PDA

View Full Version : Moneyball and the stolen base



NJReds
10-16-2006, 11:16 AM
I just finished reading Moneyball. Facinating read. I give Beane credit for what he's done in Oakland.

My question to our top analytical minds on the board is about speed and stolen bases w/respect to the philosophy behind Moneyball.

Obviously there are studies and statistics that show that the stolen base is overrated unless the success rate is extremely high.

But in watching the A's and Tigers, I was wonder if the complete lack of fear of the stolen base has an adverse affect on the offense.

If Team A doesn't utilize the stolen base, and everyone knows that they loathe giving up outs on the basepaths, I would think the following:

- With a runner on first, even one with good speed. The pitcher doesn't have to worry about him taking off and thus can focus on the batter. Also, he doesn't have to worry about pitchouts.

- The defense can settle in at double-play depth and not "cheat" to cover 2nd base.

- The hitter is less likely to get a steady diet of fastballs.

I doubt it can be proved, but there has to be some benefit to a team's offense when the threat of the stolen base exists. Even if they give the green light that often.

GAC
10-16-2006, 11:27 AM
I just finished reading Moneyball. Facinating read. I give Beane credit for what he's done in Oakland.

My question to our top analytical minds on the board is about speed and stolen bases w/respect to the philosophy behind Moneyball.

Obviously there are studies and statistics that show that the stolen base is overrated unless the success rate is extremely high.

But in watching the A's and Tigers, I was wonder if the complete lack of fear of the stolen base has an adverse affect on the offense.

If Team A doesn't utilize the stolen base, and everyone knows that they loathe giving up outs on the basepaths, I would think the following:

- With a runner on first, even one with good speed. The pitcher doesn't have to worry about him taking off and thus can focus on the batter. Also, he doesn't have to worry about pitchouts.

- The defense can settle in at double-play depth and not "cheat" to cover 2nd base.

- The hitter is less likely to get a steady diet of fastballs.

I doubt it can be proved, but there has to be some benefit to a team's offense when the threat of the stolen base exists. Even if they give the green light that often.

Personally, I don't care if it can be proved or not. Just because numbers can be shown that it's success rate is low, does not mean that it should be forgotten or ignored. While it is obviously not for everyone, not everyone should be considering themselves a "base stealer", and it really depends on the situation, there are advantages to base stealing.

But watching guys like Brock, Henderson, Morgan, and Coleman, once they got on base, practice the "art", and the way they analyzed pitcher's movement, trying to steal signs, and disrupt defensive alignments, was a joy to watch IMO.

But these guys had a higher then normal percentage of success rate at stealing.

M2
10-16-2006, 11:29 AM
I've always looked at the metrics on stolen bases as a case for aggression on the basepaths, not against it.

Think about it, you need to make it roughly 70% of the time for the SB to be an effective weapon. Here's my question. In this era of no-arm catchers and pitchers with no discernable move to 1B, what kind of reasonable quick player can't steal bases at at least a 70% clip? Seems to me that if you can run, MLB has set up an open invitation for increased scoring. Most teams literally can't stop the running game.

And if you've got speed and are willing to use it, the steady diet of fastballs and taking the defense out of the DP setup will follow. I also think it causes pitchers to rush their deliveries, leading to a greater number of mistakes.

Anyway, I think the numbers behind stolen bases make a compelling case to put the speed of the modern athlete to better use. IMO, it feeds into the core principles of big offense. In fact, I think we saw it on display with the Reds during the first half of the season.

NJReds
10-16-2006, 11:38 AM
Most teams literally can't stop the running game.



Although it may be totally unrelated to their team's respective success, two catchers left in the playoffs are Pudge and Molina, who may be the best catchers in the games with respect to shutting down the oppositions running game.

RedsManRick
10-16-2006, 11:43 AM
Personally, I don't care if it can be proved or not. There are advantages to base stealing. Base stealing is an "art" in my opinion and does take study and skill. It's not for everyone, nor should it be attempted in a majority of the time by alot of ballplayers that "think" they are base stealers.

It really depends on the situation.

But watching guys like Brock, Henderson, Morgan, and Coleman, once they got on base, and the way they analyzed pitcher's movement, trying to steal signs, and disrupt defensive alignments, was a joy to watch IMO.

There are lots of things that requires copious skill and artfullness. That does not necessarily mean it contributes to winning baseball games. I'm not saying doesn't contribute to winning baseball games, that it doesn't actually take skill, or that it's not fun to watch. In fact, I agree with NJReds' premise that the true effects of speed (not just stolen bases per se') are nearly impossible to measure, and probably greater than sabermatricians currently give credit for. It's just that your points don't necessarily say anything about the value of speed in winning games.

