THE BOOK--Playing The Percentages In Baseball

Here is what I want, probably a small improvement but I want the computer to make all the calls. You can set your optimal percentage of fastballs, away pitches, steals of second base, whatever, but the computer is the one that actually calls for the steal, or for the curveball on the 2-0 count. I’d like to see the same thing for football on third and short (run vs. pass).

Not sure what you mean, Ken.  Do you just mean to randomize the actions around a certain pre-determined mean?

On #15, you mean runner on third, right?

Tango, whenever I try to post from my iPhone, it goes through the preview process and I then submit, but the post does not show up in the comments.  Does it go to a spam filter?  Can you see these posts at all on your end?

Talk to me about #3.  What’s the advantage of a set lineup?  I’d think player “comfort"… but has that ever been demonstrated to be helpful in actual performance?

Rally/3: I check the spam queue at least once (in the morning), and I’ve yet to see anything that I had to unqueue over the last several days.  Unless you’ve been trying to direct us to some various adult sites.  It seems that we here are not iPhone-compliant.

Re #6: If you were to popularize WPA, relievers would be begging to NOT have to pitch a 3-run lead in the 9th, and instead come into the game with men on base in the 8th.

When a human decision is based on a meanginless cold-hearted number… well, why are the sabermetricians the inhumans living in basements, while the managers are the thoughtful ones who are not slave to the numbers?

Re #9: taking a pitch is worth something like .08 runs (to the opponent).  If a steal adds .19 runs and a CS takes away .45 runs, then a steal+takePitch only gives you +.11 runs.  The break-even point is now at 80%!  The only guys who should run when the batter takes a pitch is the guys who don’t need a batter to take that pitch!

#12: We have the chart in the book, and we have the Bonds chart on my site.  NFL coaches don’t mind referring to charts.  Why is baseball different?  Also, NFL coaches don’t chew and spit tobacco.  Related?

"NFL coaches don’t mind referring to charts.  Why is baseball different?  Also, NFL coaches don’t chew and spit tobacco.  Related?”

Perhaps- http://adr.iadrjournals.org/cgi/reprint/11/3/330.pdf.

In regards to #9 - this is a point where sabremetrics gets in the way of reality.  If the runner is stealing on a signal, the batter must not swing.  The reason here is that the runner is going on command and puts everything into stealing the base - supposedly because the manager perceives a need or advantage in doing so.

So now if the batter swings a few things can happen - a hit and everyone is happy (except that the runner is confused because he is not expecting to look for a hit ball, may even stop to see if it is a LD or FB), a groundball and a DP is avoided (again if the runner does not stop to find the ball), a LD or FB results in a DP and the runner scrambling to get back, an actual steal if the batter swings and misses, and finally the biggest turn off to the runner - the runner has second stolen, done everything right, but the batter fouls off the picth and the internal build-up to steal and a hard sprint is wasted.

If a runner steals on his own, then he knows the batter may or may not swing and accepts the above consequences.  Thus the master thief that appreciate the #2 batter that is patient.

Not every baseball situation can be played purely by the numbers since players are human (why else would Nash have done so much with game theory?).  Stats and analysis can guide what we do, just like in stock markets the human element can make the most sophisticated model inefficient.  You have to understand the complete dynamics and not just the opportunity cost of an attempted steal.

"No one EVER attempts a steal versus a LHP”

I’ve never heard this before.  Is there a thread on here talking about this?

Some comments

#3) To some extent I think rearanging the lineup is something that prevents certain managers from not beeing even worse than they are. When “In charge” there is often a strong desire (mentaly) and also sometimes public pressure (from outside) to “do something”. This is by no means unique to baseball, buisnes, miltary or political leaders suffer from this as well.

When this urge appears (and it will), I would much rather see that my manager starts doing some usually inconsequential moves in the batting order rather than getting stupid ideas such as “beeing more aggressive on the basepaths” or “pushing our starters deeper” or some other things that actually do damage rather than just beeing pointless. (Plus it distracts the media.)

#9) I think that most decisions about base stealing should be made by the baserunner, rather than the manager. Basicly I would have three lights, red (No steals), orange (runner decides) and green (you have to steal) and with green almost never beeing used. This of course requires some education of the players or they can simply use a chart, I assume most major leaguers can read?

#11) In theory it could be that for some teams it is correct to shift due to the skills of their infielders and tendencies of their pitcher and for some other teams it is incorrect. (I highly doubt that is the reason why some teams shift and some don’t but it is possible.)

