THE BOOK--Playing The Percentages In Baseball

We’d have regular 2 hour games.  The complete game would become as common as it was in Philadelphia 2011.

While a pitcher should be able to pitch a 9 inning game on about 75 pitches, I think a few outstanding relievers would still be used - you’d still want Mariano out there for the 9th.

No, scratch that, Mo could easily handle the 8th and 9th 65 times per year and 6th-9th with extended rest in the playoffs.

You could probably go back to a 4, maybe even 3 man rotation.  Maybe 2-3 of the best relief specialists, plus a long man to make up an 8 man staff.  That leaves 17 position players - bring back a few platoons, plus more pinch hitting/ defensive replacements depending on game situation.

Sign me up.

There are tweaks that would make all of the major sports a little more watchable (which soccer could especially benefit from), and this is a pretty good/equitable/logical example of one. The only issue is, where do you stop? It kind of opens the door to make any change that is logical and could improve the game… and while that actually sounds kind of great in theory, if I had to choose between baseball as is, and baseball with a half dozen tweaks to make the game quicker/smoother/better, I’m really not sure what I’d pick… but I know I’d lean towards the existing game.

We’d always lean towards the existing, for the same reason we want Vlad to be Vlad: he’s proven great, just as baseball is proven great with a 4-3 system.  And the room for improvement is smaller than it is for the room for disaster, which is why changes are always slow in coming.

This is why we can test these things in lower levels.

I agree with Rally’s assessment of the great benefits.

We talked about this a while ago, about what makes baseball “baseball”.  Basically: pitcher, batter, and 4 bases.  Everything else is negotiable.

One question I’d have is to what extent would this cause the percentage of plate appearances that result in walks or strikeouts to increase.  As has been written here before, more balls in play are generally good.

There was a great article about a year ago, by I think Sky at Baseball Analysts, that looked at the question of what would happen if everyone started at a 1-1 count.  You can find it there, or somewhere in this blog.

Tom, do you happen to play on a rec league softball team?

I ask because the 1-1 rule is fairly common in lower level softball leagues. The purpose, as you might guess, is to shift a little advantage back to the pitcher, who is otherwise at a fantastic disadvantage in slowpitch softball. Lower scoring means games are less likely to exceed the scheduled 1 hour per game.

As a hitter, I will say that starting the count at 1-1 definitely puts you on your heels, even in slow pitch. If you let a first pitch strike go by, suddenly you are forced to hack at nearly anything near the plate.

Interestingly, I’ve never heard of a league describing it as a “3 ball, 2 strike” count, but rather always that the count starts 1-1.

I’d love the lower count to speed the games and reduce the size of the pitching staff.

Rally,

Why would your pitchers be able to pitch more often though? If they’re going deeper into games, wouldn’t they still be throwing about the same number of pitches?

Tango,

Down mistakes in football probably aren’t as uncommon as you think. In 1990, for instance, Colorado won a game against Missouri thanks to a fifth down. Colorado went on to finish the season ranked #1 in the AP poll, which is why that game is still remembered 20 years later. I can’t think of any obvious examples from the NFL, but I’d guess they exist. At least prior to replay. We just notice counting mistakes more in baseball because there are exponentially more opportunities for them to happen.

"Why would your pitchers be able to pitch more often though? If they’re going deeper into games, wouldn’t they still be throwing about the same number of pitches?”

I don’t think they’d throw the same number of pitches, even if they go 9 innings.  I think 9 innings would usually arrive before 100 pitches.  Not 100% sure though.

This would definitely have a larger negative effect on hitting beyond starting each at bat at 1-1.

A real 1-1 count means the batter has already seen two pitches. Starting at that count cold will be a big disadvantage. If you believe Ted Willians, a big reason that batters take so much on 0-0 is that they want to get calibrated against that pitcher. Williams would take until he saw the fastball - with certain outstanding pitchers excepted.

I would also expect that the times through the order penalty against pitchers will be lessened as well. Still hoping Tango will do a per-pitch study of that.

While I’m generally favorable to the idea I think the biggest knock would be that it would upset the balance of talent. Patient hitters would be more affected than hackers.

