The Red Sox and the New Practice of Baseball

[Reverend]

There's been lots of back and forth, as we all know, about the meaning and significance of the new analytics movement in baseball. Also, there's been much made about whether or not it detracts from the purity and enjoyment of the game.
 
For some, the analytics alienate the fans from the game because we don't know on what basis decisions are made, such as why one player is preferred for the roster over another, or why one is played in one spot versus another. Also, some argue that it alters the play in ways that move us away from the real game where great players just see who can outplay whom and make it more about stat geeks finding tricks--so it's like the game is taken off of the field.
 
In my view, though, this understanding misses the newer developments in analytics that point us back towards actual play that requires good old fashioned scouting and coaching.
 
Maybe the first or second wave of analytics was about assessing players in ways that aggregated results and made baseball an abstraction. The new wave coming through, though, is taking high end information and giving it back to players and coaches to make the players better at baseball. In other words, many of the new analyticsare moving the focus back on a deeper thinking of the mechanics of physical play and how to play the game right. 
 
The Red Sox, of course, are among the league leaders in using advanced analytics. Therefore, I thought it would be fun and educational to start a thread where we could collect insights (references to which are showing up more and more in articles and columns) into how baseball teams and the Red Sox especially are using analytics to inform not only their personnel but their day to day play to help us watch what they're doing.

 

[Reverend]

While I'm at it, meet Brian Bannister, Sabermetrics, Scouting and Science of Baseball participant and the original poster boy for a guy who applied analytics to his own game so that he could find some success and prolong his career in major league baseball.
 
Here's a neat piece by Jeff Passan at Yahoo! Sports on Bannister's use of the new stuff:
 

This is rather fortuitous, because Bannister is the type of player the sabermetric community tends to frown upon, the one with seemingly middling stuff, a terrible strikeout rate and a flyball tendency. To them, Bannister is the pitcher you don't want on your staff.
 
And yet he's exactly the kind of person you do, the one who's self-aware enough to realize that he can't get by on just an 89-mph fastball and needs something supplementary to make him into the pitcher he wants to be. Bannister, a noted amateur photographer, sees pitching as an art, and for much of the 2007 season, he was Escher, the same incongruous scene playing again and again.
 
Bannister finished 12-9 with a 3.87 earned-run average, and that was after his two final starts kicked up the ERA nearly a half-point. Otherwise, in his rookie season, Bannister would have finished among the top five in the American League with a record well above .500 in spite of playing for the moribund Royals.
 
To explain Bannister's success in spite of his inability to overpower hitters is the crux of the scouts vs. stats debate that has raged for years but took root with the arrival of "Moneyball" five years ago. Scouts attribute Bannister's success to intangibles – wiliness, toughness and other -nesses – as well as the ability to keep hitters off-balance with his slow curveball.
 
Statistical analysts? Well, they just think Bannister was lucky.
 

 
There's more in there about his learning about analytics, and this HardballTimes article by Pat Andriola talks about how Bannister uses FIP which a lot of commentators couldn't wrap their heads around, but here's the kicker:
 

For Bannister, that just doesn't suffice. He wants to see which of his pitches get hit for home runs – the count, the situation, the speed, the break. He wants to know that in 0-2 counts, hitters are 3 for 53 against him in his major-league career, and, accordingly, he wants to figure out what pitches best get him to that count. He wants to increase his strikeout rate from 4.2 per nine innings – the sixth-lowest in the major leagues last year – to at least 5.5 and learn to do so with the same pitches he currently employs.
 
The first generation of sabermetrics – really, the first 25 years – served more as tools of retrograde analysis and, based on those numbers, future projection. Very little of it, however, translated to the field, because numbers aren't playable.
 
Today's data is. And so comes the revolution, one Bannister is happy to lead.
 
"I'm willing to be the guinea pig," he said. "A lot of people would look at me from a scouting standpoint and go, 'He doesn't belong in the major leagues.' I want people to never be able to question how much time and preparation I put into what I do.
 
"You can be the best pitcher in the world and break 10 bats in a row, and they can all go for hits. Or you can be the worst pitcher in the world, throw terrible pitches, and 10 guys in a row can pop out. It's an amazing game.
 
"One thing sabermetrics and statistics have allowed me to do is relax. I know the odds. I know percentages. I know that three out of every 10 batted balls should go for hits, and I deal with it. It's helped me be a better player."
 
He sees it, in his own unique way, every day.
 
 

 

[Reverend]

We have one too many guys right now.
 
The differences aren't significant enough right now to amount to real expected wins differences over a month or two, so we are talking about development and roster construction down the road, not our record on May 1st.
 
In terms of "earning it" and the negative impact of excluding from the opening day roster, JBJ is coming up short this year, just like he came up big last year.  He is being outhit by Carp, Nava, and Sizemore, and while Spring Training stats are not very indicative, I don't think he is really showing signs that he ready to outperform those guys at the plate to start the season.  Another struggle could hurt confidence, so I would like to see him earn his roster spot and be in a hot streak when he gets into the lineup.

 
 

For those who have watched the games, is JBJ hitting the ball well but just having bad luck?  Or are his stats indicative of how's he truly hitting?  With such a small sample size a couple of line drives right at someone or a couple of great plays on grounders by infielders could obviously skew the numbers.