#13) Here as well I think it is better if the batter choses at which times he bunts and when he should swing away. He has the whole defense in fornt of him and should have a better chance of guessing what they are trying to do than the manager in the dugout.

#14) Shouldn’t the quality of the pitcher AS A HITTER matter as much or more than his quality as a pitcher for when you take him out early. Just simply because the differences in their hitting abilty is potentially much greater.

The best hitting pitchers are likely (it’s hard to tell due to small sample size) pretty close to some of the utilty type players that could come of the bench to PH for them while others are an almost automatic out.

Bjorn/10: ditto of #3 and 9.

No way on #13.  Look up the thread from a few years ago with Adam Laroche (coincidentally, with the Braves at the time).  I’d have the same rule as #3, and that is, a red light on bunting, orange, and green, with orange almost never showing.

What I mean, is set fastball % for a certain pitcher to 61%, curveball 18%,changeup 15% whatever, and have a computer randomly spit out what pitch will come next. The point is to truly randomize your pitch selections or stolen base attempts or whatever decisions that are actually made. The idea was first for football which is easier since there are two choices (run/pass). Ideally you should be just as sucessfull on 3rd/4th and short with the pass as the run, so when that equilibrium is discovered(perhaps 71% run), make it random from then on out (with a mean of 71% runs).  My thought is that the randomness would make it harder for other teams to guess a steal, or a fastball is coming, which should benefit the user of randomness.

Of course, runner on 1st should be runner on 3rd (the foul ball).  Tango, can you change that? 

(Amazing that some people on BTF thought that it wasn’t a typo!)

All of these are guidelines and not hard and fast, although if a manager implemented them with NO exceptions, he would still be much better off than doing what he does now.

Batting order

1) Since batting order does not matter that much, managers can tinker if they really want to, but it is generally a waste of their valuable time.

2) If players are comfortable or not comfortable in a particular slot, I have no problem with the manager altering my order and going with that.  The difference “on paper” between alternative (but reasonable) slots is usually so small, that any increase or decrease in production due to comfort or familiarity is likely to outweigh that.

3) Mainly I want managers to know that it is not worth the time and aggravation to tinker with lineups, or I want a manager that already knows that, so that he can put in his time in other more valuable, fruitful areas.

Hot and cold streaks

If managers want to use those as an excuse to rest regulars and give some playing time to reserves, I have no problem with that. In fact, that is probably the best thing to do. It would be a little difficult to tell a player who is on fire, “Hey, I think you can use a day off.” I definitely do not want a manager making decisions during a game based on players being hot or cold.  I really would prefer that managers do not have that information handy.  It is only going to get them in trouble.

Stealing

1) Managers almost never give a, “You must steal on this next pitch sign.” I am not even sure that sign exists, other than in Little League or high school.  The reason is that the base runner has to feel comfortable and may not be able to get a good jump on any particular pitch.  They generally only have a “red light” sign or no sign at all. Sometimes the manager gives a sign telling the runner he wants him to attempt a steal some time during the AB, or a “steal if you can” sign.  About the only time that the batter takes a pitch in order to let the runner steal is when the runner signals the batter.

2) It is really not necessary to ever take a pitch and let the runner steal for two reasons.  As Tango points out, the break even point is much higher when you do that, since taking a pitch (that he might swing at ordinarily) costs the batter some runs, AND if the runner is going and the batter puts the ball in play, that is an advantage too.  In other words, you are happy if the batter does or does not swing.

Ken, you don’t really need to randomize your steal attempts.  Sure, you need to make sure that you don’t steal so often or so predictably that it is correct for the defense to pitch out more than occasionally, but that is usually not much of an issue.  As far as randomizing your pitches and your sac bunt attempts, two things:  One, you don’t really need a computer for that.  Although I agree with you that human beings are poor generators of random numbers, there are tricks that a pitcher and catcher (and manager) can use to help them randomize their actions and decisions AND you can actually do non-random things that appear to be random to your opponent, which is actually better than doing them truly randomly.  That is especially true if your opponent is not all that “smart.” For example, for the sacrifice bunt, I would recommend that the manager try bunting his better and faster bunters (non-pitchers) most of the time, and not bunting his slower and poorer bunters, in a bunt situation.  Even though this is a very predictable strategy, it will appear to be somewhat random to your opponent as they are more focused on the game situation than the speed or bunting proficiency of the batter (especially the former).  If the defense starts to catch on, then you do it more genuinely randomly.  I cannot emphasize enough how important the defensive alignment is to the decision of whether to bunt or not, which is where managers generally go wrong (not understanding that concept).