If I were king, I’d start a new league with 3-2 counts and 8 defenders. It’d still be baseball, but the type of player who would excel in that kind of game would be a lot different than MLB.

#10 Matt—I was just thinking about the same thing.

Another thought --

Perhaps instead of starting pitchers working a lot further into games (ie, an explosion in complete games), they would throw with..ah...greater effort? Much like the same pitcher will add 1 or 2 mph to his fastball when moved to the bullpen (with a corresponding increase in K/9 and generally improved numbers across the board), perhaps starters still only work 6-7 innings deep on average, but with better stuff, further decreasing offense.

I like the idea of a change that will speed the game up without much cost.  (I view a decrease in offense as an advantage: epee good, saber bad; soccer good, basketball evil.)

Another advantage of reduced scoring: more extra-inning games.  (The early innings will be faster, but don’t expect close games to be shorter.)

I wouldn’t be surprised if the change allowed more complete games for ace pitchers, but don’t expect pitching staffs to shrink in favor of platoon players.  Teams already carry 4-5 outfielders, but many teams (even with the roster slot already allocated) don’t have anyone worthy of putting in a platoon.  Where are the extra, high-quality bats going to come from?

If you don’t need as many pitchers to eat lots of innings after the starters go away, isn’t it also likely that the roster slots will be used to maintain platoon advantage?  My prediction would be that you would see the average team carry the same number of pitchers, but that fewer of them would be right-handed.

With more mid-inning pitching changes, the advantage of faster games is going to be eaten up really fast.

only allowing 1 two strike foul ball is another way slow pitch softball rec leagues speed up the game. thats always been in addition to starting the count at 1-1.

it works well for softball, but so does having 4 outfielders.

#13 kendy—Yup. My league actually allows no 2-strike fouls (eg, any foul ball with a 2 strike count is an automatic out). And 4 outfielders, yes. Also, a full game is only 7 innings.

And despite all this, games regularly end with football scores.

Taking out two pitches from every at bat is essentially giving us “less baseball”; so if we’re entertaining wacky solutions here, why not just cut back the schedule, and instead of 162 games, have everyone play 150 or 140?  That’s a different way of cutting back on the amount of baseball we all get to watch…

With more off days, you would have fewer weak pitchers on the mound, less need for 8-deep bullpens, less need to rest superstar everyday players on game days, and an overall increase in the level of play that comes with concentrating the innings played in the better players. 

I’m not really advocating either of these two ideas; in fact, I don’t think anything needs changing, but of the two, cutting back the schedule is vastly preferable, to me, than screwing around with the core of the game (the pitcher-hitter confrontation).

Rally #1: can you explicitly specify what problem you think the 1-1 change would solve, and why it needs solving (i.e. make a solid case against the status quo)?

If you want less baseball, just skip a few games here and there.  You don’t have to watch every game.  Or you could simply skip a few innings.  They’ll still let you in the park during the 3rd inning, you don’t have to get there for batting practice.

"Rally #1: can you explicitly specify what problem you think the 1-1 change would solve, and why it needs solving (i.e. make a solid case against the status quo)?”

I never said anything needed to be solved.  I would prefer 2 hour games to 3 hour games though.  And I’d prefer a full schedule of 2 hour games to fewer games lasting 3 hours.

I love this as a way to increase action on the field, create more ‘dramatic’ counts, and speed up the game. 

Having said that, I would prefer a 3-3 count over 3-2.  I’d fear a two strike K would decrease action in the field by creating too many strikeouts.  Pitchers would probably strike out 20 every game.

It’s not a question of less baseball.  It’s more a question is if it’s necessary to have a 4-3 system.

As I said, 70% of first pitches are takes, while all other counts are not like that (other than the obvious 3-0).

***

Also, starting at 1-1 doesn’t mean there’s 2 less pitches per plate appearance, since some plate appearances end on the first or second pitch.

Indeed, in modeling it, starting at 1-1 instead of 0-0 results in almost exactly 1 fewer pitch per plate appearance. 

By my count, we’d go from 39 PA of 3.8 pitches per batter to 38 PA of 2.8 pitches per batter.  So, a game would go from 148 pitches to 106 pitches.  Hence the reason Rally says we’d see tons of complete games.  Or, as he noted, Rivera could come in for 2 innings (8 or 9 batters at 2.8 pitches means 20 to 25 pitches).  And that means you have starting pitchers who can go even a bit more all-out for 7 to 8 innings (80 to 90 pitches).