 
As for watching the Red Sox, the impetus for this thread came to me because I've realized that I pay no attention to box score numbers this spring trainging. And I mean even beyond the concerns with small sample size.
 
The reason for this is because the box score only tells us outcomes, but not how well the balls were hit. John Farrell has indicated that, for defensive purposes, they only make decisions based on hard hit balls. The idea here is that hard hit balls are reliable indicators of how a batter hits a ball when he's really hitting in a replicable way, whereas weakly hit balls are less reliable and flukey--statistical noise is you will. 
 
This analysis also informed the decision to trade Iggy last year.
 
So GHoff's question here is right on point. Red Sox management knows spring training is a small sample size. So for both hitters and pitchers, what matters is how well the batter hit the ball and for pitchers, how the balls came off opponents' bats and not whatever ends up in the box score.
 
I expect this will inform how I look at "slumps" too, i.e. what kind of contact is he making or giving up. This is how we figure out how people are "actually" playing.

 

[McDrew]

Regardless of the field, step 1 in analytics is using it to understand the system that it is operating in.  Step 2 is then to leverage that understanding and data to improve the system. 
 
In the most simplest of matters, this is Madden starting to shift Ortiz.  He used analytics to realize that Ortiz hit balls to the right side of the infield a vast majority of the time, so he put more defenders on that side of the field so he could get more outs.  Other teams started following suit because it was successful, and there are now a number of players who are shifted because the metrics say it will improve the likelihood that the defenders are able to make an out. 
 
Now, a whole bunch of other metrics in other fields are being used in the same way.  I do somewhat the same thing professionally.  I write political campaign software, and one of the reasons we(and our clients) were so successful in 2012 was that we were able to help the campaign target people who would be the easiest to convert into actual votes.  For the most part, this was one of two groups of people:
1) People who would vote for the candidate, but might not actually go vote.  Getting them to vote is a vote for the candidate that wasn't there before.
2) People who would vote, but were still undecided about who.  Getting them to vote is a vote for the candidate, and also potentially NOT a vote for his opponent. 
 
Campaigns are like baseball in a way. Its a game and the winner is the one who scores the most votes.
 
Campaigns waste time talking to voters that will go vote, and know who they're voting for.  You don't want to waste any time for someone who already will vote for your candidate, and it is incredibly hard to convert someone who is decided on a competing candidate.  The time/effort is much better spent on the people who will be easier to convert.  Converting 2 easy votes is more valuable than converting 1 hard vote. 

 

[Jnai]

Regardless of the field, step 1 in analytics is using it to understand the system that it is operating in.  Step 2 is then to leverage that understanding and data to improve the system. 
 
In the most simplest of matters, this is Madden starting to shift Ortiz.  He used analytics to realize that Ortiz hit balls to the right side of the infield a vast majority of the time, so he put more defenders on that side of the field so he could get more outs.  Other teams started following suit because it was successful, and there are now a number of players who are shifted because the metrics say it will improve the likelihood that the defenders are able to make an out. 

 
I love the topic (and know Brian and Jeff quite well), but this is the most revisionist history of the Williams shift that I have ever seen written.

 

[McDrew]

 
I love the topic (and know Brian and Jeff quite well), but this is the most revisionist history of the Williams shift that I have ever seen written.

I remember Madden being the first to do it, but I'll absolutely admit that I follow hockey a lot more than I follow baseball.  My point was more someone(I thought it was madden) started shifting, it worked, the data supported it, people used data to figure out who to shift on, and many teams are using it now.  Less on who, and more of the process. 

 

[Red(s)HawksFan]

I remember Madden being the first to do it, but I'll absolutely admit that I follow hockey a lot more than I follow baseball.  My point was more someone(I thought it was madden) started shifting, it worked, the data supported it, people used data to figure out who to shift on, and many teams are using it now.  Less on who, and more of the process. 

 
The Williams shift is born

 

[absintheofmalaise]

I remember Madden being the first to do it, but I'll absolutely admit that I follow hockey a lot more than I follow baseball.  My point was more someone(I thought it was madden) started shifting, it worked, the data supported it, people used data to figure out who to shift on, and many teams are using it now.  Less on who, and more of the process. 

The Rays are one of the most advanced analytics teams in MLB. I recommend reading The Extra 2%: How Wall Street Strategies Took A major League Baseball Team from Worst To First by Jonah Keri. Maddon might have utilized the shift, but I'm pretty sure the initial information came from the guys in the FO.
 
Here is a thread I started in February on shifts.

 

[Jnai]

 
The Williams shift is born

 
In fact, this is also revisionist history, the Williams shift was not named for Ted, and had existed since at least the 1920s.
 
Anyway, this is insane. Carry on.

 

[koufax37]

I find this top very interesting, and hope this thread gets some life.
 
Stepping back to the philosophical, I think that analytics are awesomely useful and interesting and exciting.
 
But I think they alienate fans on two intertwined levels:
 
1) Pretending to be a complete picture, and that in our current state the idea that WAR is a complete picture of past value, and that projected/expected WAR is an accurate and complete picture of future value.
2) Removing the perception of skill, effort, and human element that creates moments and make the game so enjoyable and replacing them with probability and random distribution. 
 
