In-game_Strategy
In-game_Strategy
Sunday, January 17, 2016
?I have a very simple rule change, and invisible enough to not be noticed by the casual fan, to do away with the DH while limiting the pitcher from batting. I've talked about it many times over the years, with the last time right here.
All you need to do is modify the substitution rule so that at the end of the half-inning, the manager has to choose which of the two players to kick out of the game: the substitutee or the substitutor. Adam Wainwright is due up to bat. The Cards manager sends a pinch hitter. At the end of the half-inning, the manager has to kick (at least) one of the two out of the game. He kicks out the PH. Wainwright stays in the game. Once again, Wainwright comes up to bat two innings later. This time, bases empty and 2 outs, and the Cards are up by 3. Manager might put in a PH, or he might let Wainwright bat. He goes for the PH again, and since it's the bottom of the 6th, kicks out both players from the game.
It's possible to burn through your bench to keep your pitcher from batting. Or, you might be strategic and pick your spots as to when the pitcher will bat. There's no reason that a player that was substituted for MUST be kicked out of the game. Happens in soccer, but not in the other three N.A. sports.
And, it's not limited to PH of a pitcher. You can apply the rule to PH for any player or pinch runners or defensive subs. A simple rule that doesn't affect the sanctity of the game. Though I will be cursed by all computer programmers.
Friday, January 15, 2016
?Ben does a terrific job giving us the lowdown on pitchouts.
He shows that when there's no pitchout and the runner tries to steal, the success rate for the runner is 73%. If we make it +.20 runs for SB and -.45 for CS, that's a run value of +.025 runs. With the pitchout, the runner is safe only 48% of the time, making it worth -.138 runs. That's one heckavu great deal. Except you also are giving the batter an extra ball.
However, what if there is a pitchout and the runner stays put? Well, in that case, the batter benefits. Normally, Ben says you'd have .64 strikes and .36 balls. If we give strikes a run value of -.05 for the batter and balls +.05 runs (I'm assuming you don't throw pitchouts come early in the count), then normally we'd have -.014 runs. But the pitchout is an automatic +.05 runs or a switch of +.064 runs.
At this point, I'd need some more data. The pitchout-stealAttempt is -.138 runs for the runner and a fraction of +.064 runs for the batter (because sometimes the runner makes the third out on the pitchout). For discussion purposes, let's say that works all out to -.09 runs. The pitchout-nonSBattempt is +.064 runs. That basically sets the breakeven point at around 40%. That's how often a manager has to be right that the runner is going. Which is crazy high.
Of course, you've have to rework all the numbers because you are going to pitchout with more likely runners who are better percentage runners than what we're assuming up there. Anyway, great job by Ben to bring it out in the open. And there's some work out there for an aspiring saberist to take the groundwork that's been laid and come up with better numbers.
Monday, November 23, 2015
?Bill James wonders if we can make a 3-man "rotation" work.
What it really comes down to is forgetting about the notion of a "starting pitcher", since in his setup, there's really no obvious benefit for the 5-inning pitcher to start the game as opposed to say entering the game in the 3rd or 5th inning. He simply is going to get his 5 innings of work in. (From a W/L perspective, this "long man" would actually benefit the most by entering the 3rd inning.) And in the case of an already lost-cause game, he could be delayed one day.
Anyway, Bill argues that this pitcher has to be capped at 5 innings or 80 pitches, whichever comes first. (I would also add the provision of a maximum of 18 batters, which really sets the effective cap at 70 pitches, but let's set that aside for now.) That would likely average out to 70 to 75 pitches per start, down from 95 average since 2002. Figuring a realistic number of games for the long man of 50, that gives us 3500 pitches, a typical workload in the pre-2000 time period.
But the harder sell is really the setup time, the warmup pitches, the whole process. There's a reason that relievers don't approach 3500 pitches. Take for example, or for extreme, Bob Stanley in 1982. He threw 48 games as the long man, facing nearly 700 batters, which would mean about 2500 pitches. With a league-leading ERA-, he was obviously extremely effective. The next season, he threw in 64 games, facing closer to 600 batters, or about 2100 pitches at an even better performance level.
So, I think that's the kind of tradeoff you get in terms of increasing games and decreasing workload. At 35 games, you can get your 3500-4000 pitches. At 50 games, it's down to 2500 pitches. At 65 games, it's down to 2000 pitches.