It is a very slipperly slope though. Could we not say the same thing about lineup order, power, weather, park design (shadows/glare off boxes), etc. There are a ton of hidden effects in the game that will never be measurable, which is definitely a good thing. I think there was a mini-debate over the course of a yahoo fantasy article series debating the value to the game of small ball -- not in terms of win probabilty, but aesthetically and historically. Even if we could prove that speed was pointless and it behooved teams to build lineups of Jason Giambis and Adam Dunns, would the game not be worse off for it? It's an interesting conversation, regardless of what perspective you have.

I would also argue that the effect of speed is perhaps just as big on extra base advancement than it is speed. Reyes moving to third on LoDuca's soft single yesterday being a perfect example. The one thing I think the stats community has yet to really quantify appropriately is the value of runner advancement, be it giving credit to good baserunners, or extra credit to SLG% for the extra bases it allows runners to advance in addition to the batter.

wheels
10-16-2006, 12:54 PM
I know Steel's got a metric that he uses, but he doesn't seem to be posting lately.

Maybe he's reading this though.

This would be a good cue for his return.

gonelong
10-16-2006, 02:00 PM
Think about it, you need to make it roughly 70% of the time for the SB to be an effective weapon. Here's my question. In this era of no-arm catchers and pitchers with no discernable move to 1B, what kind of reasonable quick player can't steal bases at at least a 70% clip?

I posted something to this effect a year ago or so. If I am the Reds I take the alternate approach to everyone else in Baseball. I load up on SB and defense guys, (of course, I would have also built the largest field that baseball would allow, but I digress) they are cheaper than sluggers and somewhat undervalued in todays game.

I'd also contend (also said this a few times before) that 70% successful SB rate against Eric Milton (5.19 ERA) probably costs you runs, while a 70% successful SB rate against Arroyo (3.29 ERA) probably nets you runs.

While SBs are likely overvalued during the season, I think they are undervalued during the post-season where the pitchers ERAs tend to be lower.

GL

RANDY IN INDY
10-16-2006, 02:37 PM
Real speed on the bases and the running game disrupts the pitcher's flow and puts added pressure on the defense. It changes the way a catcher might call pitches to a particular hitter. It causes pitchers to lose focus and make pitches they might not usually make and rush to the plate. It takes infielders out of their regualr flow and causes them to hurry throws and sometimes look up to see where the runner is before fielding the ball not watching it into their gloves. It requires middle infielders to be moving and opens up holes in the defense. It causes outfielders to charge balls hard that they usually do not. It gives opposing managers ulcers. It puts everything on edge.

From where I sit, speed is a very disruptive force in a baseball game.

RedsManRick
10-16-2006, 03:12 PM
I think the problem becomes not in asking "Is speed valuable?" but rather, "How valuable is speed compared to X?" I think the standard answer is that power is way, more valuable than speed. Knowing nothing else, would you rather have a 40 HR guy or a 60 SB guy? Who would you pay more to?

Speed doesn't do much to get on you base (unless you're Ichiro) and doesn't advance other runners. However, as you suggest, perhaps it does have a postive effect on the performance of your teammates, in addition to it's outright benefits of extra bases for the runner. The problem is measuring these benefits and properly weighting them. In the absence of thorough analysis, I think people are prone to a perception bias with speed, giving undue weight to the positive events it causes, largely because of its "novelty". People want speed to make a big difference, so in the absence of evidence to the contrary, they assume it does.

IIRC, there was a study posted on BP which showed that having a speedy player on 1B (compared to a not-speedy one) did not have a statistically significant effect on the performance of the batter. That, of course, is just one of the suggested postive effects, but take it for what you will. Perhaps somebody has the actual reference handy.

OldRightHander
10-16-2006, 03:37 PM
they are cheaper than sluggers and somewhat undervalued in todays game.


Wasn't that really the whole crux of Moneyball, to get more bang for your buck by utilizing what is undervalued by everyone else? I think a lot of people have attached Moneyball to sluggers who walk and strike out a lot, but I think the main point is more about going against trends to get more value for what you spend, or to avoid spending too much in areas that are overvalued.

RANDY IN INDY
10-16-2006, 03:47 PM
I think the problem becomes not in asking "Is speed valuable?" but rather, "How valuable is speed compared to X?" I think the standard answer is that power is way, more valuable than speed. Knowing nothing else, would you rather have a 40 HR guy or a 60 SB guy? Who would you pay more to?

Speed doesn't do much to get on you base (unless you're Ichiro) and doesn't advance other runners. However, as you suggest, perhaps it does have a postive effect on the performance of your teammates, in addition to it's outright benefits of extra bases for the runner. The problem is measuring these benefits and properly weighting them. In the absence of thorough analysis, I think people are prone to a perception bias with speed, giving undue weight to the positive events it causes, largely because of its "novelty". People want speed to make a big difference, so in the absence of evidence to the contrary, they assume it does.

IIRC, there was a study posted on BP which showed that having a speedy player on 1B (compared to a not-speedy one) did not have a statistically significant effect on the performance of the batter. That, of course, is just one of the suggested postive effects, but take it for what you will. Perhaps somebody has the actual reference handy.