I personally agree that both elements of this aren't really accurate.  We have a better understanding of value of players than every before, but it is still evolving and leaves a lot of room for uncertainty and new and refined measuring of past value and prediction of future value.

But more importantly, I think there are elements that are not fully understood, and feel with all of the analysis and understanding, we can get pretty good at predicting the huge sample size distribution, but there is still more going on than is realized and understood that affects smaller sample sizes than random distribution.
 
Defensive metrics and now pitch framing are showing some evolution from what people thought they knew as the whole picture a few years ago, but I still think there will be more areas found that guide why small samples are different than large sample expectations beyond random distribution.  I think that there will be some evolution in the understanding of "clutch" but it will be recast as "focus" or something else, and I think it will dig deeper into how batters perform against executed pitches vs mistakes, and more broadly against good pitchers vs average pitchers.
 
 
The other part that I think is confused some times and will continue to evolve is how much these analytics impact player evaluation and game strategy vs how much they impact player development.
 
I think there is still an understanding gap on how much or little a player has to be aware of some of these analytics to perform, and how much some of these things that we know which direction is more valuable (line drive rate, walk rate, pitcher per at bat, K rate, BB rate, HR rate) can be incorporated into organizational philosophy and player development and be improved vs merely selected.

 

[OttoC]

 
In fact, this is also revisionist history, the Williams shift was not named for Ted, and had existed since at least the 1920s.
 
Anyway, this is insane. Carry on.

 
I have the page from a Life magazine that reported on the shift employed as it was first employed against Ted Williams on July 14, 1946, game 2. It includes a photograph of the fielders and three of Williams at bat. In the write-up it remarks that "Boudreau in the second employed one of the most remarkable defensive overshifts in baseball history. Deploying all his infielders between first and second, and stationing two outfielders in right field, he confronted Williams with a ludicrous exaggeration of tactics used against several great sluggers two decades ago."
 
While Cy Williams was one of those sluggers who saw some form of shift, I think it really became known as the Williams' shift because of Ted.
 
Doing a Google search for "Cy Williams" shift brings up page 154 in a Google book, Ted Williams: A Baseball Life by Michael Seidel, that describes things. You might want to start with the last paragraph of page 153 and go through the first paragraph of page 156.
http://books.google.com/books?id=75ukTD7fhIUC&pg=PA154&lpg=PA154&dq=%22cy+williams%22+shift+baseball&source=bl&ots=ZYc0Ok4N5E&sig=_c_-C9-q9sZefWKnIgWeCN7YLVM&hl=en&sa=X&ei=KpQsU_iXG6HQ0gGZmICYBg&ved=0CIMBEOgBMA8#v=onepage&q=%22cy%20williams%22%20shift%20baseball&f=false

 

[lambolt]

I think an interesting aspect of this kind of application is when it becomes a question of whether the adjustments you make then force a change on the opponent, and how that then rechanges the odds. Shifting Ortiz is one thing, but is it possible he could learn to hit another way that takes advantage of that situationally. Does the defense then have to refactor that in based on their situation etc.
 
I've also seen many times extensive analysis that says you can throw any pitcher into a 3 run save and based on whats happened before it is as good as having a closer. but heres the thing, in the bazillion times a 3 run save has come up, most teams still have always sent their closer out, so how can we know that over a huge sample size that theres no effect of "pressure" in a lesser pitcher coughing up what are supposed to be "nigh on unbeatable" leads.
 
I think this new tracking system is going to have a monumental effect on how players are evaluated (I don't know to what extent the leading edge teams have had this info already in some ways), and (hopefully) how fans can get to see things they don't normally.
 
It's really a crying shame that your media market is totally catered towards morons who are only interested in their fantasy stats and how the game was played right in the old days. So much cool stuff in sabermetrics that you would think that at least ONE show out of the 43,000 you have over there might decide, hey, no one else is doing this, lets do a baseball program and really get into all this stuff. Unfortunately it only takes a single new acronym to come up or any kind of calculation including numbers with more than 0 decimal points and some people seem to just enjoy displaying their ignorance like some kind of badge of honour. Honestly, while there's all kinds of amazingly nuanced analysis going on, it's really NOT that complicated, at least most of the fundamental principles driving player analysis, analysis of how a team succeeds, how odds are applied in lots of different ways and how things are related to league averages and replacements and margins. 

 

[OCD SS]

Just to add a small and only somewhat related anecdote, I'm reminded of a reminded of a debate Richard Feynman had with his neighbor, an artist. The artist (who I imagine to be the worst sort of Sunday painter) argued that all of scientific minutiae that Feynman and his colleagues brought to bear on a rose, understanding its anatomy, internal structure and place in the environment, clouded and lessened their appreciation of its beauty. Feynman's response was that, on the contrary, he could appreciate the flower's beauty more for he and his scientific colleagues had an understanding that went far deeper than the mere surface beauty.
 
I am an artist, but I agree wholeheartedly with Feynman.

 

[OttoC]

The future of baseball...everything will be measured so precisely that there will be designated pitchers for designated hitters and designated fielders for designated hitter/pitcher combinations.
 
Do I want baseball where everyone is positioned perfectly? No.

 

[HriniakPosterChild]

Is there a proposal for a rules to allow free substitution?

 

[Savin Hillbilly]

Do I want baseball where everyone is positioned perfectly? No.