Going back to Bill's proposal, if we're going to create a series of long-men, say 3 or 4 per team, that would have them average out to 55 pitches per game with 50 games, maybe setting a hard cap at 60 pitches. Which also means setting a hard cap at 18 batters, which of course The Book has shown would be very effective at an individual pitcher level.
Therefore, going with that, can the Straight Arrow readers construct a viable plan with 3 or 4 long-men, each given a cap of 60 pitches or 18 batters (whichever comes first), with 50 or so games?
Wednesday, October 21, 2015
?MGL takes a look to see how much of an impact each decision has, and how much you have to assume in order to make the decision break-even.
Let's see if I can try to figure it out in one paragraph: a pitcher-as-hitter is some .13 runs per PA worse than league average, or .013 wins per PA, when LI = 1. If LI = 2, then it's .026 wins. If LI = 3, then it's .039 wins. So, that's your cost of letting him hit. What's the gain for letting him pitch? If you assume that an almost-great pitcher gives up runs at 80% of league average, and his replacement is league-average, and an average pitcher gives up about 0.45 runs per inning, then you gain 20% of that, or .09 runs per inning, or .009 wins per inning. If he pitches 2 innings, that's .018 wins at LI = 1. At LI = 2, it's .036 wins.
All in all, prima facie, you have to work it out, but it may be break-even. But MGL went through the machinations more closely, and he shows that the decision is pretty straight forward. All to say: everyone should roll up their sleeves and work it out. And don't forget the huge PH penalty too!
Wednesday, September 09, 2015
?A nice account from John that shows how often the AL team's regular DH when he plays in an NL park. And if the NL team draws from their regular fielders when filling up the DH in an AL park.
Saturday, July 25, 2015
?Just insert the pitcher in cleanup, and move everyone else below him down 1 rung.
In The Book, I said the cost was 0.10 runs per game. (I don't remember if I kept him batting the whole game, or it was based on being PH. Run it both ways anyway.)
I'd like to see what the other sims are saying.
***
And if you have your own simulator, please run that and report the results.
Saturday, July 18, 2015
?Ben takes a look at historical and current trends of the pitchers batting in the 8th spot.
The Book showed that the best spot for the pitcher is batting 8th and worst is (unsurprisingly) as cleanup hitter. The gap was about 16 runs per season. Basically, it means that each slot between optimal (8th) and ridiculous (4th) is worth about 4 runs. And that's with the hitter who is BY FAR the most different from everyone else. That's what we are really up against.
Thursday, July 09, 2015
?One of the big revelations from The Book back in 2006 was the Times Thru The Order Penalty. That each time through the order, the pitcher would take a big hit. Passan is looking at the Rays buying into the idea. The Rays have been sensational the 1st time thru (best performance in MLB). They've also faced 457 batters the third time thru, which is fewest in MLB so far, (However, if you include 3rd AND 4th time thru the order, Royals faced the fewest batters.)
Friday, June 12, 2015
?MGL has brought up the point that maybe you'd want a LHP to face a RHH, if you have a runner on first. But, how about Venditte? Would he ever do that? It would depend on what his platoon splits are. Let's say it costs him .03 runs per PA if he intentionally loses the platoon advantage to the RHH. Can he make that up by holding the runner? Again, we'd have to know how he does in holding runners. Let's say that as a RHP, he allows fast runners to have .32 SB and .08 CS. As a LHP, maybe he brings that down to .21 SB and .07 CS. That seem reasonable? I dunno, but let's keep going. So, he saves .11 SB (that's worth .020 runs), but costs him .01 CS (cost of .005 runs), so a final gain of +.015 runs.
Therefore, it doesn't seem that he can make up for it, but, we really need to know what his platoon splits are.
Tuesday, May 26, 2015
First and third with Trout, two outs bottom of the 9th. A hit wins it, a walk gives the next batter a chance. So, one of the very few times we care about batting average. Trout will get a hit 30% of the time there's no walk, and he'll get a walk 12% of the time. If Trout walks, then it's on Pujols who will get a hit or walk 32% of the time.
But, this is against Kimbrel. So, maybe Trout gets a hit 25% of the time when there's no walk, and he'll still walk 12%. And Pujols will get on base 25% of the time against Kimbrel.