Why do you have to choose between the two? I'd like a little of both. The key is balance. Get too overloaded either way and you are going to have problems.

RFS62
10-16-2006, 04:03 PM
Wasn't that really the whole crux of Moneyball, to get more bang for your buck by utilizing what is undervalued by everyone else? I think a lot of people have attached Moneyball to sluggers who walk and strike out a lot, but I think the main point is more about going against trends to get more value for what you spend, or to avoid spending too much in areas that are overvalued.



Exactly right.

RedsManRick
10-16-2006, 04:06 PM
Why do you have to choose between the two? I'd like a little of both. The key is balance. Get too overloaded either way and you are going to have problems.

Well you don't have to choose one or the other to be sure. Even the Red Sox needed a stolen base to get to the World Series. But there's a whole of grey area to be negotiated in between. And given fixed resources, if you spent 10% of your payroll on a Juan Pierre, that's gonna have a significant effect on your ability to sign other players with other skill sets. It's a question of the right distribution of talent and money.

RANDY IN INDY
10-16-2006, 04:10 PM
Balance.

vaticanplum
10-16-2006, 04:11 PM
Wasn't that really the whole crux of Moneyball, to get more bang for your buck by utilizing what is undervalued by everyone else? I think a lot of people have attached Moneyball to sluggers who walk and strike out a lot, but I think the main point is more about going against trends to get more value for what you spend, or to avoid spending too much in areas that are overvalued.

That's how I see it. The story of Beane and the A's is a more of a business one than a baseball one to me, though certainly I learned a ton about baseball from it and I think it sparked a new way of looking at baseball for a lot of people. The market of baseball now has disparities between individual teams that did not always exist, and management of the teams should adjust as such, which may sometimes utilize new ways of looking at the sport itself. This is what Beane did, to good success, when other managers were refusing to do so.

It's a how-to manual for management, not a how-to manual for baseball. Though the baseball info, taken in accordance with a lot of other things, could help a lot of teams.

GAC
10-16-2006, 04:45 PM
Why do you have to choose between the two? I'd like a little of both. The key is balance. Get too overloaded either way and you are going to have problems.

And there you go. :thumbup:

And while most think that speed and base stealing go hand-in-hand.... and it does for the most part.... not all really good base stealers were the "fastest men on the planet" (or team). It does take smarts, along with the ability.

NJReds
10-16-2006, 04:46 PM
The story of Beane and the A's is a more of a business one than a baseball one to me, though certainly I learned a ton about baseball from it and I think it sparked a new way of looking at baseball for a lot of people.


It was a business story...but Beane took a lot of what Bill James had been saying for years about the misconception of baseball stats and put that into practice. It was a tremendous risk, but he was in a low-risk, high-reward situation.

Ultimately, more and more baseball decisions were made based on this information -- the A's success may have driven the price up on certain types of players that used to be cheap. But that doesn't mean that those statistics are any less important when it comes to building winning baseball teams. Or that you'll be able to find some commodity on the cheap that will help you win baseball games in other ways.

I come here not to praise Beane, or to bury him. His story is interesting. From the business side, it's a true example of "thinking outside the box."
However some of his beliefs (which I think are deep seeded regrets to how his development as a player progressed and ultimately ended) also hinder his ability in some cases.

When looking at the draft, it seems that he goes out of his way to make some kind of odd splash with the "old timers" just to make a point. Where is Jeremy Brown these days? Why was Bonderman such a bad pick?
His insistance on harping on his managers' baseball decisions is also somewhat disturbing. If he wants to manage so bad, maybe he should move down to the dugout. That's why Howe was bounced...it's probably in the background of Macha's firing today.

There's a lot to digest -- he's a complex figure.

vaticanplum
10-16-2006, 04:52 PM
Ultimately, more and more baseball decisions were made based on this information -- the A's success may have driven the price up on certain types of players that used to be cheap. But that doesn't mean that those statistics are any less important when it comes to building winning baseball teams. Or that you'll be able to find some commodity on the cheap that will help you win baseball games in other ways.

I agree with all of that. I just think that sometimes people look at it as a how-to baseball manual -- this is good, this is bad; this is the right way to play baseball, this is the wrong way. I think that's a gross misinterpretation of the main point of the book.

NJReds
10-16-2006, 04:57 PM
I agree with all of that. I just think that sometimes people look at it as a how-to baseball manual -- this is good, this is bad; this is the right way to play baseball, this is the wrong way. I think that's a gross misinterpretation of the main point of the book.

True. Although even the author admits that what he started out with was an idea to interview different people around baseball -- but the "Billy Beane" story eventually took over. I imagine that he changed his mind 5 minutes after a riviting Dan O'Brien interview.

From there it became a "how-to" guide to win with a small market team. Billy took a lot of heat for it, which is too bad. Because a lot of folks missed the point.

SteelSD
10-16-2006, 07:05 PM
I know Steel's got a metric that he uses, but he doesn't seem to be posting lately.

Maybe he's reading this though.