I would prefer something in-between, at 3.3 pitches per batter, as the “best”.  In order to do that, I’d have to start the count at 1-0 or 0-1.  0-1 is an obvious non-starter, because that makes it a 4-2 system (4 balls for a walk and 2 strikes for a K).  Starting at 1-0 (or making it a 3-3 system) is appealing, but will lead to a pretty high run environment.

***

The biggest “problem” I have is the fungible nature of the bullpen: it’s just a rotating group of pitchers who offer little differentiation among each other.  It’s not like a bench player, who can offer something unique (power, speed, defense, etc).  Instead, it’s one power arm or another power arm, neither who can locate a pitch.

Of course, with 106 pitches for a game, the opposite can happen, and you just get a bunch of two-inning pitchers for a game.  They each throw 25 pitches, and 4 pitchers later, you’ve got a game.

I’m curious: for those who prefer 2 hour games to 3 hour games: why? Don’t you like baseball? Isn’t 3 of baseball better than 2?

The idea of 8 fielders is also interesting, as it would be a little like football, where the defense has to decide whether to defend against the run or the pass.  So, do you move your fielder to the middle infield or the corner outfield after each batter (or even go into a rover position).  Heck, make it 10-fielders, but start the count at 1-0.

I find it fun when they do the shifts already.  This would add something extra to that.

Does anyone know of any organized baseball leagues that feature alternate/experimental rules? Something like the And1 league is (was?) for basketball, or the Arena Football League is for football?

I think two hours must be some sort of limit for sitting on your butt. The typical movie is about that long.

Also, I wouldn’t have to leave in the seventh inning when my kids are ready to go home.

I know someone will say how our attention spans have eroded etc, but games used to be much shorter; PPA’s were much lower.

"Isn’t 3 of baseball better than 2? “

Again, it’s the dead time.  70% of first pitches are takes.  By the logic that 3 is better than 2, why not 4 is better than 3, make the count system as 6 balls and 5 strikes, and see 90% of first pitches as takes?

So, forget about the amount of time it’s legally called baseball.  It’s the amount of action time, confrontation time.  One of the worst counts is 3-0, because over 90% of the time, it’s a take.  Over half of 3-0 pitches are called strikes, and 35% are called balls. The 3-0 count is a terrible count to enjoy baseball.  The 0-0 count is a bad count.

The better question to ask yourself is: how many pitches do you want to see per batter?  Is it closer to 3 or closer to 4?

When I was a kid, I’d go through the Expos schedule, and circle all the days where they played a double header.  Those were a treat, I was getting a 2-for-1 deal on the price, and I was getting 2 ball games in one day.  I’d go to those games.

Now? I would try to avoid those games.  MLB has turned a “let’s play two” into a “let’s avoid two” scenario.

My favorite thing to watch on TV is triple-header Olympic hockey: 6 hours, 3 games.

It’s all about the action/confrontation time.

Maybe part of the reason I can’t quite fathom the desire to shorten games is that I never got tired of looking at the grass at Fenway Park.  No game there was ever too long for me. 

Nowadays when I watch a game on TV, I almost always have something else handy (book, computer, etc) for the pitching changes and commercials, or if I’ve got it on the DVR, the delays are minimal.

I know this is 99% thought exercise, but I still like to approach these ideas from the standpoint of status quo, meaning the very first thing you have to do before introducing a change is make a credible, strong case for why the status quo is unacceptable.  Then start thinking of remedies.

And the word remedy is apropos: changes to the game should be remedies, i.e. fixes for something that is wrong, like instant replay for boundary calls (but not robo-umpires, God forbid).

We ought not pursue change for the purpose of making something already good better (dare I call such a thing a “supplement”?)…

2010 slash line for AB’s after a 1-1 count.

.237/.306/.372

Hello 1967.  Next change is to lower the mounds???

