THE BOOK--Playing The Percentages In Baseball

I can’t go there.

But I’ve been thinking about the old Bill James idea of hitting your best hitters in order a lot lately (i.e. Alex Rodriguez hitting leadoff). Any perspective on how that would actually work based on any work done here? Fans would go nuts on that one.

Meh, seems silly and pointless here (this is a rule that will never change and is one i wouldn’t want to change).

It does remind me of an NL point though that someone once mentioned to me.

Why do away teams in NL games start the game with the starting pitcher for that game in the 9 spot?  If the team bats around, the pitcher gets up and generally wastes an out. 

I would start a starting pitcher who pitched the game before in the 9 spot.  If the team bats around, PH for that pitcher, and you have a chance of increasing your lead.  If a team doesn’t bat around, when you come up to pitch, just sub in the real starting pitcher. 

Sure this would probably matter maybe only a few times in the whole season, but i don’t see the reasoning against it.

I know MLBPA would never go for this, but I’ve considered a solution to the DH/noDH debate.  The big argument for the proDh crowd is that pitchers are not really fielders and no one wants to watch them hit, therefore it makes sense to exclude them from the lineup.  I can agree with that logic.  But, it seems the answer of a DH counters that logic.  So, no DHs, no pitchers batting.  The batting line up is the 8 position players only.

Try to adhere to the instructions here.  If you can’t play for fun, don’t play.  It’s really that simple.

Greg asked a specific question, a what-if scenario that doesn’t ask us to consider if it’s logistically feasible from the standpoint of MLB.  He’s asking us to simply consider the scenario on its own.

I have to always add a provision that I understand that baseball fans treat baseball differently than any sport fan treats any other sports.  Please, set aside your biases, and stick to the scenario.  And if it’s really that hard for you to do that, then don’t play along. 

Don’t tell me that you don’t want to play along!

Some of you guys would never be a lawyer for a murderer, that’s for sure!

@garik16: The only problem with slotting in a different pitcher is that, IIRC, a pitcher who enters the game (or is in the starting lineup, of course) must pitch to one batter.  This would force that visiting team to have their fake starter throw at least a pitch.

Bases full, nobody out. Pujols strikes out. Then the manager choses to insert Pujols as a batter again. No thanks.

"Bases full, nobody out. Pujols strikes out. Then the manager choses to insert Pujols as a batter again. No thanks. “

But that could only happen if the manager had saved Pujols for the ninth position in the lineup.  How often is that going to happen?

It would be pretty interesting.  Seems like someone could play a few Strat-o-Matic games and see how that works out.  Hmm, I’ve got the 1990 season at home and my older son’s getting old enough to play it…

To be closer to other sports, allow substituted players to come back in at any point in the game.  Allow managers to decide who comes to bat at any point in time - no set lineups.  Maybe have a restriction that the same player can only bat once an inning or he can bat a second time if 8 others have also batted. 

It would lengthen games I think, so instead of 9 innings, cut it down to a 7 inning game.

I love the idea, I think it adds more strategic thinking to the game without changing much. From what I understand, with this system offense would get more effective and score more runs. What sort of run difference are we talking about here? maybe 0.1-0.2 per nine innings? Is there any way a simulator can run this type of thing?

Goroyals: you can’t envision a provision to add to Greg’s list that in addition to the batting once per 9 slots, that he can’t come to bat again until at least 3 other batters come to bat?

This is definitely not fun for you, so, please, go try some other thread, and let others have their fun.

This sort of reminds me of Fischer chess (where the starting position of all pieces is randomized instead of lining them up as described in the rules every time).  Fischer felt that the game’s strategy had become stagnant and that abandoning the rigid starting position would force more dynamic strategies and better determine the winner based on who had a deeper understanding of the game and could adapt his strategy to fit unfamiliar positions.

I don’t know that I would go for the change, though.  I like the idea of having to work within the current restrictions.  I like the nature of run scoring where it often requires a chain of the right events and successes, and that teams often need to keep plugging at it and building up small successes before they finally break through on the scoreboard.  Making those events easier to chain together adds more excitement and run scoring, but those are not things I relate directly to enjoying the game.  I think there is a balance where making the events too easy to link and making run scoring a more common occurrence begins to detract from the things I enjoy in watching the game, and I think the current offensive environment is already close to that.

