Playing Small Ball

[ProfessorHank]

Right now, my take is high OBP, SPD, STE guy lead off. #2 is low SO guy who can do some bunting. #3 is best overall batter, #4 power guy. Typical, traditional lineup, I suppose.


If lead off get's a single, #2 guy bunts him over. Then I hope for #3 or #4 guy to knock him in.


If #2 doesn't get him over with a bunt, then I'll have #3 guy bunt too.
I'll only have #1 guy steal if pitcher's hold is way low.


If we get behind by more than 2, I abandon this strategy and go for more of a long ball strategy.


Current season is 1910.


Haven't had much luck with this yet, just wondering what your take might be.

[Reed]

I'm not a fan of a position player giving the up an out (bunting) unless it's in the late innings and you really need that run. (Even in 1910). I'm not much of a fan of stealing unless the chance of success is very high (80% or higher) so that would depend on you players. If you have someone like Tris Speaker coming up, I wouldn't bunt. Anyway, that's my opinion for whatever it's worth.

[MBarrett]

Ever since I committed to Sabermetrics and advanced stats, the word "bunt" is like nails on a chalkboard to me.

Don't do it. Ever.

[The Game]

Originally Posted by MBarrett

Ever since I committed to Sabermetrics and advanced stats, the word "bunt" is like nails on a chalkboard to me.

Don't do it. Ever.

Would you rather have a man on 2nd with one out or a man an 1st with 1 out and a guy that could GIDP at the dish?

[MBarrett]

Originally Posted by The Game

Would you rather have a man on 2nd with one out or a man an 1st with 1 out and a guy that could GIDP at the dish?

I'd rather the man at-bat gets a chance to drive the run in or move the runner over and get on without giving up a free out. You only get 27 outs per game, you should treat them more preciously.

[drhay53]

Bunting with 2 guys in a row, including a number 3 hitter, is madness.

The short answer is that outs are the most precious thing in baseball and you should never, ever give them away freely. "Never, ever" can be qualified by saying that sometimes in high leverage it's not quite as bad of an idea as it is every other time.

Let's take a look at some run expectancies, shall we? http://www.tangotiger.net/re24.html
I'll be looking at the average runs after that situation chart (top), from 2010-2015 (though for most purposes, the overall story from these charts stays pretty constant even in different run scoring environments).

First, let's evaluate the stolen base.

Runner on 1st, nobody out: 0.859 runs
Runner on 2nd, nobody out (successful SB): 1.1 runs
Nobody on, 1 out (failed SB): 0.481
Reward for SB success: 0.241 runs
penalty for SB failure: 0.378 runs
reward - penalty: -0.137

This is the reason for the statement that in order to run you must be successful ~80% of the time. The penalty for a CS is significantly higher than the reward for succeeding.

Now, let's look at the runner on 1st, nobody out sac bunt.
runner on 1st, nobody out: 0.859 runs
runner on 2nd, 1 out: 0.664 runs
and for giggles, first sac bunt fails, and you try again:
runner on 1st, 1 out: 0.509
runner on 2nd, 2 out: 0.319

Net effect of successful sac bunt, nobody out: -0.195 runs
Net effect of successful sac bunt, one out: -0.19 runs

By giving up an out, you've actually reduced the amount of runs you expect to score in the inning, even though the runner is now in scoring position. Therefore, the only time you should even remotely consider a sacrifice is if the run has extremely high leverage and you're desperate.

It's important to note that sacrifice bunts are not 100% successful either. Sometimes they result in double plays, sometimes the lead runner is thrown out and your situation is even worse than it started. But sometimes both runners are safe and there are errors and you end up 1st and 3rd, 2nd and 3rd, etc. These outcomes are all captured in the run expectancy matrix though, so it's pretty safe to say that the sac bunt is a securely negative run proposition. Don't do it.

[NoOne]

To the OP: man, jsut run with it... if you really want to know if it's better or worse than a different strategy you'd need a long sample size simulation and you'd have to look over that data.... but, you'd also need a baseline in order to make any meaningful comparison or assessment of if it's better or worse than somethign else.

what's true in RL may not be the same odds in the game... do not assume that. your LTM/LT will be the biggest factor of whether bunting often will be the best odds to play and how often it's true given all the situations you will encounter over time in their propoer proportions (ie suitable sample size).

1 season or even 2-3 probably won't tell you enough on it's own. best way to hash it out would require turning off any other moving part that would affect this or masks its effect in some possible way.

with good data/knowledge, this is a simple break-even analysis. you don't need a major in math major, algebra is enough. if you don't know the probabilities of the various outcomes, it's just a flat-out guess. watching 20-50-1000 games won't do it... the human brain can't work with that much data. you need to know probabilities from large pools of data to form how you make those decisions and at what point.

i don't have that research handy, do you? i'm sure you can google it, though. since you play in a different era, you can't use modern mlb info, though, lol... historical play may require you to repeat the same process that the sabrematricians have done with modern data but with that older data.

a sabermatrician, or rather just a ratonal mind, lets the relevant facts dictate decision-making. they would happily bunt in any situation that the numbers dictate that as being the best odds of winning the game. i'd suggest this mantra for any aspect of baseball, relative to your LTM/LT and settings... this won't always translate to other people's leagues.

there are so many moving parts to consider, speed, BA (including whomevev may get to AB after that guy is a consideration), power, bunting ability, speed of baserunning, situations (outs, count, where baserunner is etc)...

so, that's the other problem... sabermetrics may give a black and white answer for a decision, but if they didn't use the right information and if all the probabilities of outcomes are not thoroughly known, it's not going to give you accurate information about what to do.

it's never about preference, or it shouldn't be... it should be about facts. whatever they are for your league.