So, Trout hitting means: (1-.12) x .25 + .12 x .25 = 25% of the time Angels win, 75% goes to extra innings.
If Pujols hits: 25% they win, 75% goes to Extra innings.
So, same thing.
***
Interestingly, and historically, 1st and 3rd 2 outs has been a 69% chance of winning, compared to bases loaded of 68%. That's in the 9th inning.
But in extra innings, it's 66% with 1st and 3rd, and 68% with bases loaded.
***
All to say, it's pretty much a pick-your-poison kind of situation, and you can look at Trout and Pujols specifically against Kimbrel, a very extreme kind of pitcher. It's possible that one guy matches up very differently against Kimbrel than the other.
***
Anyway, that's enough to get some aspiring saberist to take this to the next level. Either that, or we get MGL to run his simulator against these very extreme matchup types.
Saturday, May 23, 2015
?I was asked this question on Twitter. The game is in extra innings, and the home Marlins had a runner on first (and two outs). A single was hit. This could have left runners on first and second and Stanton batting. But instead, the runner on first went to third, making it first and third, with Stanton now being IBB to load the bases for Prado. We could have figured Stanton would have been walked with the open base, so, what should the runner have done? We could consult the charts in The Book. But, let's try to figure it out.
First thing, we need the leverage index. First and second, 2 outs, tied game, bottom of extra innings is 4.4. That means the impact of an event is 4.4x what it would normally be.
The second thing we need is the win expectancy with an average batter. That's .613.
The third thing we need is Stanton's talent level. That's about +.006 wins per PA. (Interlude: I'm using a shortcut here. We REALLY want to know the component breakdown for Stanton, but I'm watching hockey right now. I'm going with .006 and someone else can figure it out.)
Now, we put the three things together, and the chance of Marlins winning with Stanton batting is .613 + .006 x 4.4 = .639. (Interlude: you always get someone saying about win expectancy how it's all about everything average happening. Well, yes, that's true, but as I have just shown you, you can use that as the base, and figure it out for a great hitter. Don't let people who give up stop you from finding a path to a solution.)
Now, how about bases loaded and Prado hitting? Repeat from above:
1. LI is now 6.4
2. WE for bases loaded is .662
3. Prado is close to league average hitter
So, this is an easy decision: the runner DEFINITELY should take the extra base, even if that means taking the bat out of Stanton's hand.
However, the reader asked how bad a hitter would you need to NOT take that extra base, and leave Stanton batting?
In that case, you do .662 - .639 = .023 wins lost. The runs per win converter is 10/6.4 = 1.56, so .036 runs per PA below average. That's about 70% of league average, or, close to the worst hitter in baseball (outside of pitchers). So, if you somehow had the worst hitter in baseball on deck while the best hitter in baseball was at the plate (has this ever happened?), then yes, the runner should NOT have taken the extra base.
Sunday, March 08, 2015
One of the great joys of my baseball life is that I get to email with Pete Palmer and Bill James. It's not every day that the two guys that are most responsible for inspiring one on their journey end up being in that journey. I'm extremely lucky.
The Hidden Game of Baseball is THE book that has set me on the path that I've been on, with Baseball Abstract propelling me further. Before you read The Book, before you read any baseball book, The Hidden Game of Baseball is what you should read. And it's being re-released. Pete said there is going to be nothing new, other than an update to his "Top 500" list. Regardless, his book is timeless, as relevant then as it is now.
***
He did however share some unrelated research with regards to the sac fly. Using Retrosheet data, he breaks down the flyouts by batted ball depth, and 0/1 outs. He then checks to see how often the runner was sent (goRate). And of those that did go, how often they were successful. Virtually every deep flyout resulted in a "go", and virtually all were successful.
Those at the medium depth went around 80% of the time, with a success rate of around 97%.
Pete points out that the interesting one especially at short depth, where the go rate was 21% with 0 outs and 29% with 1 out, as the breakeven point is far lower with 1 out: Pete goes through the run expectancy chart and shows it's around 75% with 0 outs, and half that with 1 out. One would think the go rate should be much higher with 1 out, not just somewhat higher. The success rate was barely any different at 92% with 0 and 90% with 1 out. Again, we'd expect the success rate to be much lower with 1 out.