I have always been interested in the idea of 3 balls is a walk, and 3 strikes is a strikeout (with 2-strike fouls counting as strikes). This would make the number of walks more similar to the number of strikeouts. Why should there be twice as many “end of the count” results that favor the pitcher, as there are now? Also, it would set up a bit more drama, with a definite “final pitch” to each PA. It would reduce some of the ‘nibbling’ that pitchers do. It would give a greater reward for pitch judgement skills by hitters.

Of course, there are always ‘unintended consequences’ to worry about. I suppose somebody could sort of test this idea (and others) with a simulation.

Dave: I can do that easily enough.

If you start everyone at 1-1, and keep the 2-strike foul, you have 2.73 pitches per batter, 21% of PA ending on a K, 9% on a BB, and 70% in play.

If on the other hand you start at 1-1, but there is no 2-strike foul (i.e., a foul ball is a strike call), then you get 2.33 pitches per batter, with 35% of PA ending on a K, 6% on a BB, and 59% in play.

As you can see, a huge difference!

***

Greg: I take the complete opposite approach.  You start with the status quo, and ask how you can improve it.  I go against the status quo, and ask why is that the right way to start with it.

Dave #28

I have to disagree - if you make a 2-strike foul the third strike, you will demolish the pitcher-hitter balance we currently enjoy.  Demolish, I say!

So, this is what needs to be determined:

- number of pitches per batter you like to see
- % of PA that end on a K
- % of PA that end on a BB
- % of PA that are in play

If you give me that number, than I can see what we can come up with.

7/Mark
I used to umpire for a 3 and 2 slow pitch softball league that didn’t start at 1-1.  Strike two, you’re out!  People caught on pretty quick.

Tango, why are you using 1-1 as the starting point when I am retaining 3 strikes for a strikeout? Am I missing something?

Greg, care to explain? To me, going from 4 balls to 3 for a walk while keeping strikouts at 3 will obviously increase offense. Making the 2-strike foul a K will help mitigate/decrease some of that offensive improvement.

#32 Jim A—Huh, interesting. I can’t decide if I think that would be easier to remember or not; I guess I’d have to try both for a period of time to see how easily I adjust.

This is only tangentially related, but talking about speeding up games reminded me of a particular umpire in my church league, who alternately calls out either “Need a batter, need a decker!” (when requesting that the offense hurry up), or “Got a batter, got a decker!” (when requesting that the defense/pitcher hurry up).

Or how about this:

Watch a game tonight or tomorrow, and make a note of each at bat you enjoyed watching.  Write down how many pitches each batter saw, and whether it was inplay, K or BB.

Compile that for your 10 or 12 favorite PA, and let’s see what you guys have got.

Dave: my fault for being sloppy.

Ok, so starting at 1-0, using current rules, that’s:
3.21 pitches
13% K
14% BB
73% in play

Removing the 2-strike foul rule, and we have:
2.96 pitches
22% K
12% BB
66% in play

So, your rule would retain the current K/BB balance we currently enjoy, but less balls in play (and less pitches).

Dave, removing the hitter’s ability (privelege) to fight off a tough 2-strike pitch will fundamentally change their approach to hitting, since it adds a massive disincentive (danger) to hitting late in the count.  If the hitter is facing a pitcher with a live fastball who is even average at locating it, they are dead meat once they get to two strikes, because where before there was balance (mistakes get hit hard, great pitches are strike three, everything else gets fouled off), now there is imbalance (pitcher wins on everything except a mistake).

Knowing this, hitters will have to try to hit the first pitch they can put in play, and they won’t be able any longer to wait for the first pitch they think they can hit WELL. 

Pitchers, knowing that the end of each at bat is drastically tilted in their favor, will be able to nibble more on early pitches, knowing that hitters will have to hack early. 

What you will end up with is a ton of weak ground balls to the middle infielders or lazy flies to the shallow outfield on pitch 1 or 2, as the hitters make the best of a terrible situation.

Forgot about the all-star game… my little experiment will have to wait.

Tango, so by your analysis, retaining the 2-strike safe foul will result in similar numbers of BB and K, with similar BIP numbers to what we have now, and with fewer pitches per PA. So, we are essentially exchanging some K for some extra BB. Sounds good to me. I quickly estimate that this would result in about a 1 RPG increase in scoring.  So, 5.5 RPG with only 3.2 pitches per PA, the same proportion of batted balls per PA, and more BB and fewer K. All sounds good to me.