Bjorn, yes of course the quality of the pitcher as a hitter matters. Most pitchers are bad hitters though.  Obviously for Micah Owings, Carlos Zambranos, and Livan Hernandez, your strategy is different.  A good hitting pitcher is fairly rare though.  You won’t need to worry about that too much.  Remember that some of these things are just guidelines with obvious (and not-so obvious) tweaks and exceptions.

#13, sure a manager can tell a batter that if the defense is playing a certain way then he can bunt and if they are playing another way he can hit away.  In fact, that is a good idea.  For every batter though, depending on his overall hitting and bunting ability (and speed), he has a different “break even point” for where the defense needs to be playing.

#11, I don’t think the quality of the defenders has much to do with it.  Like a lot of things, some managers believe in the shift and others don’t.  It is one of those (many) silly, stupid things whereby managers (and other people) think that “opinion counts.” The only way to know whether the shift is correct or not is to do the research.  Now, it is possible that it is so close that we don’t care whether a manager chooses to do it or not.  And I will concede that to some small extent the range of the fielders factors into the equation, but I think the effect is small.

Stealing versus a LHP is like the IBB.  If no one ever attempted a steal versus a LHP, they would be better off (I think).  But I am sure that there are some basestealers and some LHP, against whom the runner can exceed the break even point.  In general, though, runners tend to ignore the times they get picked off by the LHP.  And as with all basestealing attempts, the BE point depends on the game situation.

Speaking of LHP and basestealing, one thing I forgot to mention in the initial post was that an underused strategy late in a game is to bring in a lefty pitcher to prevent a runner on first from stealing (one who is going to be around 90% or so against a RHP and the particular catcher).

You sir, are an idiot.  Have you ever played baseball, or any sport?  You want to abandon the “book”, but institute your own instead.  Nice.  While some of your ideas do have some merit, the beauty of baseball is that you never know what the result is going to be.  The only rule I would have is, “Don’t be afraid to go against the Book”.  And the only question that I would have for you would be, what in sports is 100%?  Even if the percentage is 95% pro, there is still a 5% chance of failure.  Then what do you do?

Leeeeeeeroy Jeeeenkins!

Guru: do the first words you post on this site have to be those?  This site is to foster discussion (see the TOS at the bottom of the main page), not for drive-by posts.  There’s plenty of other forums for that kind of party.  This isn’t it.

Regarding #2 on the list: When does a sample size cease to become small? And how long does it take for data to become irrelevant? I mean, I understand that Lance Berkman is not a .392/.470/.790 hitter, even though he is so far this year. And I feel comfortable that his 2005 stats aren’t relevant to his current at-bats. But what time period/number of PAs do you use to decide how good he is right now?

At about 200-250 PA, half of what you see is real, and half is not.

r=x/(x+200)

So, put in the number of PA as x, and that tells you the reliability.  Nice and simple rule.

Well, #2 is really about the “hot” and “cold” hitters (and pitches to some extent) as in over the last 10, 20, or 30 AB.  The thing you hear on TV and the radio all the time.  We have some nice research on that in The Book.  If you haven’t read it yet, I won’t spoil the plot.

As far as longer-term short-term performance (like the month of April), read the two threads we have on it here.  One for batters and one for pitchers. I think the research is interesting.

You are absolutely right about the 3-2 count. On Monday at my baseball game, I was in that situation twice. The first time, the pitcher made a borderline outside pitch, and I walked. The second time, the same pitcher threw another borderline pitch low and away. I took it; however, I was called out on strikes. For a poor hitter like me, one pitch away from first looks like a good opportunity. I think I would have only had a 50-50 chance of even making contact on the strikeout pitch. Ted Williams’ hitting chart flashed before my eyes. Skillfully fouling off bad pitches to hit is far beyond my abilities.

