THE BOOK--Playing The Percentages In Baseball

Answer: During the course of playing each other an infinite number of times, Ichiro gets an infinity game hitting streak, shattering Joe Dimaggio’s record, and its so awesome that no one cares how many games each team won.

One thing is for certain.  Over the course of an infinite number of games, everyone will beat Joe Dimaggio’s streak!

On a related note, if two teams played an infinite number of games, what is the longest hitting streak you are guaranteed to see?

Just a thought for a potential future question of the day:

In the BAL-SEA game today, bottom of the 9th, tie game, runnner on 3rd, 1 out, Baltimore decided to intentionally walk Ichiro and then Branyan to load the bases, with Beltre coming up.  A very interesting decision!  I was wondering if it was reasonable, or a mistake.

To be honest, I don’t really see how we’re supposed to do this, unless it’s just a guessing game.  In that case, I’ll take the Warlocks and the Elderberries (o_O) to win 52% of games.

MGL, wouldn’t you want them to play an insanely large amount of times as opposed to infinite? Because even if you could say that the talent level for one of the teams made it that they were likely to win 60% of the time, 60% of infinity is infinity. As is 40% of infinity. So you could never really tell how many times each team won.

Can we use the age of the players?

For the AL Teams I am taking Elderberries with a winning % of .570.
My stab

Took RAR to date of each team from Fangraphs.
Elder = 124.4
Avs = 90.9

~ 52.5 games played per team

Elderberries were 0.64 runs better per game.

So are this year 4.72 scored per game, so it put Elderberries at 5.04 r/g and Avs at 4.40 and using Pythag I get .570 for the Dingleberries

The should be regression some where there, but too many categories to regress.

Age? For what?  Too see how old they would get after a couple of hundred thousand games?

Yes, I guess just a grip of games, rather than an infinite number.

Yup, just a guess.  You can look at their current stats if you think that will help.

I think walking the bases loaded is really bad there.  A player’s OBP is just too high with the bases loaded.

And then when the bases are loaded, there is no real reason to play the IF in (which I think they were) with a fairly slow RHB who hits the ball hard (Beltre).  I think that was a very bad play.  You can walk the first batter and then bring in a lefty to pitch to Branyan and try and strike him out.  He is also slow enough for the GDP.

As was Trembley’s hit and run (bad) in the top of the 9th with Huff on first and the 3-1 count.  Terrible play.  You love the walk there and you force Mora to swing at a marginal or even a bad (if he does not know it is bad) pitch with a hit and run.  As it was, he swing at ball 4 high and fouled it off.  On top of that, one of the reasons for the hit and run, even though it is almost always a bad play, is to get some extra hits through the hole vacated by the SS or 2B. 

The announcers had just remarked that there was already a gigantic hole on the second base side because the 2B was playing rally close to the 2B bag.  Is Trembley blind?  Could he not see that?  The 2B was going to cover the bag on the hit and run and he was already only 5 or 10 feet away from the bag in the first place, so that advantage practically was non-existent.  And, another reason for the hit and run is to advance the runner an extra base.  With no outs and a tied game in the 9th advancing the extra base is not all that valuable (around 10% more I think).

Some managers are just obsessed with the hit and run or feel a need to “do something” in a tight game to show how “smart” they are. They usually end up doing something dumb.  I don’t know much about Trembley, but that hit and run was a REALLY dumb play.  And then of course there is no “excuse” for Mora to swing through a fastball right down the middle, especially when he knows the runner is going (you have to take a little off your swing to make sure you make contact) but there is nothing you can do about that.  I don’t even like a slow runner going on a 3-2 count (less than 2 outs of course).  It forces the batters to swing at more marginal pitches.  Remember the question about the 3-2 count and 50-50 pitch?  What if you knew that 70% of the time there would be 2 outs (runner gets thrown out) on a K?

I probably should say, “No looking at the stats for this year...”