Greg, the increase in hitter aggressiveness and pitcher nibbling that you expect should be mitigated greatly by the 3 balls per walk part of my proposal. Works against both of those expectations.

Two hours is the American entertainment sweet spot. Part of my motivation to change to the 1-1 count is to avoid losing the next generation of MLB fans, which I think is a legitimate concern that the league should be mindful of. Most people on this website did not grow up with 3-d video games, Facebook, etc. There is more competition for youth’s attention, and getting your kid to sit through a three hour game is really difficult. Sure, you could leave in the 7th inning, but isn’t a movie more satisfying when you don’t have to leave the theatre 30 minutes before it ends?

How about the regular 4-3 system but penalize batters for taking a first pitch strike.  Make any strike taken on a 0-0 count, count as two strikes.  Then you will have batter swinging at more first pitches - more strikes than balls if they have a decent eye.  There would also be more incentive for the pitcher to throw a first pitch strike.
vr, Xei

Xei: that’s interesting.  If it’s a called strike, it’s two strikes.  But if it’s a swinging strike, it’s one strike.

Would a player swing late, in the hopes of turning an eventual called strike into a swinging strike?

Since you are transferring a huge advantage to the pitcher, why not bring it down to a 3-3 system?

Tango: I would think that leverage would be a much bigger deal than the other stuff in such a small sampling. Maybe I’m wrong. I also think that more “baseball guys” are going to like a long duel between pitcher and batter of 15 pitches or something than the average fan or, especially, people who are apathetic to baseball. So it somewhat depends on who you’re targeting.
There is an issue that nowadays, a lot of the kids find baseball “boring” and think that there’s 2 minutes of action in a 3-hour game. I’ve actually had several tell me things along those lines. 2 minutes of action in a 2 hour game isn’t going to help much though, I think, especially because to them, offense = action.
I’m interested in how you’re doing these sims, because I’d think there would be tons of unintended consequences to such a change (approaches would shift a lot after a while, oh game theory, how I do love thee). I think a great place to test this out would be spring training. Heck, the all-star game wouldn’t be too bad either.

But my initial reaction to your question was “it’ll turn back the clock 100 years.” - in terms of style more than anything.

I think I’ll be an outlier here, for if I was to detail what at-bats I enjoyed watching the most, it is the ones where the pitcher and hitter have a real battle (>8 pitch at-bats). By that logic, I would prefer longer games, so bring on the 6 hr marathon games.

Fun thought experiment, but I really don’t see the problem with 3 hr games. The easiest solution to quicken the pace of games (if pace/length of game is even the issue here) is to get the umpires to call more strikes. Either expand the zone by making the plate wider or alter the height in which strikes are allowed. That will increase hitters need to swing and get the game over quicker. Or just have Selig send out an edict to the umpires to call more strikes. That way you don’t have to alter something like the count so drastically.

I would dislike such a drastic change, because they would basically have to start over with records (for the same reason that baseball doesn’t count records from before 1900 - the game is too different).

Currently, you can compare statistics from today’s game with those of the 1920s. There may be a lot of differences, but it is recognizably the same game.

There other less drastic ways that the game could be sped up. Quit automatically calling time every time the batter steps out of the box. Put a limit on time between pitches and enforce it. Limit visits to the mound by managers, coaches, and other players.

No rules change necessary for this thought, just a tactics (and corresponding roster) change.

Is there a reason three outfielders are needed?  The existence of the shift and the emphasis on defense for middle infielders makes me wonder if going to 5 infielders and 2 outfielders might be an improvement.

For discussion, let’s move the 2nd baseman behind the bag and create a “longstop” between 1st and 2nd base, adjusting spacing as needed.  (You could have an overshift with 3IF on the right side plus 3B and SS in the normal places if that’s the right way to play that batter.)

The longstop would be kind of like the current 2B, but with less range needed.  SS and 3B become easier to play, and the new 2B position becomes more like a corner position.