This would be a good cue for his return.

You rang?

Anyway, I have Speed Adjusted OPS (SAOPS) that I use when I'm looking to do a quick hit comparison between different player types. But that's probably not going to tell us what we're looking for (i.e. a measurement of "hidden" value).

Stolen Bases are interesting though. The threat of one can change pitch selection, defensive positioning, and create brain noise for the pitcher. On the flip side, they also create distraction for the hitter who now has to ignore a secondary moving object crossing his field of vision. And it's not uncommon that a Stolen Base attempt ends up wasting a Strike if a hitter is trying to protect the runner. Subjectively, there are some potential positives. But we also see associated negatives. In the end, it's possible that it's a wash; meaning that a Stolen Base might be actually only as valuable as the additional base (or out) it represents.

I'd actually like to see testing done with players to determine who might be able to best filter out that kind of distraction. Might be a good idea especially for teams that like to run a lot in order to avoid putting more easily distracted hitters behind the rabbits.

I've always been a big fan of functional speed because extra bases are good. In my mind, I like it when a team that can steal does...but not too often. Just enough in high percentage situations (right runner, right catcher, right time) that certain players become a threat because it's the threat that keeps defenses thinking.

And the interesting thing is that this topic was posted with the A's/Tigers series as the catalyst. One of those two teams stole more bases than the other and one of those teams posted the worst success rate in Major League Baseball this season.

Anyone know which is which?

pahster
10-16-2006, 07:21 PM
Oakland stole 61 and was caught 20 times - 75%
Detroit stole 60 and was caught 40 times - 60%

gonelong
10-16-2006, 07:35 PM
Wasn't that really the whole crux of Moneyball, to get more bang for your buck by utilizing what is undervalued by everyone else?

Yup. I always wonder how someone could read that book (or professed to have) and come to some of the conclusions that they have.

Its an age old truth. When you are on a budget you stretch your dollar as far as it can go and you darn well don't waste any. The way to do that is to know what everything is worth in time and/or money.

GL

mth123
10-16-2006, 08:50 PM
You rang?

Anyway, I have Speed Adjusted OPS (SAOPS) that I use when I'm looking to do a quick hit comparison between different player types. But that's probably not going to tell us what we're looking for (i.e. a measurement of "hidden" value).

Stolen Bases are interesting though. The threat of one can change pitch selection, defensive positioning, and create brain noise for the pitcher. On the flip side, they also create distraction for the hitter who now has to ignore a secondary moving object crossing his field of vision. And it's not uncommon that a Stolen Base attempt ends up wasting a Strike if a hitter is trying to protect the runner. Subjectively, there are some potential positives. But we also see associated negatives. In the end, it's possible that it's a wash; meaning that a Stolen Base might be actually only as valuable as the additional base (or out) it represents.

I'd actually like to see testing done with players to determine who might be able to best filter out that kind of distraction. Might be a good idea especially for teams that like to run a lot in order to avoid putting more easily distracted hitters behind the rabbits.

I've always been a big fan of functional speed because extra bases are good. In my mind, I like it when a team that can steal does...but not too often. Just enough in high percentage situations (right runner, right catcher, right time) that certain players become a threat because it's the threat that keeps defenses thinking.

And the interesting thing is that this topic was posted with the A's/Tigers series as the catalyst. One of those two teams stole more bases than the other and one of those teams posted the worst success rate in Major League Baseball this season.

Anyone know which is which?

I threw this out about a month ago on Redslive. I'm not a stat guru so my approach probably wasn't any good. I'd be interested in Speed Adjusted OPS. How is it calculated? Does it just account for Steals and Caught stealing? Does it take into effect guys getting picked-off or thrown out on the bases? Some guys speed may even hurt their value because they take so many chances. I understand the common wisdom says a 70% success rate, but a guy who gets picked-off a lot would need to be more successful I would think. The reds best steal guy makes a lot of outs on the bases that probably don't count against him in many formulas.

Some one mentioned Brock, Henderson, Morgan, etc earlier. Those guys took over games with speed and it was obviously valuable. That seems a lot different than the run of the mill 30 steal guy. I'm not sure how much they really add. If stealing bases is the primary skill, I don't think its that attractive unless its a guy who regularly changes games with it like the big names mentioned earlier. BTW Morgan, Brock and Henderson were obviously fine hitters w/o the steals.

RedsManRick
10-16-2006, 09:33 PM
I threw this out about a month ago on Redslive. I'm not a stat guru so my approach probably wasn't any good. I'd be interested in Speed Adjusted OPS. How is it calculated? Does it just account for Steals and Caught stealing? Does it take into effect guys getting picked-off or thrown out on the bases? Some guys speed may even hurt their value because they take so many chances. I understand the common wisdom says a 70% success rate, but a guy who gets picked-off a lot would need to be more successful I would think. The reds best steal guy makes a lot of outs on the bases that probably don't count against him in many formulas.