It was terrific stuff that Pete shared with me. I asked him if I could post it on my blog, and he said yes.
Thursday, January 29, 2015
Five years ago, I had a pair of threads at Fangraphs. Someone linked to them, and I quite enjoyed re-reading them. Anti thread and Pro thread. I think the reason that pair of threads worked is because I basically kept each side away from each other.
Re-reading those threads, and I didn't have to ?suffer through shouting. They were very civil, very respectful, and both sides got to state their view, without needing to compete with each other.
About two years ago, I offered an idea on revamping the substitution rules.? I cleaned it up a little, and posted it on Fangraphs. There were a couple of good comments there. Here were my responses:
***
You add about .12 runs each time the PH takes over from the pitcher.
If a pitcher normally bat 2.5 times per game, but let’s say with this new rule he might bat 1 time per game, that’ll increase runs (in the NL parks) by 1.5 x .12 = 0.18 runs per game.
***
Right, that’s why I figure you’d still see the batter at the plate once a game. If they are up by 3+ runs, pitcher might still bat, etc.
***
Excellent question. The rule as I’ve stated would indeed allow for that, since this would just be a defensive sub.
Obviously, you wouldn’t see that in practice with a starting pitcher, but I can certainly see that with a relief pitcher. Say for example you bring in your regular setup guy in the top of the 8th, and he gets the first two hitters out.
Then you bring in your LOOGY to face Ortiz. He gets Ortiz out. End of inning. At the end of the inning, the manager has to choose which of the two pitchers to knock out of the game. He’ll knock out the LOOGY.
At all times though, it’s always a choice of the last two players. In the above example, when the setup guy replaced the starting pitcher, it was those two guys that the manager would have had to make a choice. Once the LOOGY came in, it became between the LOOGY and the setup guy.
The danger for the manager is if the LOOGY lets Ortiz get on base, and there’s a RHH now at the plate. The manager would then replace the LOOGY with yet another reliever, thereby automatically knocking the original setup guy out of the game.
Having said all that: if you want to exempt the pitcher-as-fielder from this rule, I’d be ok with that. I’d prefer not having an exemption, but I also wouldn’t want this rule to become a focal point of strategy every time a pitcher gets relieved.
***
That is a good point, thanks for thinking it through.
We could do two things:
1. I posted at the bottom in these comments that we could exempt the pitcher-as-fielder from the sub rule. So, once a pitcher is swapped out for a reliever (as a defensive sub), then he’s out.
2. Allow the “option c” only once per player. Suddenly, you couldn’t PH for the pitcher each time he came to bat and allow the pitcher to remain in the game. You could only do it once. This is the same for any player on the field or at bat. You couldn’t always protect say Adam Dunn in LF or John MacDonald at bat. They get just one freebie per game of being replaced without being knocked out.
Just a couple of ideas so that the rule itself doesn’t become too central to the strategy.
This would mean therefore that a pitcher, who normally comes to bat 2.5 times per game, will come to bat 1.5 times per game.
This puts pressure on the manager to decide when he gets the freebie of PH for the pitcher, while letting the pitcher remain in the game.
Friday, November 07, 2014
?A terrific piece by Jeff.
Friday, August 01, 2014
?Way back in the early 1970s, he was making his mark with the decision makers.
In each game, a reliever should start, go two or three innings and come out for a pinch hitter his first time up. He would be followed by a starter type, who would go about five innings, batting the first time he came up, but going out for a pinch hitter his second time up. Another short man would finish the game. This sounds revolutionary, but essentially the only difference is that the "starter" appears in the middle of the game instead of at the beginning.
Sunday, June 22, 2014
?Seeing the lines spray painted for free kicks is a genius move. It's such an obvious thing once you see it, you wonder why it wasn't invented 20 years ago. Even when players try to stand in front of the line, the referee will immediately blow the whistle, and the defender will obviously get back to the line. (I don't know what happens if the defender AGAIN tries to cheat in an obvious manner. I'd LOVE to see a yellow card there.)
Anyway, is baseball best played when strong hitters do their best to swing at balls, and not bunt to keep the defense to play them honestly? Or, are we ok that we need to let this play out organically, without legislating behaviour?