Brent, good deal!  I try and teach my cousin who is a very good high school player a lot of the concepts we discuss on this blog. He is also really into sabermetrics.  The 3-2 thing is really just a mathematical problem.  On a borderline pitch, how often will you walk (and then apply the win value of the walk) versus what is the win value if you swing?  Because batters are taught from an early age not to get called out on strikes and to “protect the plate” with 2 strikes, they swing too often at the 3-2 pitch.  And managers and hitting coaches do nothing to help them to unlearn this tendency (because they don’t understand it themselves).  Again, the pitch f/x data can tell us exactly what we need to know about the 3-2 count.  And what you do depends on the value of the walk as compared to the single and extra base hit.  If it is a situation where you mainly want to get on base (such as with 0 outs, no one base, or late in a game when you are down by several runs), then you are more likely to take on the 3-2 close pitch.

If you read the BTF blog which links to this post, it is amazing how many people who have read my stuff think I mean these recommendations “literally.” I don’t.  Maybe I did not make that clear.  (Also, people need to recognize the difference between a well-thought out edited article and a blog post made in the middle of the night.) Two things about these recommendations:  One, obviously not all of these are literal or hard and fast.  Two, despite that, in many cases, I am confident that making them a hard and fast rule (such as no IBB’s, ever), even though it is not technically correct, is a lot better than what they do now.  That was the tone of these recommendations:  Here are some rules of thumb that although technically there is plenty of “wiggle room” within some of them, if you did nothing else but abide by these like clockwork, you would substantially increase your team’s WE at the end of the year.

Anyone who thinks that I think a a manager should pay no attention to who might be sick, injured, or otherwise not feeling up to par (mentally or physically), or should not be aware of, and act accordingly on, any mechanical or other physical or mental changes a player has made or has gone through, for the good or for the bad, is either really dense or just likes to criticize for the sake of doing so. I am not referring to the comments on this blog of course.

Batting order

I think taking five minutes to make a few random (but sensible changes) in batting order can save both the manager and the players both time and grief in terms of media hazzle when the team’s offence has “in a slump” (i.e. has had bad luck) for a couple of games.

Bunting

These guidelines are for sac bunting i guess, what about bunting for a base hit? Do some players have a “licence” to do that at their own discretion? That (to me) appears to be the way it is done today…

Pinch hitting for pitchers

One thing I thought about yesterday is that in some situations you also have to take into account the flexibilty (or rather lack flexibilty) you have with your bench players.

With the trend for carrying bigger picthing staffs on the 25-man roster you have less room for extra position players. Also PHing for your starters earlier would increase the need or at least desire for a bigger bullpen.

I can remember at least a couple of games this year where the anouncers has made comments about how the manager is somewhat “handcuffed” by the fact that they are currently carrying 14 pitchers. And even with 13 pitchers you only have four guys, backup catcher, 5th infielder, 4th outfielder and one more guy and that is assuming all 25 players are availible.

Even if I think that managers are beeing to risk-adverse in that respect I can certanly understand the desire to not “use up” the bench players to early and then be without a suitable replacement if one of your starters would suffer an injury.

Hot and cold hitters

This is one area where I think “the naked eye” can actually beat out statistics. A batter might have mechanicly great at bats and go 0-4 and vice versa.

A batter might have mechanicly great at bats and go 0-4 and vice versa.

Willie Randolph, after Pagan was hitting lights out, said: “We know he won’t keep this up.”

So, here is a manager that presumes that Pagan has a certain talent level, and was exceedingly lucky.  And therefore, Randolph would be a manager that might see a guy go 3-4, and decide to bench him, because he knows better.

***

MGL’s choices of: a) hard and fast rules versus b) keep doing what they are doing is a good trade.  Obviously… OBVIOUSLY… there are exceptions.  Some of the readers at BTF give no benefit for the doubt. I also don’t understand why they wouldn’t question mgl here (no registration!) so that he could set the record straight, rather than giving him no benefit of the doubt, and creating phony outrage.

I get the impression that had MGL’s 18th rule been “I will not consider skin color when hiring my manager”, there would be somebody on BTF saying “MGL doesn’t think it’s a problem to have a manager with neon yellow skin as a result of an infectious alien disease.  He doesn’t care if the whole team is infected with this terrible plague!”

Yup, I am a punching bag on BTF!  Hey, at least I generate some responses.  Better to be infamous than unknown I guess.

Bjorn, batters are on their as far as bunting for a hit (and occasionally they even sac on their own). 

Batters that bunt generally do so at a .400 clip or so, and the better ones at more than a .450 clip (and then there is Taveras, who can seemingly bunt for a base hit at will). 