 
Why not? Why is it a negative for any aspect of the game to be executed as well as possible? Fielders are already trying to be positioned as perfectly as they can; if better information makes them better at it, how is that a bad thing? If the issue is that this will suppress offense, well, that will just bring different skills to the fore, as changes in the game have always done. In a game of (almost) perfectly positioned fielders, contact skills will be slightly less valuable, and power and plate discipline will be slightly more valuable. And if this has a negative impact on entertainment value, by making games longer or what not, then MLB can adjust the rules, as it has before.

 

[67WasBest]

Do I want baseball where everyone is positioned perfectly? No.

IMO, there are too many variables for this placement to ever be realized.  Players positioned optimally could be acheived; and if so, the batters will modify their approach.  The constant pitcher / batter adjustments that have always existed, are morphing into pitcher and fielder / batter adjustments.  And the beat goes on!

 

[Idabomb333]

I've also seen many times extensive analysis that says you can throw any pitcher into a 3 run save and based on whats happened before it is as good as having a closer. but heres the thing, in the bazillion times a 3 run save has come up, most teams still have always sent their closer out, so how can we know that over a huge sample size that theres no effect of "pressure" in a lesser pitcher coughing up what are supposed to be "nigh on unbeatable" leads.
 

I'm probably making too big a deal out of this, because I agree with the other stuff you're saying, but what are you talking about here?  Teams put pitchers other than their closer out in save situations of any lead size a lot.  Closers get injured, or need days off, and it's perfectly possible to observe and compare how pitchers do in high leverage situations vs. low leverage situations.  The idea of a closer is also relatively new.

 

[OttoC]

Is there a proposal for a rules to allow free substitution?

 
Rules can be changes, can they not?

 

[OttoC]

 
Why not?...

 
I want to watch my baseball played by (to err is) humans and officiated by (to err is) humans. Sorry, but I have been watching baseball a long time and I still have a fondness for the game I saw in the '50s.
 
And just to make some things clear about me, I learned how to score games on August 21, 1952, and after that I scored every game I listened to or watched. And not long after I began scoring games, I started compiling stats from my score sheets. It wasn't long after that that I embarked on trying to prove that Ted Williams was the best. Since I lived in an analytical wasteland, I had to develop my own metrics (which I later discovered others had already done). I've been an active SABR member for more than 30 years and have known and discussed baseball statistical analysis with many of the people on whose shoulders rest the rise of sabermetrics. I am not a curmudgeon when it comes to advances in statistical analysis but I do prefer the baseball of my childhood to today's game.

 

[HriniakPosterChild]

 
Rules can be changes, can they not?

 
In secret?!
 
Do you really think that there will be a serious proposal to change baseball into football? Really?

 

[Alcohol&Overcalls]

 
In secret?!
 
Do you really think that there will be a serious proposal to change baseball into football? Really?

 
Why not?
 
The players would love it (more members for MLBPA!), and the owners would love it if it created larger revenues.

 

[alwyn96]

Just so we're clear, I tend to think of the various stats waves this way:
 
First: batting average, ERA, RBI, etc
 
Second: OBP, OPS, three true outcomes, DIPS
 
Third: WAR, advanced fielding stats, UZR
 
Fourth: Pitch f/x, MLBAM, this crazy shit: 
 
I'm not sure if any of these measures are more abstract than another, aside from WAR, which is pretty tough to calculate. DIPS stuff isn't that intuitive sometimes. But I don't think use of WAR-like stats is particularly new. WAR itself has been around for years, and uberstats have been around for decades. Anyone remember total average? Win shares? WARP? Guys have been playing with sliderules in their mother's basements for a long time. The inputs for WAR (aside from UZR) aren't new. It could have been invented 50 years ago. 
 
Now, these fourth wave stats look crazy different to me. The measurement of batted ball speed and players positioning and running speed using imaging technology and computer algorithms is categorically different than merely aggregating traditional outcomes the way WAR does. These are process measures. We're just scratching the surface with that stuff, as evidenced by the MLBAM video. That is where the real change in statistics is going to come (and maybe it's come already to some teams' FOs). These are statistics that make scouts nearly obsolete, and ironically I bet these kinds of stats are probably of the most interest to players. How fast the ball comes off your bat is something you can easily conceptualize, how much WAR you just earned from that game is not. 

 

[Vermonter At Large]

First, lets get away from the talk of shifts.  You certainly don't need advanced measurements and algorithms to tell you when to reposition a fielder and that argument
really dilutes this discussion.
 
Second, I think what is happening is that baseball people have sort of separated stats into two different categories, with both strategic and tactical applications.  Things like
WAR might give you an idea of a player's relative contribution and help the front office make personnel decisions regarding roster management and contract negotiations, 
but do little to advance the playing of the game.  Tactical applications can add value to player development, training and actual in-game management.  I think we're seeing a lot more of the latter usage in the present tense and I think that is a very good thing.  The constant challenge is to somehow marry the abstraction with the reality of the situation and glean as much useful information (and discard the disinformation).  I think this enhances the game for both the teams and the fans.

 

[OCD SS]

Just so we're clear, I tend to think of the various stats waves this way:
 
...snip...