Some one mentioned Brock, Henderson, Morgan, etc earlier. Those guys took over games with speed and it was obviously valuable. That seems a lot different than the run of the mill 30 steal guy. I'm not sure how much they really add. If stealing bases is the primary skill, I don't think its that attractive unless its a guy who regularly changes games with it like the big names mentioned earlier. BTW Morgan, Brock and Henderson were obviously fine hitters w/o the steals.

Well, in terms of just SB and CS, figure it this way. 7 SB and 3 CS is break even, which means that every 3 CS negate 7 SB (or 1 CS negates 2.333 SB). Let's look at some of those guys best (most) SB seasons:

Henderson '82, 130(1) - 42(2.333) = 32 net SB
Morgan '75, 60(1) - 10(2.333) = 37 net SB
Brock '74, 118(1) - 33(2.333) = 41 net SB

Not saying these are these guys best "net SB" seasons, but consider how SB don't really scale the same way as Homers do. You might have twice as many SB as somebody else, but not have helped your team any more. Homers are always good and there's not bad a side. Speed is risky, and often the negative aspect of the SB is ignored, or at least brushed aside.

Joe Morgan, who stole "only" 689 SB in his career to Brock's 938, had over 100 more "net SB" in his career.

mth123
10-16-2006, 09:38 PM
Well, in terms of just SB and CS, figure it this way. 7 SB and 3 CS is break even, which means that every 3 CS negate 7 SB (or 1 CS negates 2.333 SB). Let's look at some of those guys best (most) SB seasons:

Henderson '82, 130(1) - 42(2.333) = 32 net SB
Morgan '75, 60(1) - 10(2.333) = 37 net SB
Brock '74, 118(1) - 33(2.333) = 41 net SB

Not saying these are these guys best "net SB" seasons, but consider how SB don't really scale the same way as Homers do. You might have twice as many SB as somebody else, but not have helped your team any more. Homers are always good and there's not bad a side. Speed is risky, and often the negative aspect of the SB is ignored, or at least brushed aside.

Joe Morgan, who stole "only" 689 SB in his career to Brock's 938, had over 100 more "net SB" in his career.

Thanks. That's pretty simple math, but what do Joe Morgan's net 37 Sb really mean in terms of adding runs?

cincinnati chili
10-16-2006, 10:09 PM
I've always looked at the metrics on stolen bases as a case for aggression on the basepaths, not against it.

Think about it, you need to make it roughly 70% of the time for the SB to be an effective weapon. Here's my question. In this era of no-arm catchers and pitchers with no discernable move to 1B, what kind of reasonable quick player can't steal bases at at least a 70% clip? Seems to me that if you can run, MLB has set up an open invitation for increased scoring. Most teams literally can't stop the running game.




I would love to see data on this (e.g. stopwatches in 1981 and 2006), but I would guess that pitchers in general are MUCH quicker to the plate now than they were 20-30 years ago. I can sum it up in two words:

SLIDE STEP. What was a balk 20 years ago, is allowed today.

Human beings are getting faster. You'd think that in the last couple decades one or two guys would come around who could steal the bases that Rickey Henderson or even Eric Davis could steal. I think there are external factors.

Bill James wrote about how the fluctuation of the stolen base is cyclical. As emphasis decreases, teams start trying to sneak in more good-hit-bad-throw catchers. Then the stolen base comes back. Then catchers get better, and it gets to be a bad bargain.

My untested theory is that we'll stay stuck in this gear of low-sb's until baseball enacts some rule changes/enforcement changes.

Cyclone792
10-16-2006, 10:11 PM
Thanks. That's pretty simple math, but what do Joe Morgan's net 37 Sb really mean in terms of adding runs?

One "net SB" is worth about 0.20 runs, which means that 37 "net SB" will net you 7.4 runs.

Dan Fox at The Hardball Times (http://www.hardballtimes.com/main/article/whitey-ball/) wrote a pretty good article last spring about the effect of stolen bases and its run value for the 1985 Cardinals.

IslandRed
10-16-2006, 10:24 PM
My untested theory is that we'll stay stuck in this gear of low-sb's until baseball enacts some rule changes/enforcement changes.

You're right. It's really just a function of the run-scoring environment. As scoring goes up, the relative value of a base to an out goes down. Mostly, I think there's a bit of an equilibrium at the moment where the break-even point and the real-world success rate are very close together, which makes the running game more of a tactical consideration than a philosophical one. It is not obviously in a team's favor to run, and it's not obviously better to not run, so it's about picking spots.

OldRightHander
10-16-2006, 10:25 PM
Human beings are getting faster.

Except Griffey and me. We've both lost a step or two over the years.

vaticanplum
10-16-2006, 10:33 PM
SLIDE STEP. What was a balk 20 years ago, is allowed today.

Ok, I'm not educated enough to understand this just from this sentence, but I'm dying to. Can anyone elaborate?