[drhay53]

2 things. 1, bunting is a negative run outcome in basically every era that has a run expectancy matrix available. 2, the proper sabermetrician knows that baseball is a game and therefore game theory applies, and the sac bunt should be used appropriately in that context.

[venflaalachi]

Bunting with your #3 hitter makes no sense. Bunting with your number 2 guy doesn't make much sense either, if you're gonna give up an out why not just try to steal second?

[NoOne]

First, i never bunt and i probably wouldn't bunt even if its 51/49 in favor of it... but that's not the point. (and i would be wrong for being a stubborn ignorant twit for doing so, but i am okay with that.)..

the game isn't restricted by RL eras, and i wasn't referencing his specific game situation, but all situations possible.

It definitely isn't exactly like the real mlb, so direct translations of concepts can have some sever drawbacks if you assume it is the same.

http://www.baseballprospectus.com/ar...articleid=2844

looks like good info... provides basic equations you can plug the variables into form your specific situation. However, if your league baseline isn't exactly the same, that information is going to be off for you. (i didn't read closely, if he explained methodoloy and data set used to get baselines, you'd have to recreate that knowledge for your league - doable but require effort and likely not what anyone considers fun)

but, you can't apply 'averages' to all players either... that average includes speedy guys as well as a victor martinez-type... applying some matrix based ont he league to ricky henderson isn't going to work well, for example.

he mentions this in the last paragraph... those equations in that article are too simplistic and naive. you would need solid info on each of your players capabilities and how they relate to baseline, then aco**** for that... plus the other team's players and how they relate to the situation at hand. not likely to have a suitable sample for that type of knowledge.

Last paragraph quote:

"Finally, even given our extremely light-hitting Batter One, the upper range of Breakeven--where sacrifices increase run expectation--indicates that there is a decent chance that there are some cases around the league where sacrificing may be beneficial. Enhancing the equations should allow us to further estimate where that threshold may lie."

And, that's a modern league with real power... take that power away and alll of a sudden that bunting might become a really good idea. if early 1900's is somewhere near default, ... well the elevated BA may offset the loss of power? no idea... again... it's not so obvious to the eyeball, ever.

even so, i doubt bunting 2-3 times in a row can be beneficial for most people's leagues... but, it's not impossible in OotP.

[PSUColonel]

I a big fan of hit & run more so than bunting

[HonusWagner]

my 2 cents is that most here haven't seen competent bunting. what is in the game now and has been for decades isn't bunting. it's desperation -

1. place bat horizontally across plate
2. hope for best

there are such things as infield hits (accidentally). so why can't they occur intentionally? for the same reason that everyone believes a HR is somehow exciting rather than a game killer.

imagine having a guy that could line it 6 ft off the ground at 100mph down the line - or - place it lovingly to die just inside that same line. result: just another variable for the defense to cover, meaning more opps for errors and offense.

but, for the same reason that HRs are considered manly and made by heros, bunting isn't. bunting is only nuanced. "boooh - ring"! it also no longer exists, hasnt for generations.

[BBGiovanni]

Originally Posted by venflaalachi

Bunting with your #3 hitter makes no sense. Bunting with your number 2 guy doesn't make much sense either, if you're gonna give up an out why not just try to steal second?

I once watched a major college baseball team with bases loaded with no out and their cleanup hitter at the plate bunt. Pretty sure the run expectancy on that strategy is not too good

[drhay53]

If you spend the time to get really exceptional at bunting the defense will adjust and force you to swing the bat. If you get really good at hitting homeruns there is nothing the defense can do about it. The only game theory to play is then between the pitcher and the batter, and for a batter with good discipline, that means the compromise the pitcher will settle on is walking you. The closest we have to game theory on power hitters is the extreme shifting, which, frankly, doesn't really seem to be creating more outs at a statistically noticeable rate.

[frangipard]

Originally Posted by drhay53

2 things. 1, bunting is a negative run outcome in basically every era that has a run expectancy matrix available. 2, the proper sabermetrician knows that baseball is a game and therefore game theory applies, and the sac bunt should be used appropriately in that context.

Run expectancy is not the best metric; win expectancy is (and looking at WE, there are a few high-leverage situations where the bunt does make mathematical sense).

Your note about game theory is quite correct, and often underappreciated.

[drhay53]

Originally Posted by frangipard

Run expectancy is not the best metric; win expectancy is (and looking at WE, there are a few high-leverage situations where the bunt does make mathematical sense).

Your note about game theory is quite correct, and often underappreciated.

I agree that the win expectancy chart is more important, and I did point out in a couple of places that both bunting and stealing bases are more valuable in extremely high leverage situations.

On the game theory point, I don't think that matters as much in OOTP because I don't think that AI managers have any real ability to know tendencies and adjust for them.

If I had time, I'd like to do a test in OOTP where all players and pitchers are identical, with identical managers except half of them with bunting at 0 and the other half with bunting maxxed out. It would be interesting to run a few seasons like that and see what happens.

[fcabanski]

You have your best hitter sac bunt?

You're fired.

[Rosco Peabody]

I got a player who actually wasn't rated super highly because all they had was a high contact rating and high speed and stealing, but I use him SUPER effectively.

He leads off and gets on a lot (due to his high contact, he easily hits well into the .300s). Then as long as I avoid the pitchout, I can steal 2nd and 3rd and have three outs to play with to drive him in (which happens most times. He's only about a 60 overall, but he's added such a crazy dynamic to my team

[jpeters1734]

What are the odds of grounding into a double play? If it's greater than the expected obp, then bunt.

[pgjocki]

Originally Posted by jpeters1734

What are the odds of grounding into a double play? If it's greater than the expected obp, then bunt.

This or the guy at the plate strikes out more than Mark Reynolds...

Sent from my VS985 4G using Tapatalk