 
If you haven't read it already, I highly recommend Alan Schwartz's "The Numbers Game" which is basically a history of baseball statistics.
 
As for the human element, I think chaos (or more specifically a sensitivity to initial conditions that magnifies the outcomes) will be OttoC's friend. No matter how much more effeciently teams may play the probabilities, they are only still probabilities, and the exciting outcome that goes against them will be put in much starker relief.

 

[Frisbetarian]

 
I want to watch my baseball played by (to err is) humans and officiated by (to err is) humans. Sorry, but I have been watching baseball a long time and I still have a fondness for the game I saw in the '50s.
 
And just to make some things clear about me, I learned how to score games on August 21, 1952, and after that I scored every game I listened to or watched. And not long after I began scoring games, I started compiling stats from my score sheets. It wasn't long after that that I embarked on trying to prove that Ted Williams was the best. Since I lived in an analytical wasteland, I had to develop my own metrics (which I later discovered others had already done). I've been an active SABR member for more than 30 years and have known and discussed baseball statistical analysis with many of the people on whose shoulders rest the rise of sabermetrics. I am not a curmudgeon when it comes to advances in statistical analysis but I do prefer the baseball of my childhood to today's game.

 
Why do you prefer baseball from the 50's and 60's to today's game, and what do you see as the differences? I'm seriously interested in your take, and not trying to be critical. I started watching baseball in 1967, like so many others of a certain age in this area, and love the game today as much, if not more, than I did as a kid. 
 
 

Just so we're clear, I tend to think of the various stats waves this way:
 
First: batting average, ERA, RBI, etc
 
Second: OBP, OPS, three true outcomes, DIPS
 
Third: WAR, advanced fielding stats, UZR
 
Fourth: Pitch f/x, MLBAM, this crazy shit: 
 
I'm not sure if any of these measures are more abstract than another, aside from WAR, which is pretty tough to calculate. DIPS stuff isn't that intuitive sometimes. But I don't think use of WAR-like stats is particularly new. WAR itself has been around for years, and uberstats have been around for decades. Anyone remember total average? Win shares? WARP? Guys have been playing with sliderules in their mother's basements for a long time. The inputs for WAR (aside from UZR) aren't new. It could have been invented 50 years ago. 
 
Now, these fourth wave stats look crazy different to me. The measurement of batted ball speed and players positioning and running speed using imaging technology and computer algorithms is categorically different than merely aggregating traditional outcomes the way WAR does. These are process measures. We're just scratching the surface with that stuff, as evidenced by the MLBAM video. That is where the real change in statistics is going to come (and maybe it's come already to some teams' FOs). These are statistics that make scouts nearly obsolete, and ironically I bet these kinds of stats are probably of the most interest to players. How fast the ball comes off your bat is something you can easily conceptualize, how much WAR you just earned from that game is not. 

 
At an HSSP MIT seminar I put together a few summers back, someone asked Joe Bohringer, then a scout for the Diamondbacks and now the Director of Professional Scouting for the Cubs, about the "future of sabermetrics." Joe is a very smart guy, a graduate of MIT's Sloan School Management and very much a proponent of advanced statistics. He cited the "4th wave" info you mentioned, the use of laser cameras and physics to assess defense, but also said he thought the pure stats/numerical advancements had been taken close to as far as they could. He thought that old time scouting was going to become a more sought after skill, and it would be enhanced by the new measurement tools. I tend to agree; there will always be a need for scouts who can recognize a hitch in a batter's swing causing him to have trouble with a breaking pitch, poor footwork by an infielder on a double play, a young pitcher's potential to gain 3 MPH on his fastball, etc. 
 
Re: the use of WAR, see my signature.

 

[lambolt]

I'm probably making too big a deal out of this, because I agree with the other stuff you're saying, but what are you talking about here?  Teams put pitchers other than their closer out in save situations of any lead size a lot.  Closers get injured, or need days off, and it's perfectly possible to observe and compare how pitchers do in high leverage situations vs. low leverage situations.  The idea of a closer is also relatively new.

yeah I didnt word that as well as I could have. I just remember reading an article that was basically saying no one needs closers because even bad pitchers get people out most of the time. Factor in the number of save opportunities (I think they included 3 run saves but now that I think about it, maybe they just meant 1 and 2 run saves i.e. "close" saves), and the conclusion was that you could choose ANY of  your bullpen in those games and over a long sample size, it wouldn't make that much difference but you would save crap loads on paying a "closer". But it used a lot of numbers from actual "save stats" and save appearances, and it's clear that there is a very large bias (or maybe I'm wrong?) that in the closer era, the "closer" (or the next best guy) predominantly takes most, if not all of those 1 and 2 run "nailbiter" saves. So all the numbers they'd used to say "X% of the time a save goes without problems" failed to address the in built selection bias that that was with teams best pitchers. Theres nowhere near enough data to say that over a large sample size, running out your crappiest bullpen guy in all your 1 run saves isn't in fact going to bite you in the ass because there are associated effects, psychological maybe, of the team being twitchy compared to calm depending on if your closer or your "meatball guy" is on the mound. Also the effect of losing close games on a teams W-L record. 
 
