man he better drive him home because if he doesn't we got our 3 guy coming up with a base open and oh my god they might walk him which causes the wheels to fall right off the cart, and then if we can figure out by looking at 10 AB's or so off the opposing pitchers and maybe shuffle our order up a little bit to take advantage of this guy owning that guy or vice a versa... well then I think we have a ballclub that can score some runs.

I'm old school though.

Well you said in about 400 words what I said in about 20. So yeah, I guess you can make things complicated if that's what you want to do.

I think most would agree that one of Larussa's philosophies as a manager is to get his role-players a lot of playing time. So the question becomes, which games should he get starts in for the Miles/Taguchis/Bartons/Spieizios etc? He can just do it randomly, or he can take into account things like past performance and maybe talk to the players and see if they feel they have a good read on a particular pitcher or not. And 75 at-bats is actually a pretty decent sample size. 5 not so much, but still it might be better than complete randomness. These are at-bats we're talking about not coin tosses.

[jim]

Tony said recently that Albert's numbers this year are even more impressive than they look given the kinds of pitches he sees. That's not just a protection argument; it's also about pitchers and managers realizing how tough Pujols is and being extra stingy with him regardless of who is batting behind him. But it's just common sense that if you have runners on base before him, and serious threats behind him, pitchers will be more constrained in their ability to pitch around Albert. Traditionalists call it protection, sabremetricians call it putting as many good hitters in a lineup as possible. Neither is wrong. While someone like Jeff Kent has a point about sample sizes, when the results of a small sample produces the expected outcome, there is nothing wrong with pointing it out.

I'd also say that both of the examples given where Tony Larussa is "wrong" are debatable, and therefore not wrong.

I still say if you get the right guy in the 2 hole to protect AP in the 3 hole, and then get some guy with some thump in the 4 hole and get some little speed daemon in the 1 hole, and make sure that guy in the 2 hole has good bat control and doesn't K much, and back that up with a real slugger in the 5 spot who can protect the guy in the 4 spot who protects AP who drives in the guys in the 1 and 2 hole unless we decide to bunt with the 2 guy which doesn't make sense because the guy in the 1 hole has the speed to take the base on his own in which case we would want the 2 guy to swing away plus we really don't want our 3 hitter up with a base open so they can pitch around us and completely nullify what we are trying to do, then we hit the pitcher 8th so we get an extra leadoff hitter so that our 1 guy who really isn't fast could bunt him over so our 2 guy could drive him home.... oh man he better drive him home because if he doesn't we got our 3 guy coming up with a base open and oh my god they might walk him which causes the wheels to fall right off the cart, and then if we can figure out by looking at 10 AB's or so off the opposing pitchers and maybe shuffle our order up a little bit to take advantage of this guy owning that guy or vice a versa... well then I think we have a ballclub that can score some runs.

I'm old school though.

Well you said in about 400 words what I said in about 20. So yeah, I guess you can make things complicated if that's what you want to do.

I think most would agree that one of Larussa's philosophies as a manager is to get his role-players a lot of playing time. So the question becomes, which games should he get starts in for the Miles/Taguchis/Bartons/Spieizios etc? He can just do it randomly, or he can take into account things like past performance and maybe talk to the players and see if they feel they have a good read on a particular pitcher or not. And 75 at-bats is actually a pretty decent sample size. 5 not so much, but still it might be better than complete randomness. These are at-bats we're talking about not coin tosses.

Ah, but they are coin tosses. Because he is using a purely statistical approach, based on purely statistical data and not observational data. Those are his words, I've heard them repeated enough to believe that there are times where he really is going simply off the data.

It is exceptionally rare for a player to have enough AB's to make an informed decision based simply on the results. A player with 1,500 random AB's you are going to have to exceed 60 to get to any sort of statistical meaning. More random AB's, add more.

You say the notecards are better than nothing, even if it's not accurate it essentially shuffles the deck and results in a positive outcome - that being players get their playing time in which has benefits. So here we are, people claiming that Tony (or other managers) don't need to resort to fancy math because of their infinite baseball wisdom - and yet it's the math nerds calling for the manager to just go old school and use his baseball instincts.

Basically, it seems to me that if you criticize a manager for using statistics incorrectly or you criticize a manager for not using data available to him - all the same. In the former case, there is some reason (excuse really) to justify it, and the latter you get the ol' mysticism of baseball explanation that only a very few and chosen have been able to figure out.

His notecards are a bad idea, he should throw them away and use his baseball judgement as if player X will be successful against player Y.