Another thing I think would be compromised is the unsung hero.  You can have Yadier Molina hit a huge home run to win the game, and that’s part of the charm of the game, seeing the unexpected hero come through in a key situation where he would just be passed over for another hitter most of the time under this rule.  I also like the idea that my team’s pitcher can get bailed out of a tough situation if he has the bottom of the order coming up.  It might be cool to see Albert coming up whenever a huge situation arises, but at the same time, as a fan, I can see the strain of knowing every time your pitcher gets in trouble, he’s facing the gauntlet taking some enjoyment out of the game for me.

Beyond just the tradition involved (which doesn’t really enter into the constraints of this discussion), I really like the pace of the game as it is, and the balance of exciting edge-of-your-seat situations against the normal flow of the game.  I think the ways this rule changes the dynamics of the game in forcing more exciting situations would actually detract from my enjoyment, because part of baseball that I like so much is that normal flow of the game between the high points.  There are some cool strategic situations that would come up to add interest, and maybe it’s something I would get used to, but I prefer the current lineup rule.  It’s certainly an interesting thing to discuss, though.

I will do an experiment to estimate the impact of such a change on scoring (it’s pretty clear it would increase it).

This year the AL scores 4.86 R/G. The team that’s closest to that is Texas, who score 4.85 R/G.

I’ll take their players that have had the most PA for each position. That gives me:

C Jarrod Saltalamacchia
1B Chris Davis
2B Ian Kinsler
SS Elvis Andrus
3B Michael Young
LF David Murphy
CF Marlon Byrd
RF Nelson Cruz
DH Andruw Jones

I’ll use a OBP-sorted batting order, which gives:
Young, Murphy, Cruz, Jones, Andrus, Byrd, Kinsler, Saltalamacchia, Davis

Ok, using that in my simulator over 1 million games gives an average R/G of 4.91. That’s close enough, I suppose.

Now let’s add the rule change. So now instead of batting in order, you can reorder the 9 players until they’ve all been used. Since it’s not super obvious what rule we should use for this, we can try a few.

1. Random players… each time the batting order is randomized. We expect this to be worse than the OBP-sorted batting order. Sure enough, we get 4.86 R/G.

2. Simple rule: if someone is on base, choose the best player by OBP, else choose the worst available player by OBP. I don’t have much intuition about this one, but perhaps it should be better.

Running the simulation gives 4.98 R/G. So even with this simple rule that’s almost 0.1 R/G increase. But could more complex rules do much better? I’m not sure. For each specific situation it wouldn’t be too hard to compute the best permutation of the remaining players. That could give some more benefit.

Overall I don’t think it would be a huge increase. Maybe with the best optimizations 0.2 R/G extra, which isn’t huge, considering the difference in the AL in the past 10 years between the largest and smallest has been over half a run per game.

Fantastic stuff!

I wouldn’t want to see this done to baseball, because it’s too drastic a change I think. That said, I LOVE the idea and the mental exercise of it all.

If I were a billionaire looking to start up my version of the XFL/UFL in baseball, I would think about trying this. It’s a gimmick, sure, but it would make my new league unique.

The strategic elements seem fairly endless, but I wonder - wouldn’t, most of the time, you end up batting a fairly conventional order with the exception of some guys near the bottom? The restrictions, which I do believe are logical and necessary, almost make it to there won’t be too many times where you’ll be able to make a major switch (Pujols for the pitcher) at a key moment in the game. It would most often be used for L/R matchup purposes, I think. I would love to see some more work on this though.

I think this would be awesome.  It would making choosing and organizing a lineup a much more important task than it is now.

Andy, I think the idea behind the post is that you would be able to maximize runs scored situationally.  So I don’t neccesarily think a sim is best for this. 

You could really manage!  Down by 1 with a runner on third?  Put in your best contact hitter. 

I LOVE this idea!

Dave Trembley’s head just exploded.

While I think this would be incredibly unwieldy and unworkable in practice (can you imagine how long a playoff game involving Joe Torre would take?), I love the outside the box thinking.  This would definitely make for lots of fun second guessing in-game moves.

I think the time constraint is legitimate.  You could make the rule that you have to send the batter that is on deck to bat.  So, you have to make the decision on batter ahead.

Furthermore, you can even make it that the guy on deck that ends the inning has to leadoff.  That could add yet another dimension, as the guy on deck with 2 outs, likely your power hitter, may end up leadoff off the next inning if the guy at bat doesn’t get on.

Right, I like the outside the box thinking, basically asking, if you were to start baseball from scratch, how would you design it?  Obviously, the 1893 rules are not the ones set in stone as being so perfect.

Kincaid/11 was marked for moderation and is now open.