Obviously it is usually done (bunting for a hit) when the bunt single is worth a lot relative to the extra base hit (e.g., you don’t want to bunt with a runner on 2nd and 2 outs, unless you can be safe like 60-70% of the time - I don’t know the BE point off the top of my head).

Last night a young player on the Giants (I forgot who) tried to bunt for a hit with 1 out and a runner on second.  While that is not as bad as with 2 outs, it is generally a bad play.  But, as with most bad plays (like trying for third base with 0 our 2 outs), there is still a BE point above which it is not bad anymore.  To tell a player, for example, “Never make the first or third out at third, is doing him a disservice by not telling him the complete story.  Why can’t you tell him the truth, which is, “Only try for third with 0 or 2 outs when you think you can make it 85 or 90% of the time,” and that when he makes out 10 or 15% of the time he is not going to get chewed out?

Anyway, with 1 out and a runner on 2nd, a bunt single picks up .52 runs or so.  An out loses .381 runs, so the BE point is around 42-43%.  Since a bunt attempt also ends in a foul ball a lot of the time, we have to up that figure a little.  So 45% is a reasonable BE point I think for that situation. So it really is not a bad situation to bunt in, if you are a good bunter and the defense is not expecting it (which they usually are not in that situation).

That is a good example of where conventional wisdom, or the wisdom of the crowds, or whatever you want to call it, can easily go afoul. That happens when there are other “forces” or motivations besides maximizing win or run expectancy. That happens all the time in baseball.  Examples of those “forces” in baseball are “risk aversion,” “ass-covering,” and machismo.  In this case, the “conventional” thought is that, “You are passing off the chance to score the runner on the next guy.” The only consideration of course is whether your WE is increased or decreased by the bunt attempt.  Baseball players and managers do not understand that concept at all.  There is only “wisdom of the crowds” when there is some requisite level of knowledge and expertise in that crowd.  You can give a group of hundreds of thousands of people a complicated math or physics problem, and unless there are some experts in math or physics in the crowd, there ain’t going to be any “wisdom of the crowds.”

The only thing that stops good bunters from bunting more often are the number of opportunities and game theory.  The more they bunt, the more the third baseman (and to some extent, the first baseman) must play in (thereby increasing the value of swinging away and decreasing the value of the bunt).

As far as hot and cold streaks, I think that the eye, even a trained one, will fool you just as much as the stats will.  And if I were in complete charge of a team, I really don’t care that much what a manager does with his lineup or even with hot and cold streaks as long as he does not use it to make too many suboptimal decisions.  If he wants to tinker with the lineup a little, that’s fine, and if he wants to use hot and cold streaks to give starters rest and bench players some playing time, that’s fine too.  If he starts IBB’ing hot players on the opposing team or benching cold players for hot ones (when the cold players are better players) on a regular basis, I DO have a problem with that.

Mainly, I just want a manager who does not believe in all that nonsense.

I have one: this will be the most exaggerated circumstance I can think of, but I am sure there are situations where you would want a base runner to purposely allow himself to be picked off:

1st inning, two-outs, your clean-up hitter up, runner on first. The strikeout pitcher with a huge fastball has thrown 2 breaking balls for strikes, and is ahead on the count 0-2.

Your clean-up hitter, who is prone to striking out anyway, and also owns a .500 OPS with two strikes, hasn’t even seen his fastball yet.

In this situation, I would like my runner to just start strolling to second base while the pitcher is checking the sign.

This is certainly the most extreme example I can think of, but I am certain there are other less extreme circumstances where a pick off and a clean at bat are preferable to having the at bat continue.

I had no intent to come off as a drive-by-poster. 
What bothers me is for all this talk of the “Book” being bad and not of absolute sense, all I ever read are theories that substitute “old” rules for “new” rules, and thusly creating another “Book” for which the next generation of fans (SABRs included) will dispute as again, not being absolute.  I know this is just a forum to discuss our favorite sport, but everyone needs to realize that stats are acquired after the fact, not prior to the action.  Percentages are great, but never add up to 100%.  Would you be comfortable betting the house on a 75% success rate?  Maybe in a regular season game, but not in a playoff game.  Managers have a tendency to go with the safe route, and there is a reason for it.  It’s not the “Book”.  That’s all I was trying to say.  Please continue with your manipulations.

Matt, interesting scenario.  Let’s forget about the “500 OPS etc, etc” as if that’s his true talent or whatever.  Let’s get into the meat of it:

You have a great hitter, one who is +.08 to +.10 runs per PA in an average situation, on an average count.