Off the top of my head, .550 for the Warlocks and the Elderberries.

On a related note, if two teams played an infinite number of games, what is the longest hitting streak you are guaranteed to see?

Well, it is possible that every single one of the infinite number of games is a no-hitter on both sides. (The runs would have to score in other ways.) Of course the probability of this is zero.

With probability 1 (which doesn’t mean every single time—see above), every player will have arbitrarily long (but not infinitely long) hitting streaks.

Oh, and Adam, if you just keep track of the winning percentage as they play more and more games, it will converge to the right number. There’s no need to divide infinity by infinity.

I am adding one more part to the question:

How many games would each team win versus an average team in the same league?

I thought the question was using just this year’s stats what the numbers would be and I figured you picked two even them and the only difference might be age allowing for better base running with hits.

Can we know the average split between right and left handed hitting vs right handed pitching?

You have to infer each player’s true talent as of right now.  IOW, a projection for each player, hitting, fielding, and base running.  Assume this is a real game, played right now.  You can look up if you don’t already know which of these players are RH, LH, or switchers.

Basically the question is which of the two lineups in each league would you rather have.  Imagine you are a sports writer or broadcaster and these teams are playing one another.  Which ones do you think are better and how do you think they stack up against an average team in their league.

IOW, would you say, wow, this a really good lineup, a pretty good lineup, an average lineup, a below average one, or a poor one.  Hitting, fielding, and base running combined.

Simple question.  I don’t really want any stats analysis.  I assume that almost everyone on this blog is pretty familiar with all of these players, more or less.

Hill 2B, Zobrist 3B, not the other way around, I presume?

I think the question is pretty tough just visually.  You’d have to put pen to paper, since I don’t see much bias.  You have Gutierrez, Hill, and Span on the same team (good D), but then you also have Jack Cust and Zobrist.

The only thing I can think that MGL did was take a real team, and found a player’s “twin”.

The Elderberries appear to be a much stronger team than the Avalanche, both offensively and defensively. Actually, you could make a case that the Avalanche is just a collection of the most overrated players in the AL, while the Elderberries are the most underrated.

The NL matchup is tougher, but might be a similar idea. I’d favor the Warlocks over the Fireplugs, for similar reasons (offensive reputations precede actual ability, especially with regards to OBP, plus defense).

I fear for the Elderberries playing Jack Cust at first since he hasn’t played there in more than a decade and seems to struggle in the outfield most on balls that are on the ground.

Actually, I think #16 gets the exact point of the exercise: in each league, MGL created two teams, one of which would be considered the better team by the typical fan/sportswriter while in fact being the inferior team.

I’ll say the Warlocks win 52.5% against the Fireplugs and 55.0% against average NL team. 

Elderberries win 51.0% v. Avalanche and 52.0% v. average AL team.

the Avalanche is just a collection of the most overrated players in the AL, while the Elderberries are the most underrated.

This is exactly what I thought when I saw the lists, which is why I made the choice that I did.  I think any “analyst” in the MSM would take the Avalanche to cream the American League.  Since this is an mgl question I’ll say they go .450.  Is that cheating? : -)

In the NL it’s not as clear to me.  Obviously Lee and probably Blake are overrated, but I don’t know much about Bonifacio or Loney.  I guess I don’t follow the NL as closely.  I think the Warlocks have a much better outfield but in the infield I’m not as sure.

Without looking at any stats or using any kind of tool there is no way of making an educated guess.
vr, Xei

For part B, the answer will be inflated because real teams don’t put their best lineup on the field every game.

That said, I take Fireplugs .490, Warlocks .540, Avalanche .520, Elderberries .570.

Hmm....
I’ll say the Warlocks are at .510 v. the Fireplugs, or .520 v. the ‘NL.’
The Elderberries may be at .510-.515 v. the Avalanche and .520 v. the ‘AL.’

