Synergy
Tapping the chin over Michael Bourn

.

Again splicing general baseball strategic insight ---> against --- > current tactical situations:

.

 

I was wondering if players' offensive or defensive values are mutually reinforcing enough with their teammates to change their relative values for different teams, or is value value? Ok, that doesn't make a lot of sense, so let me try an example...if a good fielding / mediocre hitting team and an all-bat, no-glove team are both looking for a shortstop, should they value the available players differently? Should the first team look to improve their strength or mitigate their weakness? Should the second team just give up on defense and get the best hitter available? Or is the best available shortstop just the best available shortstop?
Asked by: Zeke**
Answered: 12/15/2012
Well. . .I don't know if this is the RIGHT answer, but then, neither does anybody else.    I would consider the other things at the margin.   If you have a slow left fielder and a slow right fielder, you probably need a fast center fielder.   If you have a bad defensive third baseman, you probably need a good shortstop.  
 
I think there's a rational basis for that, which is this.   While we tend to think of plays as "belonging" to one fielder or another (and we tend to measure defense in that way, and we tend to model the game that way in simulations and table games). ..while we think of the issue in those terms, it is easily observable that there are many plays in the field which can be made by either of two fielders (and sometimes more than two.)   It stands to reason, then, that when one player's range contracts, his neighbor can cover that to some extent.. .whereas if two neighboring fielders both have poor range, there is probably an interactive effect.   
 
There is a second reason to avoid stacking up liabilities in the field, which is the curvature of the lines.    If you increase hits by 10%, you increase runs by 20%.   If you increase runs by 20%, you increase losses by 44%.   When you stack up parallel liabilities in the field, there may be a more-than-proportional cost because of the curvature of the lines.   

.

I do believe this to be true, though it hasn't been published yet as a study on Hardball Times, at least as far as I know.  That's okay; we believe millions of things we don't have conclusive proof for.

Dr. D has talked several times about redundancy in chess and in game theory.  You don't want four units of energy doing the same thing; you want them diversifying their possibilities.  In chess the concept is "mobility," and it increases one's options.  In basketball, you wouldn't want three players standing next to each other on the low post.

In baseball, it's easy to imagine that if you put Adrian Beltre next to Brendan Ryan, that Beltre would actually take away from Ryan's ability to affect a game.  ASSUMING THAT YOU ALREADY HAD ADRIAN BELTRE, you would be LESS motivated to ADD Brendan Ryan to your team.  Ryan wouldn't get you as many defensive runs as he'd get somebody else, right?

................

It's an old-timey baseball truism that you need a couple of great defenders out there somewhere.  Every, and we mean every, manager wants a great center fielder.  But, oddly, there isn't the same universal clamor for three great outfielders.  Rather, they think in terms of a Norm Cash "covering for" a Ted Williams in left.

Supposing you were to add Nick Swisher in right field, ALONGSIDE Michael Bourn in center field, you would exploit this principle.  The fact that defense is not a part of Swisher's 4.0 WAR would actually enhance Michael Bourn's defensive 2.0 WAR, perhaps raising it to 2.2 or 2.3 WAR.

Bourn isn't my idea of a 5.0 WAR player, exactly, but it's interesting to reflect that if some team were to purchase them as a pair, they might wind up with an impact outfield.

From an offensive standpoint, it might work the other way:  Bourn gets his stolen base for that series, and then Swisher coming up behind him draws a walk.  The SB is wasted in that event, and you still risked the CS.  Better to hit Swisher #5, perhaps, or Bourn #9.

....................

There's a beauty and elegance to this kind of ballclub construction, having a great defensive player at second base standing out there next to a thumper playing shortstop.  And a true greyhound in center field does allow you the luxury of putting WWF wrestlers in the corners.

.

Comments

1
Lonnie of MC's picture

The mixing and combining players based upon their skill-sets if very interesting. A lot of people do just that when they make their suggestion as to who to add (big blog). The problem with that approach is that if chemistry is completely thrown out the window, what you end up with is a quantifiably perfect, but dysfunctional product.

2
ghost's picture

And one with which I wholeheartedly agree. And I think that this is the fundamental problem with my previous evaluations of the potential for great defense to win many ballgames. I still don't think pitching is 80% of the defensive puzzle the way that James assumes, nor that defense is less important in net than offense. BUT...people have looked at the defensive wins my evaluations have claimed (4.7 for Cameron in 2003, for example) and said - "wow...that can't be right - you're saying that if you had 8 fielders as talented as Cameron at each position, your team would win 37 games from defense ALONE??" I've always blinked and said, "I guess that's what I AM saying..."
The answer is...no...because there are a limited number of balls in play and 8 great fielders will be robbing plays from each other like crazy and all of their values will go down.
And that calls into question the entire notion of rating defense skill in a team-based construct and treating dWAR as bankable commodities. EVEN THOUGH defense is a team sport, I think we need to figure out (somehow) how many plays each fielder would have made with an all-average team defense and pitching staff if we truly wish to compare players historically. But then we need to make a change to a team-specific context for the purpose of projecting. GIven that Brendan Ryan covers X plays to his right more than normal per chance, how many runs is that if the third baseman steals Y chances in 2013?
This may actually be a big part of why defensive performance appears to be so chaotic.

4

But I don't know that I buy it.
There could be an entire McCracken-level system update needed here Matty - we wouldn't think of looking at pitching or hitting stats without normalizing them for park and conditions.  But sabermigos are not used to thinking in terms of normalizing for TEAMMATES.

5
ghost's picture

UZR can't normalize for teammates. By definition, if a good fielder takes a ball in a bad fielder's zone...it's a miss for the bad fielder an OOZ play for the good fielder. The good fielder steals credit from the bad, resulting in an amplifying impact on the results. Two good fielders side by side will be stealing lots of plays from each other...both of their lines will go DOWN relative to what they would be with an average fielder for a partner.
I think I will make this my top priority if ever I get any time to do more stat analysis

Add comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd><p><br>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

shout_filter

  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.