The Fox commentators were talking about pitchers/catchers and the stolen base with the shockingly articulate AJ Pierzynski last night, and it was fascinating. Apparently Molina has a 1.6 second release time, which is extraordinary (Pierzynski made the joke that he aims for 2.6 seconds, but evidently 2.0 seconds and under is considered good). But the general concensus was that most of the reponsibility for that lies on the pitcher, who ideally should have a 1.4 second or less release time.

Obviously I realized the details of this, but I never really thought about it in terms of time. It was also interesting how adamant they were that this is the pitcher's responsibility, since I've always thought of it as more the catcher's, and since I have such an unfortunate love of high leg kicks in pitchers.

mth123
10-16-2006, 10:43 PM
One "net SB" is worth about 0.20 runs, which means that 37 "net SB" will net you 7.4 runs.

Dan Fox at The Hardball Times (http://www.hardballtimes.com/main/article/whitey-ball/) wrote a pretty good article last spring about the effect of stolen bases and its run value for the 1985 Cardinals.

Thanks for the link. This means that a guy who steals 30 bases is only worth 6 runs all year in comparison to some one equal in all other areas who doesn't steal any. This assumes he is never caught and that his hidden negatives (Times Picked Off which don't count as CS) aren't even factored in. Doesn't seem worth paying anything for to me. A couple 3 run bombs takes care of it pretty quickly. Heck 6 "rally killers" do the trick as well.

I still want speed on the bases though. I just don't want to pay $ or talent to get the steal statistic.

blumj
10-16-2006, 10:48 PM
It was also interesting how adamant they were that this is the pitcher's responsibility, since I've always thought of it as more the catcher's, and since I have such an unfortunate love of high leg kicks in pitchers.
I'll give you a really easy example of this. There were 5 bases stolen off Bronson Arroyo this season, with 5 caught stealing. In his 2 full seasons with the Red Sox, with a totally different catcher, Varitek, he averaged 5 SB, 4 CS. No real difference, right? Now, Varitek also caught Derek Lowe regularly for 3 seasons, during which time an average of 23 bases were stolen off him with 4 caught stealing. Same catcher, different pitcher, huge difference. And you still get the high leg kick, just not from the stretch.

OldRightHander
10-16-2006, 11:42 PM
Personally I've always thought speed was more of a benefit when the ball is put in play, like if you have a guy who can score from first on a double, go first to third on a single, or have a reasonably high chance of scoring from second on a single. I'd also rather a contact hitter be fast to at least avoid a fiew of those annoying double plays. If you're a turtle, you'd better be mashing the ball.

SteelSD
10-17-2006, 01:14 AM
I threw this out about a month ago on Redslive. I'm not a stat guru so my approach probably wasn't any good. I'd be interested in Speed Adjusted OPS. How is it calculated? Does it just account for Steals and Caught stealing? Does it take into effect guys getting picked-off or thrown out on the bases? Some guys speed may even hurt their value because they take so many chances. I understand the common wisdom says a 70% success rate, but a guy who gets picked-off a lot would need to be more successful I would think. The reds best steal guy makes a lot of outs on the bases that probably don't count against him in many formulas.

The formula for each component:

SAOBP= (Hits+BB+HBP-CS)/TPA
SASLG= (Total Bases+SB-CS)/AB

The reason CS is subtracted from both is that a Caught Stealing is the erasure of a Base and it's the addition of an Out. But there are a couple of issues. Pickoffs don't count as a CS unless the runner actually breaks for the next bag. Without knowing that, we're left with less than optimal data. Secondly, some CS events (steals of 3B and Home) erase more than one base. Because we're missing those two data pieces, the "Speed Adjusted" metrics correlate with actual Runs Scored at a rate lower than the actual OBP, SLG, and OPS metrics. But they still correlate at very high rates. Here are the correlations for the 2006 NL:

OBP to RS/Game: 74.3%
SAOBP to RS/Game: 73.1%

SLG to RS/Game: 88.6%
SASLG to RS/Game: 83.1%

OPS to RS/Game: 94.7%
SAOPS to RS/Game: 90.1%

Now, those small correlation gaps may be narrowed with the addition of pickoffs and actual bases lost. That being said, even if we had the missing information, I'm not sure we'd see the SA metrics correlate at rates higher than their contemporary counterparts. So let's cross-check against Equivalent Runs (EQR). Basically, EQR is a measurement of how many Runs a team should have scored given their offensive performance.

OBP to EQR/Game: 87.3%
SAOBP to EQR/Game: 84.8%

SLG to EQR/Game: 82.3%
SASLG to EQR/Game: 85.0%

OPS to EQR/Game: 95.5%
SAOPS to EQR/Game: 96.4%

Well, that's interesting. SAOPS correlates less with actual results in each case but when we use EQR, SAOPS comes out on top in the two most highly correlated metrics (the SLG and OPS segments). Now, we need to remember that's nothing but suggestive as our study is simply one season of one league, but maybe there's more bite than I originally anticipated- i.e. that we might expect players with higher SAOPS numbers to project better than if using just OPS. Again, "might" is the key word there. Insert pickoffs and actual bases lost, and it's also possible that the correlations to EQR increase to near-perfect. That would be pretty cool.

