View From The Bleachers

Talking Cubs Baseball Since 2003

Sunday

18

March 2012

16

COMMENTS

The Impact of Good Plays and Misplays

Written by , Posted in General

by John Dewan

There are 54 separate Defensive Misplays and 28 separate Good Fielding Plays that Baseball Info Solutions has “scouted” going back to 2004.  One of our biggest undertakings in the last three years has been to convert Good Fielding Plays and Defensive Misplays into Runs Saved.  Today we’re going to walk through an example that shows the magnitude of what we now refer to as Good Play/Misplay Runs Saved.

Take, for example, Alfonso Soriano of the Chicago Cubs.  He’s now 36 years old and a below-average outfielder, according to Good Plays and Misplays.  In fact, his 13 runs lost on Good Plays and Misplays in the last three years is the worst among all outfielders. He had 22 Good Plays, 73 Misplays and 25 Errors. That’s 76 more misplays and errors than good plays. The next worst left fielder is exactly half as bad. Logan Morrison had 38 more misplays and errors than good plays. The best left fielder in GPF/DME runs saved is Jason Bay. He had 73 good plays and only 47 misplays and errors in the three years, a net of 26. Compared to Soriano’s net of -76. Here is how they compare in runs saved:

Good Play/Misplay Run Impact Chart 2009-11

Good Play/Misplay Type

Bay
Runs Saved

Soriano
Runs Saved

Mishandling ball after safe hit

3

-4

Outfield assist after hit or error

2

-1

Holds to single

1

-2

Wasted throw after hit/error

1

0

Cutting off runner at home

0

1

Giving up on a play

0

-1

Hesitating before throwing

0

-1

Slow to recover

0

-1

Robs home run

0

-1

Missing the cutoff man

0

-1

Overrunning the play

0

-1

Slipping

0

-1

Total

7

-13

Soriano has cost his team 13 runs with his poor play in the field since 2009 on Good Plays and Misplays alone, while Bay has saved his team 7 runs in that time.  That’s a difference of 20 runs, or roughly two wins.  That’s huge.

What we can see from the chart is that Soriano struggles in a number of areas.  In fact, the only areas where he rates as average or better are “Wasted throw after hit/error” and “Cutting off runner at home.”   In these types of plays, Soriano performs at least how we’d expect an average fielder to perform, in the same opportunities as Soriano.  In every other way, Soriano rates below-average.  The biggest problem Soriano has, according to Good Plays and Misplays, is “Mishandling the ball after a safe hit”, where he cost his team four runs since 2009.  Jason Bay, on the other hand, excelled in that department, saving his ream three runs by having far fewer Misplays for “Mishandling the ball after a safe hit” than an average fielder would have in the same number of opportunities as Bay.  Bay is also slightly above-average in three other categories: “Outfield assist after hit or error”, “Holds to single” and “Wasted throw after hit/error.”

For more on Good Play/Misplay Runs Saved, check out The Fielding Bible – Volume III, available now.

 

Used with permission from John Dewan’s Stat of the Week®, www.statoftheweek.com

  • Buddy

    Soriano is a bad OF? Really? I’m shocked! In other breaking news, the sun came up today.

  • Buddy

    Soriano is a bad OF? Really? I’m shocked! In other breaking news, the sun came up today.

  • Doc Raker

    It is good to see that someone is quantifying in the saber world what MOTE already know. What is surprising is Soriano only has cost the Cubs 13 runs in 3 seasons with his poor outfield play. I would have thought that Soriano has cost the Cubs much more with his poor play. This stat has many subjective variables to it so it certainly isn’t something I would believe over MOTE.

  • Doc Raker

    It is good to see that someone is quantifying in the saber world what MOTE already know. What is surprising is Soriano only has cost the Cubs 13 runs in 3 seasons with his poor outfield play. I would have thought that Soriano has cost the Cubs much more with his poor play. This stat has many subjective variables to it so it certainly isn’t something I would believe over MOTE.

  • Buddy

    Four+ runs per season is a lot to me considering how few chances a LF really gets over the course of a typical year.