I'm not really saying *anything* (haha!) just that I read a long piece of analysis but it was all based on closer era data. I don't think it was valid to say that because 90% of saves are converted, you may as well throw anyone out there, it won't make that much of a difference.
 
anyway, I made a mess of explaining and remembering this so feel free to disregard as unreliable evidence

 

[HriniakPosterChild]

 
Why not?
 
The players would love it (more members for MLBPA!), and the owners would love it if it created larger revenues.

 
C'mon.
 
We've had the DH rule (on an experimental basis) on the books for 40 years now, and we can't get half of MLB to agree that it's a good idea.
 
I think you can relax.

 

[Reverend]

The Red Sox just went to the over-shift on a batter once the count when to 2-2 even though they were in a more traditional alignment up to that point.

 

[absintheofmalaise]

This is very interesting to me. Many hitters do change their approach with 2 strikes. Protect the plate mode. Who was the hitter and the pitcher and the outcome?

 

[Reverend]

This is very interesting to me. Many hitters do change their approach with 2 strikes. Protect the plate mode. Who was the hitter and the pitcher and the outcome?

 
OK, I was inaccurate in what I said, but because 1) what they actually did was way weirder, and 2) Lowe and Don were talking about how the concussion resistant hats don't look sexy.
 
So, the batter was Joyce, a lefty. To start they had a kind of funky shift with WMD playing around the normal SS spot and XB to the left of normal 2B and so far back he was on the outfield grass and the 2B playing pretty deep, even for a shift.
 
The, at 2-2, they swapped WMB and XB, and XB played further to the right than did WMB at the rough SS spot and out on the outfield grass, and WMB went to the normal 2B spot, but further to the right of where XB had been while still on the OF grass. 2B still deep and playing really close to the line.
 
Ended in a strikeout after a couple of really lame defensive swings.
 
Last batter in the 4th inning if anyone wants to check it out on mlb.tv. Or post screen shots (*hint* *hint*).

 

[Reverend]

OK, here's the over-shift that they started with on Joyce:
 

 
 
Here you can see WMB and XB changing their positions once the count goes to 2-2:
 

 
 
And here's where they ended up, now with WMB playing to the 1B side of 2B with XB to the 3B side:

 
 
And here's Joyce's hit chart from last year--you can play around with v LHP, v RHP, counts etc. here at Texas Leaguers.
 

 

[SouthernBoSox]

I'm assuming they like WMB charging in on bunt attempt more so than Xander.

With 2 strikes the bunt attempt is less likely thus they move Xander back to the more traditional SS position.

 

[HriniakPosterChild]

Madden would love it.

 

[Plympton91]

 
I want to watch my baseball played by (to err is) humans and officiated by (to err is) humans. Sorry, but I have been watching baseball a long time and I still have a fondness for the game I saw in the '50s.
 
And just to make some things clear about me, I learned how to score games on August 21, 1952, and after that I scored every game I listened to or watched. And not long after I began scoring games, I started compiling stats from my score sheets. It wasn't long after that that I embarked on trying to prove that Ted Williams was the best. Since I lived in an analytical wasteland, I had to develop my own metrics (which I later discovered others had already done). I've been an active SABR member for more than 30 years and have known and discussed baseball statistical analysis with many of the people on whose shoulders rest the rise of sabermetrics. I am not a curmudgeon when it comes to advances in statistical analysis but I do prefer the baseball of my childhood to today's game.

 
This is exactly how I feel.  I make my living doing statistical analysis, so I'm not afraid of numbers.  But sports are not math, and shouldn't be.  That's why I posted in the MLB forum the idea of limiting the ability of teams to shift their defensive players around in the same way the NBA limits the ability of teams to deploy zone defenses.  To me, the shift in baseball has evolved to the same point as if the NBA disallowed goaltending and allowed zones such that every team had a Manute Bol type who just stood under the basket and swatted away any shot he could reach.  It changes the game, and makes it less enjoyable.

 

[Toe Nash]

 
This is exactly how I feel.  I make my living doing statistical analysis, so I'm not afraid of numbers.  But sports are not math, and shouldn't be.  That's why I posted in the MLB forum the idea of limiting the ability of teams to shift their defensive players around in the same way the NBA limits the ability of teams to deploy zone defenses.  To me, the shift in baseball has evolved to the same point as if the NBA disallowed goaltending and allowed zones such that every team had a Manute Bol type who just stood under the basket and swatted away any shot he could reach.  It changes the game, and makes it less enjoyable.

BABIP is actually higher than it was in the 60s and shows a steady increase over time:
 

 
While the shift has taken away some hits recently, balls are largely being hit harder than they were in the "old days," which means a higher percentage of contact turns into hits. The reduced offense in the past few years is mostly attributable to the historically high level of strikeouts we're seeing. I'm not sure why you'd pick the shift to regulate out when if anything it's swinging things back to the way they used to be (once you do something about all those strikeouts!). You might also regulate weight training or tell players they can't swing as hard as they can and should choke up to make more contact. 
 
In Little League when the one kid who's hit his growth spurt already comes up, everyone tells their teammates to step back. Using data to move fielders is just a slightly smarter way of something you'd do intuitively, and it's not changed the game more than on the margins. I really doubt we'll see a game where everyone is positioned perfectly and BABIPs are absurdly low...if shifting had that big of an effect it would have shown up already.