On an 0-2 count, that goes down by .09 runs:
http://tangotiger.net/halejon/allcounts.html

In short, just as a 2-0 count turns an average hitter (at 0-0) into a great hitter, an 0-2 count turns a great hitter (at 0-0) into an average one.

Now, the run expectancy of man on 1B, 2 outs (0-0 count), Table 1 of The Book: +.117 runs, with an average batter.

When you have a great hitter, with an 0-2 count, you really have an average hitter at the plate.  And so, that RE table works for him.  And so, by purposefully being picked off, you are giving up around .12 runs.

The benefit is that now your great hitter comes up to bat in the next inning (but with bases empty).  As we said, your great hitter is around +.08 to +.10 runs per PA, in average conditions. For a great hitter, his +.09 runs per PA breaks down as around +.06 with bases empty and +.12 with men on base (average out situation, meaning outs = 1).  With 0 outs, those numbers go up by 50%, and so, bases empty 0 outs means that he’s worth +.09 runs to start the PA, compared to the -.12 in the previous inning that you gave up. 

It’s not a good trade, but close enough that it is possible that you can come up with some scenario where you want someone to get intentionally picked off on an 0-2 count with runner on 1B only, with 2 outs.

Ballsy…

Wow, interesting.

Well, if the trade off is .03 runs, rather than a get picked off situation, does that make it sort of a “must attempt a steal” situation?

Even if the runner is only 50% successful, the RE would RE would only need to increase by .06 for it to be a wash, and the RE must increase more than .06 with a runner on 2nd, as opposed to a runner on first.

Of course, the batter has a high possibility of striking out on the next pitch, which would ruin my objective.

BUT (to question the facts), how does a great hitter have a RE of just .06? An Ortiz-like hitter has a .064 chance of hitting a home run (45 home runs in 700 PAs) all by himself--so that number seems wrong.

Matt, an out has a negative value.  Your typical great hitter will be say +60 runs in 600 PA.  I suggest you read all you can on Linear Weights.  Or get The Book.  Feel free to submit any specific questions you have through our mailbag (see link at top right corner).  Your questions may help others who are new around here.

How about one for Charlie Manuel...stop taking out Pat Burrell for defense and losing him an AB late in games.  The amount of runs he will cost you in the field is much smaller than the amount he will gain for you at the plate.

Bjorn/10:

“I would much rather see that my manager starts doing some usually inconsequential moves in the batting order rather than getting stupid ideas such as “beeing more aggressive on the basepaths” or “pushing our starters deeper” or some other things that actually do damage rather than just beeing pointless. (Plus it distracts the media.)”

It seems to me that this is a place where a manager who’s skilled at manipulating the press can literally make something up and that will be enough. 

Look at Dusty Baker and Corey Patterson - CP’s not hitting, his true talent doesn’t justify his playing time or his batting leadoff - but Dusty doesn’t want to change anything he does, so he just leaves Patterson in the leadoff spot against righties and says “Corey’s a good hitter, don’t worry about him.”

So I don’t think you’d ever have to worry about appeasing the media because as long as you threw a cigar-chomper up there at the post-game press conference, you’d be fine.  Managers have said so many stupid and incorrect things over the years, that if that guy said “Joe Bob’s a #3 hitter,” nobody would care that you weren’t doing anything.

Billmac/31: the fielding gap between a great OF and a really bad one is 40 runs per 162 9-inning games.  That’s roughly .03 runs saved per inning.

The gap between a great hitter and a really bad one is roughly 85 runs per 700 PA.  That’s roughly .12 runs cost per PA.

So, that’s your tradeoff.  If you replace a great hitter/horrible fielder with a great fielder/horrible hitter for 1 inning, then the break-even is if that lineup slot only ends up coming up once every 4 games.  If it comes up less often, good deal.

If you do it for 2 innings, then the breakeven is now at 50%.  So, if that batting lineup slot will come up less than 50% of the time, then good deal. 

You can’t do it for 3 innings or more obviously.

So, figure out how much Pat and his replacement are worth at the bat and on the field (as per PA and per inning), and give us the answer!

I have an addendum to #13, but it only applies if you are running the Dodgers: Juan Pierre must bunt every time up.  It’s not like he’s ever going to take a walk OR hit the ball out of the infield anyway, so just bunt.  And if the third baseman plays 20 feet from the plate, just do what I saw Pierre do a couple weeks ago: bunt it over the third baseman’s head.