I like his two points regarding:
- unsung hero
- ordered flow of the game

In hockey, you have 4 forward lines and 3 defensemen pairs.  Suppose that you always had to rotate those guys all the time (even during PP and SH time), and you can only pull the players after 30 seconds, and you can’t leave them on for more than 90 seconds at a time.  What would happen?  Well, you’d get the structured flow, but is that what you want out of hockey?  This is the way it would work in house leagues and kid leagues (everyone gets the same ice time).  The not knowing of when your top players come in is the cool part.

Would you want that in baseball too?  As noted, the unsung hero aspect is fantastic in baseball.  It works in baseball, and not in hockey, because of the individualized nature of the sport.  Indeed, baseball is all about giving each kid an equal chance, and always letting the big guy having the higher leverage opps doesn’t lend itself to that.  Every boy/man facing his own battles.

So, I concur with Kincaid that given the choice, the ordered batting lineup is one of the appeals of baseball, much as the batter/pitcher matchup.  I’d therefore say, in terms of what makes baseball baseball, it’s the following, in order, so far:
1. pitcher faces batter
2. each fielder has a set role
3. the batting lineup is ordered

While once in a blue moon, you will see the LF/RF swap spots, it’s not really something managers want to do.  If they were smart, they would.  They happen to not.  And I think they don’t because they understand that fundamentally, they shouldn’t on a sportsman level.  That, and they are too exposed to everyone that that’s what they are doing.

Let’s say you started with that top-down OBP lineup, no matter what happens, wouldn’t you send up the guy with the best chance of getting on base next?

The only difference would maybe be two outs and nobody on, but the run-expectancy, I would think, would still be greatest if you sent up your best hitter in that situation, not your worst.  Otherwise, you are giving away at-bats. 

What would be the run expectancy for the next inning of using your worst hitter with two outs and nobody on?  What would your run expectancy be if you used your best available hitter?  I don’t think it changes that much, but I would say use the best possible hitter in all situations.  The key is to get on base.

My alternative to the DH was to simply allow the manager to send in a substitute hitter in the lineup, but take some other player out of the game, not the player substituted for.

So the manager can send in a pinch hitter for the pitcher, but the pinch hitter substitutes for the third baseman, not the pitcher, who stays in the game.  You don’t see pitchers batting, but there is no such position as “DH”, the existence of which I understand is the main objection to the DH rule.

I like the free lineup idea too, with the provision that the guy on deck must bat next, to keep the game from dragging.

The 8-man hitting lineup will do away with Edgar, Ortiz, etc.  I don’t know that that’s necessarily good, since really, what’s the difference between Ryan Howard and David Ortiz?  Is it really necessary to put those guy on the field at 1B?

You could make it an optional 8-man or 9-man lineup.  If your bench hitter is better than the average hitter in your lineup, you go with a 9-man lineup.  If not, you go with an 8-man.

You can also introduce a “designated pinch hitter”, where you can bring in a hitter into the game, without taking a batter out.  So, this dPH hitter sits around waiting for the opportune time to come into the game.  He is inserted in-between the two spots for this one PA only, and afterwards, he’s out of the game. 

You can do this all-game long if you like, with new hitters.

The other alternative for the DH is the “home manager discretion”, and let the home manager decide if that game will be played with a DH or not.

Great premise!

#20 Do you send your best OPS hitter up 1st automatically, or do you wait and have Pujols bat with runners on? You’re guaranteed 3 plate appearances in the 1st 27, so should you wait until you have runners on if possible?
Worse case he bats with no one on 2 outs in 9th spot - best case he is batting with runners in scoring position way more than before.

I think he did that in #12, and is shows a slight advantage

I’m sure everybody want to hear this, but if Pujols came to bat with men with men on 80% of the time, he could get 190 rbis.

Just what we need, another way that managers can show us how dumb they can be!  Can you imagine the threads on how such-and-such manager did this or that?  Not to take anything away from this discussion - it is a good one - but I’ll take the least amount of potential strategic decisions by a manager just to keep my head intact!

This is a really interesting idea, and I would love to see it implemented, but can you imagine the clutch debates that would spawn from this?

Well, the best hitters usually hit up in the batting order, so I don’t think they could get any more than 1 extra at bat (as long as the order is slotted by 9). It might increase fan excitement in tense situations. However, I think it would increase the amount of pitching changes (if that’s even possible anymore) as the dangerous hitters will be coming up in more RBI situations. I think Greg wanted to reduce specialized relievers, but I think the opposite would happen.