 

[Plympton91]

I don't see why you'd look at BABip without accounting for more strikeouts (and walks) though. They are flip sides of the same phenomenon , which is a recognition of how important it is to swing at strikes. I bet another factor is less foul territory league wide.

None of that changes the calculus of the fan experience at a game in the same way as a shift that causes the optimal decision for David Ortiz to be a bunt down the third base line instead of trying at ripping a double into right center. I don't pay to watch David Ortiz bunt. Just like I don't pay the same price to watch minor league baseball as major league. YMMV.

 

[alwyn96]

I don't see why you'd look at BABip without accounting for more strikeouts (and walks) though. They are flip sides of the same phenomenon , which is a recognition of how important it is to swing at strikes. I bet another factor is less foul territory league wide.

None of that changes the calculus of the fan experience at a game in the same way as a shift that causes the optimal decision for David Ortiz to be a bunt down the third base line instead of trying at ripping a double into right center. I don't pay to watch David Ortiz bunt. Just like I don't pay the same price to watch minor league baseball as major league. YMMV.

 
I think he was looking at BABIP with regards to fielder positioning. I would guess that fielder positioning has a minimal effect (if any) on walks and strikeouts. 
 
That said, I think there are fewer balls put into play than there used to be, for the reasons you cite, although I absolutely pay to watch Papi bunt. I don't think baseball will ever become a game where it makes sense for Papi to bunt all the time, but a few well-timed bunts for game-theory reasons (keeping the infield honest) are absolutely awesome. That's part of what I love about baseball - the ever-shifting tactics and countertactics. I don't understand why people would want the game to be frozen and static. It's like complaining about someone moving your cheese. I mean, we all love the style of baseball from our childhoods best of all, but I love seeing the crazy new stuff guys will do to win. Towering homeruns and overpowering fastballs are great, but I also love smart guys who know how to sneak one past the defense to help the team or junkballers who get by on deception or a trick pitch like the knuckleball. That's why baseball is awesome. 

 

[Reverend]

 
I think he was looking at BABIP with regards to fielder positioning. I would guess that fielder positioning has a minimal effect (if any) on walks and strikeouts. 
 
That said, I think there are fewer balls put into play than there used to be, for the reasons you cite, although I absolutely pay to watch Papi bunt. I don't think baseball will ever become a game where it makes sense for Papi to bunt all the time, but a few well-timed bunts for game-theory reasons (keeping the infield honest) are absolutely awesome. That's part of what I love about baseball - the ever-shifting tactics and countertactics. I don't understand why people would want the game to be frozen and static. It's like complaining about someone moving your cheese. I mean, we all love the style of baseball from our childhoods best of all, but I love seeing the crazy new stuff guys will do to win. Towering homeruns and overpowering fastballs are great, but I also love smart guys who know how to sneak one past the defense to help the team or junkballers who get by on deception or a trick pitch like the knuckleball. That's why baseball is awesome. 

 
Indeed. If the best part of sport was identifying who had the greatest ability in linear fashion, we'd all be watching track and field. And in that vein, if Papi was good enough to hit the ball to left, then other teams wouldn't be using the large shift on him, so the argument that the large shift dilutes the game by hurting him because he's so good has a sort of begging the question aspect to it.
 
As interesting as the large shift stuff is, though, the smaller shifts are where the real action is now, I think. The advanced teams are putting a ton into figuring out precisely where to position players within the traditional zones. My real interest in the pictures of the over shift I posted above isn't even as much in the switching of the players--though SouthernBoSox made an excellent point about trying to take advantage of different skill sets--but in the different positions they had for the fielders after the count went to 2-2, even within the context of the overshift.
 
What we can learn from this is that the team at least thinks there is enough advantage to field position tweaking to make it worth spending time and effort on, and that that's true even bracketing the advantages from the over-shift.

 

[koufax37]

The best thing about widespread shifts is for the first time it really is breaking free of "this is where my T-Ball coach told me to play and if a ball is hit here and I'm not here it is my fault" to really analytical balancing of weighted risks.
 
I am driven crazy when "no doubles" old school managers have someone two feet off the line, like the risk of a ball getting by their backhand if they were five feet off the line is somehow a greater negative than opening up the 5.5 hole.  I am similarly very pleased when a Pujols/Goldschmidt situationally plays insanely far from the line instead of being a DH with a mitt.
 
I also think that there are so many variables and counter actions going on, that we don't have to worry about positioning data making anything cold or robotic or precise, and instead add a thoughtful richness to the game that has sometimes been missing when players position themselves on engrained autopilot and not based on what is really likely to occur.

 

[OttoC]

 
Why do you prefer baseball from the 50's and 60's to today's game, and what do you see as the differences? I'm seriously interested in your take, and not trying to be critical. I started watching baseball in 1967, like so many others of a certain age in this area, and love the game today as much, if not more, than I did as a kid....

 
It's a gestalt.
 

I dislike inter-league play…In my opinion it detracts from the All-Star Game, detracts from the World Series. I would prefer to see the talent spread over fewer teams. Imagine if they went back to two 8-team leagues. Now, there would be 30 starting shortsops looking to fill 16 starting spots, on down the line. This could strengthen the minor league clubs and it might loosen the grip MLB teams have on the minor leagues. Once upon a time, minor league owners could sign their own players, and would sign major leaguers who were on their way out to help boost attendence. It also would shorten the season…beginning in mid-April and finishing up the World Series around mid-October would suit me.