BCubb2003
10-17-2006, 01:56 AM
With a base stealer on base and a hitter at the plate, how does the base stealer's ability to acquire a base compare with the batter's on-base percentage?

Obviously base stealing relies on getting on base to begin with, but getting past that for a moment...

If the batter gets a walk, moving the runner to second, that's good OBP.

If the runner steals second, moving himself over, is that as a good as the hitter's OBP moment would have been?

If the hitter then gets a walk to the open base, is his OBP moment not as good? It did avoid an out, but was not the mover of the runner. It did provide another runner for someone else to move over, though. Should a runner who steals a base get credit for on-base percentage? Slugging percentage? I know that's probably why bases acquired was invented.

How good a base stealer does a runner have to be to equate to a good OBP batter?

How bad does a batter's OBP have to be to make the runner's stealing a base a good risk?

Jpup
10-17-2006, 04:52 AM
It's the pitching.

Tigers LCS ERA 2.25 WHIP 1.19 K9 7.00
A's LCS ERA 5.71 WHIP 1.67 K9 6.49

It's almost always the pitching.

mth123
10-17-2006, 06:41 AM
The formula for each component:

SAOBP= (Hits+BB+HBP-CS)/TPA
SASLG= (Total Bases+SB-CS)/AB

The reason CS is subtracted from both is that a Caught Stealing is the erasure of a Base and it's the addition of an Out. But there are a couple of issues. Pickoffs don't count as a CS unless the runner actually breaks for the next bag. Without knowing that, we're left with less than optimal data. Secondly, some CS events (steals of 3B and Home) erase more than one base. Because we're missing those two data pieces, the "Speed Adjusted" metrics correlate with actual Runs Scored at a rate lower than the actual OBP, SLG, and OPS metrics. But they still correlate at very high rates. Here are the correlations for the 2006 NL:

OBP to RS/Game: 74.3%
SAOBP to RS/Game: 73.1%

SLG to RS/Game: 88.6%
SASLG to RS/Game: 83.1%

OPS to RS/Game: 94.7%
SAOPS to RS/Game: 90.1%

Now, those small correlation gaps may be narrowed with the addition of pickoffs and actual bases lost. That being said, even if we had the missing information, I'm not sure we'd see the SA metrics correlate at rates higher than their contemporary counterparts. So let's cross-check against Equivalent Runs (EQR). Basically, EQR is a measurement of how many Runs a team should have scored given their offensive performance.

OBP to EQR/Game: 87.3%
SAOBP to EQR/Game: 84.8%

SLG to EQR/Game: 82.3%
SASLG to EQR/Game: 85.0%

OPS to EQR/Game: 95.5%
SAOPS to EQR/Game: 96.4%

Well, that's interesting. SAOPS correlates less with actual results in each case but when we use EQR, SAOPS comes out on top in the two most highly correlated metrics (the SLG and OPS segments). Now, we need to remember that's nothing but suggestive as our study is simply one season of one league, but maybe there's more bite than I originally anticipated- i.e. that we might expect players with higher SAOPS numbers to project better than if using just OPS. Again, "might" is the key word there. Insert pickoffs and actual bases lost, and it's also possible that the correlations to EQR increase to near-perfect. That would be pretty cool.

Thanks. This is similar to what I tried to do. Like you I wanted to adjust slugging as well as OBP when a CS occurred. I also wanted to subtract Pick-offs and multiple bases in slugging when the CS was 3B and Home. I couldn't find data either and gave-up.

The other missing piece is dumb baserunning. You don't have to be a steal guy to get thown out a lot, and if you are a good steal guy it doesn't mean you are not out on the bases a lot. For example, guys who are thrown out multiple times per year trying to stretch have inflated OBP because they get credit for a single and suffer no consequences for the out made at 2B. And when they do make it, it adds 2 bases to SLG and 1 even when they don't. If OPS is your primary bargaining chip when it comes time to talk contract, the best play would be to try to stretch everytime. I know that is nutty and other factors would be considered eventually, but the math works out that way. I can hear the agents talking now. "Keep running. If you make an out it doesn't hurt your OPS and if you make it it helps your slugging.";)

Add to that the fact that some guys seem to be doubled off on line drives more than others or run full bore on a fly ball with 1 out and get doubled off that way. These things are probably pretty random and may not make a big difference in the big picture. I just think there are a few guys who make a habit out of making outs on the bases and it isn't taken into account enough in any of the formulas. It probably wouldn't be worth the work to try to identify these guys even if the data was available, but its interesting to me anyway. My own observation leads me to believe that the Reds have more than their fair share of these guys. A game during the August nosedive, where a couple of the few baserunners the team had were out on the bases, "inspired" me into looking at this.