I remember when we played on our company team, one of my buddies was a HUGE pull hitter. And I did something terrible to him: I told the 3B to go play RF-rover. My buddy was VERY upset at me. And he wanted to show me up, so he did his best to hit opposite field. Which to him meant hitting the ball just to the left of the 2B bag, where me as SS was waiting right for him. We never played him like that again. We all felt it was fun/funny the first time, but it would be very unsporting if we kept doing it.
Of course, this is MLB. Is it fun or serious? Could we put a line marker halfway between 2b and 3B and ensure that at least 1 fielder plays between this line marker and the 3B foul line? (We don't have to do this at 1B, because the 1B will always be nearby.) For some reason, the OF is different. We don't seem to have a problem how they are positioned. And then it hit me: MGL noted that almost EVERYONE tries to hit it to the OF, while the IF gets the "mistakes". So,, it's almost as if we're ok in fielders doing their utmost in positioning their OF wherever they want, but for "mistake" hits, we don't want all the fielders there waiting to get the out, that even in mistakes, hitters should have a decent chance of getting on base.
So, what do you think? Is it a pain to watch the massive overshifts in the infield? Or, it's just the cost of doing business?
Friday, June 06, 2014
Note: I've talked about this a few times, but, I won't let that stop me.
Under current substitution rules, once a PH comes into the game, the original player is knocked out of the game. The PH himself can either be brought into the game, or yet another sub takes his place. So, in a typical NL scenario, the PH comes in for the starting pitcher, only for himself to be replaced by a relief pitcher.
Let's say we change the substitution rules so that after a PH event occurs, the manager has the option, at the end of the half-inning, to either proceed as usual, or to knock the PH out of the game, and bring back the original player.?
First: will anyone object to this rule change? It seems benign enough in terms of affecting the "sanctity" of the game.
Secondly, it addresses the most embarrassing part of an MLB game, and that is being forced to watch a pitcher bat at a crucial moment.
Thirdly, it does that by not introducing the "abomination" of the DH.
Fourthly, it adds strategy to the manager, since he won't ALWAYS have a PH to bat for the pitcher, otherwise he'd burn through his bench pretty quickly. In a bases empty, two out scenario, he'll likely let the pitcher bat. And in situations where he'd be happy to sacrifice the runner over, he might let the pitcher handle that as well.
So, everyone wins here: We significantly reduce the number of times the pitchers bat; We eliminate the madeup "position" of the DH. And we do all that by a rather simple rule change that works fairly well within the normal baseball rules.
***
We could even expand the change in substitution rules to affect defensive subs. If we wanted to bring in Brendan Ryan for one half inning, and then decide whether to keep him or Jeter for the rest of the game, the option is now there.
***
So, tell me, what is it you don't like about this plan?
Thursday, June 05, 2014
Terrific stuff from him. Definitely saber-leanings with this statement.
“‘Shutdown inning’ is one of my favorite phrases,” he said laughing. “They’re like, ‘Come on, shutdown inning!’ and I go, ‘Yep, this inning is way different than any others where it’s OK to give up runs! This inning we really don’t want them to score!’?
And then right after, he completely loses me.
“I always laugh at the term shutdown inning, but it is. It is because of the momentum of the game. People are crazy if they don’t believe in momentum because there is such a thing. The team that has momentum, it seems like the good things go their way.
Unfortunately for Peavy, he is mostly wrong. To the extent that he might be right, it would be pretty imperceptible. Maybe momentum changes a 50/50 situation into 51/49 or something.
Thursday, May 29, 2014
?Good stuff from Zach.
Recent comments
Older comments
Page 1 of 150 pages 1 2 3 > Last ›Complete Archive – By Category
Complete Archive – By Date
FORUM TOPICS
Jul 12 15:22 MarcelsApr 16 14:31 Pitch Count Estimators
Mar 12 16:30 Appendix to THE BOOK - THE GORY DETAILS
Jan 29 09:41 NFL Overtime Idea
Jan 22 14:48 Weighting Years for NFL Player Projections
Jan 21 09:18 positional runs in pythagenpat
Oct 20 15:57 DRS: FG vs. BB-Ref
Apr 12 09:43 What if baseball was like survivor? You are eliminated ...
Nov 24 09:57 Win Attribution to offense, pitching, and fielding at the game level (prototype method)
Jul 13 10:20 How to watch great past games without spoilers