Of course, we are not going to see MLB going back to two 8-team leagues, so perhaps they could add a couple of franchises and split into four independent 8-team leagues.

As an aside to all this, I think it is ridiculous that they have combined the AL and NL but do not follow the same rules.

As for a game by itself, I deplore the direction that announcing has taken. Once upon a time, announcers described the game but now announcers think they are the stars. I'm sick of local celebrities (of whom I've never heard) monopolizing half-innings, innings where there is an occasional. "Whoops, we forgot to show the game," Flash down, then back to the booth. Oh, and I definitely want to see every sign fans bring in and definitely I need to hear and watch interviews with them. Instant replays...and instant replays of instant replays, ad (almost) infinitum.

Back then, I could read a book while watching the game. Crack of the bat, look up, see and hear what happened. Nowadays when I look up, I'll probably see the announcers discussing what color shirt they will wear tomorrow.

I can't say tht I care for the switches to action from other broadcasts; that is why they have post-game and sports shows.

I don't think players should be allowed to wear body armor just so they can hang over the plate. I also think pitchers should be allowed to challenge the hitters: brush 'em back, if need be.

I think the umpiring ridiculous is some regards: the de facto strike strike doesn't seem to match the de jure one; batters no longer need to attempt to get out of the way of pitches; too many phantom tags of scond base allowed. The new replay rule is apt to make umpiring worse because they'll just feel that any mistake can be overturned.

As I said before, I like my baseball to be played by (to err is) humans and to be judged by (to err is) humans. There once was a nice, leisurely pace to the game and now it seems so frantic. It's a hard question to answer but those are some of my thoughts.
 
(Besides, I remember when the announcer would do the beer commercial, which included puring a glass and having a swallow; nowadays I'll probably get a highly-intelligent Coors ad.)

 

[Papi's fan]

Watching a game earlier in the week I thought the umpires appeared very focused perhaps due to replay. We'll certainly find out if some umpires become less attentive. I don't think it would be a wise move. My biggest beef with umpires is their positioning to make the proper call.

 

[Reverend]

It's a gestalt.
 
I dislike inter-league playIn my opinion it detracts from the All-Star Game, detracts from the World Series. I would prefer to see the talent spread over fewer teams. Imagine if they went back to two 8-team leagues. Now, there would be 30 starting shortsops looking to fill 16 starting spots, on down the line. This could strengthen the minor league clubs and it might loosen the grip MLB teams have on the minor leagues. Once upon a time, minor league owners could sign their own players, and would sign major leaguers who were on their way out to help boost attendence. It also would shorten the seasonbeginning in mid-April and finishing up the World Series around mid-October would suit me.

Of course, we are not going to see MLB going back to two 8-team leagues, so perhaps they could add a couple of franchises and split into four independent 8-team leagues.

As an aside to all this, I think it is ridiculous that they have combined the AL and NL but do not follow the same rules.

As for a game by itself, I deplore the direction that announcing has taken. Once upon a time, announcers described the game but now announcers think they are the stars. I'm sick of local celebrities (of whom I've never heard) monopolizing half-innings, innings where there is an occasional. "Whoops, we forgot to show the game," Flash down, then back to the booth. Oh, and I definitely want to see every sign fans bring in and definitely I need to hear and watch interviews with them. Instant replays...and instant replays of instant replays, ad (almost) infinitum.

Back then, I could read a book while watching the game. Crack of the bat, look up, see and hear what happened. Nowadays when I look up, I'll probably see the announcers discussing what color shirt they will wear tomorrow.

I can't say tht I care for the switches to action from other broadcasts; that is why they have post-game and sports shows.

I don't think players should be allowed to wear body armor just so they can hang over the plate. I also think pitchers should be allowed to challenge the hitters: brush 'em back, if need be.

I think the umpiring ridiculous is some regards: the de facto strike strike doesn't seem to match the de jure one; batters no longer need to attempt to get out of the way of pitches; too many phantom tags of scond base allowed. The new replay rule is apt to make umpiring worse because they'll just feel that any mistake can be overturned.

As I said before, I like my baseball to be played by (to err is) humans and to be judged by (to err is) humans. There once was a nice, leisurely pace to the game and now it seems so frantic. It's a hard question to answer but those are some of my thoughts.
 
(Besides, I remember when the announcer would do the beer commercial, which included puring a glass and having a swallow; nowadays I'll probably get a highly-intelligent Coors ad.)

This is the most awesomely crotchety post I think I have ever read.

I have no idea what using data to win games would mean for any of it though.

 

[richgedman'sghost]

What do any of your complaints have to do with shifts or using data and statistics to better place fielders?

 

[AbbyNoho]

This is the most awesomely crotchety post I think I have ever read.

I have no idea what using data to win games would mean for any of it though.

 
 
Somehow the post managed to include complaining about people not paying attention to the game and not being able to read a book during a game. Also complaints that the games aren't officiated accurately enough and complaints that they use instant replay to get things right. I feel like I should get off of someone's lawn pronto. 

 

[OttoC]

I was asked, "Why do you prefer baseball from the 50's and 60's to today's game, and what do you see as the differences?" And I answered.