    • Doc Raker

      I think it is to difficult to really quantify accurately. Example, a misplay forces your pitcher to throw 15 extra pitches. You give up 1 extra run immediately due to the misplay. Later in the game that pitcher comes out of the game 2 or 3 outs before he normally would of come out and the result of the bull pen getting those extra outs costs your team another 3 runs. Now the true cost of the misplay is 4 runs but this formula only counts the original 1. Also, most misplays don’t always cost an immediate run but will cost an extra base for the baserunner. How does this formula calculate an extra base into runs?

  • Buddy

    Agreed. It is hard to pinpoint. 

  • RichBeckman

    “That’s a difference of 20 runs, or roughly two wins.”

    Really!?  20 runs only comes to two wins?  This is the kind of stat that makes some a bit leery of sabermetrics. It seems absurd on its face, and so it makes one wonder what other absurdities are laying about the discipline.

    • Jedi

      It really is that kind of garbage that gives sabermetrics a bad name – if we’re down 10-0 I couldn’t care less how Soriano misplays a double into a triple.  Same thing if we’re up 10-0.  You can’t quantify how many games he’s cost us unless you actually look at the mistakes he’s made in their context.  I think you’d find that some guys have a tendency to lose focus defensively in lopsided games, or that other guys tend to boot the ball more in pressure spots.  It could be that a difference of 20 runs amounts to 8 or 9 wins, or perhaps that if he’d been perfect in the field not a single game would’ve ended differently.  But by assigning a formula for a number of runs that equal wins you’re really assuming a lot.  I could go even farther, does 10 runs difference equal 1 win for every team?  Because I would be that the teams with a crappy offense a 10-run differential in the field is much more than a single win…and for teams in the AL East 10 runs might be more like half a win.

    • Jedi

      It really is that kind of garbage that gives sabermetrics a bad name – if we’re down 10-0 I couldn’t care less how Soriano misplays a double into a triple.  Same thing if we’re up 10-0.  You can’t quantify how many games he’s cost us unless you actually look at the mistakes he’s made in their context.  I think you’d find that some guys have a tendency to lose focus defensively in lopsided games, or that other guys tend to boot the ball more in pressure spots.  It could be that a difference of 20 runs amounts to 8 or 9 wins, or perhaps that if he’d been perfect in the field not a single game would’ve ended differently.  But by assigning a formula for a number of runs that equal wins you’re really assuming a lot.  I could go even farther, does 10 runs difference equal 1 win for every team?  Because I would be that the teams with a crappy offense a 10-run differential in the field is much more than a single win…and for teams in the AL East 10 runs might be more like half a win.

  • Doc Raker

    I think when you get into abstract stats that have subjective input in the formula you can get a lot of garbage in garbage out. I take this stat as a generalization which my own two eyes can already give me but if I am looking at players who I never saw before it gives me a general guideline to their defensive play. I would not take this stat as an absolute.

  • Doc Raker

    I think when you get into abstract stats that have subjective input in the formula you can get a lot of garbage in garbage out. I take this stat as a generalization which my own two eyes can already give me but if I am looking at players who I never saw before it gives me a general guideline to their defensive play. I would not take this stat as an absolute.

  • RichBeckman

    Upon further reflection, I suppose if you take 20 runs and randomly add them one at a time to a team’s final score, you would end up with a few ties presumably half of which would result in a victory.  So looking at it that way I guess it makes some sense.

    I have the scores from some Cubs season a few years ago (I don’t know what year!?) in a spreadsheet.  Looking at the first twenty games, adding one run to the Cubs final tally results in five ties. So two and a half wins (based on an absurdly small sample size).

    • Jedi

      The key term there is randomly.

  • RichBeckman

    Upon further reflection, I suppose if you take 20 runs and randomly add them one at a time to a team’s final score, you would end up with a few ties presumably half of which would result in a victory.  So looking at it that way I guess it makes some sense.

    I have the scores from some Cubs season a few years ago (I don’t know what year!?) in a spreadsheet.  Looking at the first twenty games, adding one run to the Cubs final tally results in five ties. So two and a half wins (based on an absurdly small sample size).

    • Jedi

      The key term there is randomly.