I enjoyed reading this.  I read it with a grain of salt, knowing that absolutes make for good conversation but should rarely be taken literally.

My first post, so go easy on me. 

Regarding #15, as an earlier poster commented, don’t you mean runner on third?  Otherwise, I can’t see how you don’t take the out in the situation you mentioned. 

Similarly, I would think there are some situations where it’d be to your benefit to drop fair balls with a runner on first and less than two outs, and then take the force at second.  For example, if the runner on first represents a “game changing” run and that runner is fast, while the batter is considerably slower.  What do you think?

#16 seems to fly in the face of #2 and #3.  Hitters go into funks, some hitters are more prone to streaks than others, I’m sure you could even come up with a number for how “streaky” a player is .. call it a “Funk Factor.” For example .. Vlad has a very high Funk Factor of .8 which means his batting ability for a given game is 80% of what he has done in the last 10 games and 20% his career average.  You are probably not going to sit Vlad out because he’s excessively funky, but I think if your have a high-funk, marginal player on a hot streak, you have to keep him in the line-up until he cools.

#6 is really the only other one I can see a problem with.  Having set roles in a bullpen goes beyond ego, psychology, or economics.  I’d argue allowing pitchers to have a set routine outweighs the leverage benefits of mixing roles game to game.  Is there any data comparing performance of relievers in a bullpen by committee situation vs.  their career/season averages?

If you could effectively enact rule #11, I think you save your team over a run per game.  You are absolutely right that fielders need to be moved more batter by batter and perhaps even pitch by pitch. 

With absolutely no data to back anything up, I believe outfielders play far too deep.  And going out on a bigger limb, I think you could improve your defense if you told the pull side outfielder to play straight away, the center fielder to shade the opposite way and two steps shallower, and the opposite fielder to play 6 steps shallower.

Chippy/34: This was a typo as noted in MGL/13.  I’ll fix that in a sec.

Justin/35: If you read the chapter in the book that deals with streaks, you will see that if you only focus on the numbers, you can’t tell if someone truly is in a funk.  To figure out if someone truly is or is not in a funk, you have to go beyond the numbers.  Basically, it’s a gut instinct call.  I gave a small allowance for gut instinct as I do for clutch hitting, somewhere around 10 wOBA points (equivalent to 10 OBP points), and no more than 20 wOBA points (20 OBP and 25 SLG points).

Anything more, and I’d like to see the evidence that someone has the gut instincts to figure out the funkiness of someone else.

Sorry about that, I overlooked the response.  What about the second part of my question? 

“Similarly, I would think there are some situations where it’d be to your benefit to drop fair balls with a runner on first and less than two outs, and then take the force at second.  For example, if the runner on first represents a “game changing” run and that runner is fast, while the batter is considerably slower.  What do you think?”

Is there just too little advantage gained (if any at all)?  Too much risk of not getting an out at all?

Chippy/34,37: I’ve seen that fairly often.  I’m all in favor of that, but I think everyone is.

Jeff/34’s post was marked for moderation and has now been opened up.  (That means all our post reference numbers from that point to now gets +1).

I want to highlight his sobering words:

I read it with a grain of salt, knowing that absolutes make for good conversation but should rarely be taken literally.

I read it with a grain of salt, knowing that absolutes make for good conversation but should rarely be taken literally.

You can say the same thing about generalizations!

Seriously…

While it is true in the sense of what is “accurate,” it is not true (IMO) in terms of advice to people.  As I said several times already, and I stick to that opinion, I would MUCH (much, much, much) rather have a manager who had to follow my advice to a T (or is that “tee”?), even though there are plenty of opportunities to correctly deviate, than do what they do now.  They would pick up several wins a year.

And, to tell you the truth, if I told my manager that he could modify my guidelines according to what he sees, knows, etc., I think it is debatable that he could increase his overall WE.  I think he would make too many mistakes that would negate any advantage he would get from what he “knows.” But it is no big deal either way.

The point about taking these with a “grain of salt” is a good and accurate one and should be obvious, at least if you are a non-BTF poster.

Chippy, I have seen an infielder drop a line drive or a pop fly in order to get a force on a faster runner a couple of time, but not very often.  I have often wondered why they don’t do it more often.  One, maybe the few times (1%?) that you don’t get the out negates the benefit (I don’t know) and of course the umpire can call the batter out if he thinks you dropped the ball on purpose (I think).