Upsides:
* This solves the shortstop shortage, as the position becomes much easier to play.  This is even more of a factor for backup SS than starters.  (You would see more Miguel Cairos and fewer Wilson Valdezes on rosters.)
* This combines very well with heavy groundball pitching.
* You can stack lots of offense in your infield without much of a penalty.
* Defending against smallball tactics (bunt, steal) becomes a little easier.

Downsides:
* This formation is less valuable in combination with flyball pitchers or high-value shortstops.
* Defense is now at more of a premium in the outfield.

Hmmm. . . this might be worth more thought.

"One of the worst counts is 3-0, because over 90% of the time, it’s a take.  Over half of 3-0 pitches are called strikes, and 35% are called balls. The 3-0 count is a terrible count to enjoy baseball.  The 0-0 count is a bad count.”

This seems backwards to me, almost akin to arguing that, since the most exciting times in baseball are in the bottom of the ninth with runners on, we should just start there. Or let each team bat its most exciting hitters again and again.

The excitement of the 9th inning is earned over the course of nine innings. The excitement of the 3-2 count is earned during the previous few pitches. The 0-0 count is a good one *because* it’s exploratory; it whets the appetite. Go to two strikes, and the first pitch will mean *too* much; batter swings and misses and the at-bat is largely over. The game becomes restless and impatient.

As for counting 2-strike fouls as strikes…as you might be aware if you’ve attended a baseball game, fans *love* long foul-ball battles. Not just hardcore, old-timey fans, but fans, the folks sitting in the stands. Do the cheers start to lose force after pitch 7 or 10? No; the crowd gets more and more into it, and when the batter actually gets a hit or even a walk the reaction is tremendous. Scott Spiezio’s home run, one of the great contemporary World Series moments (well, not great for me, I’m a Giants fan), was so exciting partly because it followed a long battle.

I think 3 hours is long for a ballgame too, but do you think that at an hour shorter baseball would be more popular than football? Football games are three hours and feature a staggering amount of dead time.

If you want to cut down on dead time, prevent batters from leaving the box and shorten the time between innings. A conversation like this one, though, is just a fun way to kill time during an All-Star Break.

Re #27:

I think the slash lines would decrease even more since the percentage of innings thrown by good pitchers would increase significantly.

Tango, does you 10% run scoring decrease estimation in your opening statement take this into account?

The 1st pitch for the batter is all about “timing”. To remove the “0-0 take” basically gives all power to the pitcher. I love it (as a former pitcher), but as a fan, it stinks.

Here’s how the 1-1 count change would play out ....

[1] Pitchers will move away from being fastball dominant and start being more “out pitch” dominant.

[2] Hitters will have to switch to the approach “you better hit the first strike you get”.

[3] Pitchers will exploit #2.

Seriously, batters are going to get the pitcher’s best pitch right off the bat and if it’s a strike, the at bat is pretty much over. If it’s a foul ball, same result.

If the idea is to shorten the game, there are other ways ... we use them in amatuer baseball (courtesy runners for catchers, limited pitches between innings, limited time between pitches, etc). DH in both leagues ensures that pitchers are always ready to pitch, and remove the “warm up pitches” between innings. Of course, that means commercials are decreased, so that won’t happen.

There’s not too many pitches in a game, there’s too much time between pitches. There’s too much time between innings.

IMHO, this is something that baseball is going to have to examine and look at in a hard way. Society has changed. There’s no more leisure pace. There’s no more “passing the time”.

I think that MLB could do what the NFL and NBA do, a “pitch clock”. You have ~X seconds from the time of last pitch/play to deliver the next pitch. If not, it’s a ball (or a walk). Batters are not allowed out of the box. Believe it or not, batting gloves really don’t need unstrapped and restrapped after every pitch, nor do they need spat on, rubbed up, etc. Many of the time delays in baseball are just anxiety actions that have become habit.

To me, the time of games, and the lack of continual action, is baseball’s greatest threat. Society isn’t going to become slower paced nor are there going to be fewer options. This current generation texts and play apps while they are AT the movies, watching an action-packed thriller on the big screen. They likely are not going to sit through a baseball game when they have other choices ... let alone follow a team for 162 games. This is where the NFL has it made. You can follow every minute of your team’s action on 3 hours a week ... all on the one day where no one’s doing anything at noon anyway.