Is there any website or data source that includes SAOBP, SASLG and SAOPS or is it your own off-line adjustment? I'm interested, but kind of lazy so a link or a site name would be great if any exist. If not, I guess I've got some work to do.

RFS62
10-17-2006, 07:43 AM
Even on the high school level in good programs, the coaches time the pitchers moves and catchers release times. It's down to math.

Pudge in his prime was 1.6. It's the gold standard. 2.0 doesn't sound a lot different, but it is.

Most bases are stolen off the pitcher. And yes, the slide step is a huge difference, but it comes with a trade off sometimes. A lot of pitchers lose velocity with the slide step, and it's not a good thing to use unless you've really perfected it.

Basestealing is both an art and a science. The greats like Joe Morgan studied pitchers moves with relentless precision. The mental game was more important to Morgan than just his speed.

I see less and less of that in today's game.

RANDY IN INDY
10-17-2006, 08:06 AM
Morgan knew what just the threat of the steal did to opposing pitchers, catchers, and defenses. Good post, RFS62.

wheels
10-17-2006, 08:06 AM
Without the ability to get on base, this conversation is not possible.

On Base Percentage is blood.

Ravenlord
10-17-2006, 08:45 AM
i'd be interested in seeing the the macro and micro stats for hitters and pitchers when you have a Juan Pierre/Carl Crawford/Willie Wilson/Ricky Henderson guy on first or second and less than two outs.

i'd also like to see how those stats break down by the good/average/bad player echelons.

IslandRed
10-17-2006, 10:09 AM
Just to semantically nitpick for a moment, what's measured for catchers is "pop to pop" time -- the time between when the pitch hits his glove and his throw hits the fielder's glove at second base.

cincinnati chili
10-17-2006, 09:50 PM
i'd be interested in seeing the the macro and micro stats for hitters and pitchers when you have a Juan Pierre/Carl Crawford/Willie Wilson/Ricky Henderson guy on first or second and less than two outs.



SABR has a statistical analysis committee that publishes a quarterly called "By the Numbers." A few years back they did a "Best of..."

Somebody wrote a great article on your topic. Maybe somebody here can dig it up.

For the record, not surprisingly, it does seem that batters hit better with fast guys on first base than slow runners. Not so much that you can justify a 65% success rate stealing bases or a lineup full of poor on base %'s, but enough that there is some measurable value in having fast guys in the lineup if possible.

OldRightHander
10-17-2006, 10:43 PM
Without the ability to get on base, this conversation is not possible.

On Base Percentage is blood.

It's kind of hard to steal first.

D-Man
10-18-2006, 03:49 PM
Wasn't that really the whole crux of Moneyball, to get more bang for your buck by utilizing what is undervalued by everyone else? I think a lot of people have attached Moneyball to sluggers who walk and strike out a lot, but I think the main point is more about going against trends to get more value for what you spend, or to avoid spending too much in areas that are overvalued.

Oakland has acquired some good basestealers (Johnny Damon and Terence Long come to mind), but I think it is a byproduct of Oakland's drafting philosophy and the relative price of speed (vs. power) on the market.

Speed is still very expensive to acquire via the draft. Athletes like Joey Gathwright generally get drafted early in the first three rounds (for millions of $) because it is a talent that any scout with a stopwatch can quickly identify. . . And unless they hit for power or take walks, speed players have a high-end projection Vince Coleman. More often than not, these speed demons turn into Terrell Godwin or Tom Goodwin washouts, and I think we can all agree that that is not good use of a 1st round draft pick.

On the other hand, guys like Jeff Bagwell or Jason Giambi hit single-digit HRs per year in the minors, yet they blossom into 40-HR hitters in the majors because power is a latent tool that often develops over time. A guy like Beane can develop a system in place to identify "good bet" candidates. In the process, he is using the inefficiency of the market (in terms of valuing power) to acquire talent.

So I think Beane's approach here is consistent with how it is portrayed in Moneyball.

[Sorry if this is probably a few pages to late. . .]

lollipopcurve
10-18-2006, 03:56 PM
guys like Jeff Bagwell or Jason Giambi hit single-digit HRs per year in the minors, yet they blossom into 40-HR hitters in the majors because power is a latent tool that often develops over time.

Uhhhhhhhh. In cases like this, the tool may very well be hypodermic.

NJReds
10-18-2006, 03:58 PM
So I think Beane's approach here is consistent with how it is portrayed in Moneyball.

It's not so much having speed, or a lack thereof. I think what strikes me is that the A's are so noteably averse to the stolen base that it no longer is a threat. So I'm not saying that the A's don't have team speed...I'm saying they've taken what team speed they have and negated its potential value as a threat.

The Tigers were a worse stolen base team statistically this season, however when a Granderson or Monroe found their way onto first base, the pitcher had to pay attention. The infielders had to cheat a bit.

When the A's had a Kotsay or Bradley on first, in general, the pitcher barely needed to acknowledge the runner.

In any event, I think this was an interesting thread and I thank all of you who took the time to answer my question.