As for motivation, the idea of putting together a ‘sportswriter’ team versus a ‘cognoscenti’ team may be part of it, but I’m thinking the teams are really, really close with the first two being guys that SABR tend to dislike for contract/age/’RBIs’ type reasons, with the second group tilted more towards potential/prospects, and even more tilted towards guys with great numbers in part-time play (Choo, Zobrist, ).  I’d imagine there are big platoon splits with the second group, and maybe not with the first?  Either that, or they’re guys that everyone *expected* to have big platoon splits.

You guys are too smart!  I didn’t pay any attention to platoon splits or much attention to how may LH and RB bats are in each lineup.

But yes, I did try and create a team that would likely be way overrated by the average sports fan or writer (or player/manager I would think) and one that would be underrated.

What prompted me to come up with this question was, I was watching a HOU/CIN game a few days or a week ago and I was (painfully) listening to Jeff Brantley (he and Harrelson have to be the absolute worst analysts).  He said something like, “Looking at this Houston lineup, it’s hard to believe they haven’t won more games.” Of course, he is looking at Lee, Berkman, Pudge, Tejada and maybe Bourne and Pence and thinking that this is a great lineup.  In fact, they have one of the worst overall lineups in baseball.  Had he even bothered to check out their team stats, he would have seen that they were around dead last in RS in both leagues.

Anyway, the Fireplugs and the Avalanche are terrible teams. Not just overrated (probably), but terrible.  The other two teams are very good. The Warlocks win 59% of the time when playing the Fireplugs and the Elderberries win 61% of the time when playing the Avalanche.  With equal pitching, that is a huge difference!

Every player on the Fireplugs is below average for their position other than Benji Molina, who is about an average hitter for a catcher, maybe a little better.  Their defense is awful and their offense is pretty bad - about 44 runs below average total, per 150 games.  Their lineup with average pitching would win only about 44% of their games against an average team in the NL.

On the Warlocks, every player but Adam Laroche is above average.  They would win around 56-57% of their games versus an average team.  Their defense is very good and their offense is excellent.

Basically, the Fireplugs are equivalent to around the worst lineup in baseball and the Warlocks the best, give or take a few percent of WE.  Although no one on the Warlocks is a superstar, Weeks, Hardy, Bradley, Hermida, and Ianetta are equivalent to major league stars, offense and defense combined.  I don’t think you can get a more underrated five-some than that.

In the AL, you basically have the same situation.  The Avalanche are a terrible team.  They will win 45% of their games versus an average AL team.  No one but Jeter is above-average at their position and even he is barely so at this point in his career (-10 on defense and +8 on offense).

The Elderberries are a great team.  They will win almost 58% of their games versus an average AL team! Their defense is fantastic and is almost half their total value.  Span, Zobrist and Cruz are stars that many people have never heard of, and everyone else is an above average player at their position (I even have Cust rated poorly at first base, around -5.5 runs per 150).  Ramon Santiago is the most underrated SS in baseball.  I’ll take him over Jeter.  How many mainstream people would laugh hysterically at that (remember when Tango used to say that Polanco and Jeter had the same value? - which they did).

How is Ichiro below average?  Are you giving him negative defense?  Just looking at Fangraphs WAR, he has been 4.2, 5.7, 4.9 the last three years, and is at 1.9 already this year.  Their rest of season ZIPS projections--which I take to be their estimate of true talent--anticipate that he’ll hit 314/360/403, which is obviously pretty mediocre for a corner outfielder, but with the baserunning and fielding numbers he has put up would still make him above average.

I am not including arm and base running, which probably add another 5 or more runs a year.  But, STATS UZR never liked Ichiro for some reason on defense.  My UZR projections are a combination of STATS and BIS UZR ratings.  I have him as below average in RF.  Right or wrong, that is what I have.  If I make him +3 runs rather than -1, it won’t change the assessment all the much.  If he is not below average for a corner, he is average.  Certainly not the star or superstar that EVERYONE thinks he is.