Baseball Savant - Infield Outs (Below) Average

[amjgt]

The metric is actually called Infield outs Above Average, but let's just say it doesn't paint a pretty picture of the Twins defense. 

 

https://baseballsavant.mlb.com/leaderboard/infield_outs_above_average

 

Polanco, 2nd to last among qualified players

 

[Brock Beauchamp]

Yeah, the middle infield is going to be ugly this year, even worse than last. Maybe Polanco can improve slightly but it's a significant drop from half a season of Schoop to a full season of Arraez.

 

Which is why it sure would be nice to get Donaldson, which removes a weak spot at third and potentially improves first base, too (though that's far from a given; I think Sano will be fine at first but we won't know that until he plays there).

[nicksaviking]

I'm not a big defensive bobo, but this is why I'm open to options moving BOTH Sano and Polanco to the right side of the infield, and that includes the unpopular opinion to dangle Arraez for pitching or 3B/SS options.

[howeda7]

Other metrics have Polanco only slightly below average. It does make me wonder if we shouldn't have pursued Iglesias though.

[USAFChief]

 

Other metrics have Polanco only slightly below average. It does make me wonder if we shouldn't have pursued Iglesias though.

I think what it this shows is either: 

 

1) How little teams (including metrics-friendly teams) value defense (Iglesias got $3M from bottom feeder Baltimore, imagine what Polanco would get on the FA market),

 

and/or

 

2) How little stock teams put into defensive metrics.

 

I think it goes back to the old saying...shake any tree, and a hundred gloves fall out...

 

There's a minimum level of defense required to play in the big leagues, beyond that, you need to hit. And hitting is a lot harder to find.

 

So...no, don't install an Iglesias into the lineup. 

 

[nicksaviking]

 

I think what it this shows is either: 

 

1) How little teams (including metrics-friendly teams) value defense (Iglesias got $3M from bottom feeder Baltimore, imagine what Polanco would get on the FA market),

 

and/or

 

2) How little stock teams put into defensive metrics.

 

I think it goes back to the old saying...shake any tree, and a hundred gloves fall out...

 

There's a minimum level of defense required to play in the big leagues, beyond that, you need to hit. And hitting is a lot harder to find.

 

So...no, don't install an Iglesias into the lineup. 

 

Yeah, I'm not up for sacrificing offense for defense, and I struggle to accept defensive numbers at face value.

 

But just watching the games last year, the Twins infield defense had more eye-roll moments than they usually do, so I'm more than open to fixing that. However, I'd rather just get additional offensive weapons who have a better glove to push the troublesome fielders to positions they'll be less troublesome.

[USAFChief]

Some questions for those smarter (or more industrious) than me:

 

is this a counting stat? Does someone who plays 1000 innings have a 2 to 1 advantage (or disadvantage) over someone who plays 500 innings? 

 

What do they measure against? 

 

 

[Mr. Brooks]

Some questions for those smarter (or more industrious) than me:

 

is this a counting stat? Does someone who plays 1000 innings have a 2 to 1 advantage (or disadvantage) over someone who plays 500 innings?

 

What do they measure against?

It's a counting stat.

 

It's basically a "difficulty bonus" for catches made combined with a penalty for those not made, if I'm understanding it correctly.

 

EDIT: For infielders it also includes force plays, with the baserunners speed being a contributing measurement.

 

If a play is determined to have a 75% out probability, you'd receive a +.25 for making the play, or a -.75 for failing to make the play.

[USAFChief]

 

It's a counting stat.

It's basically a "difficulty bonus" for catches made combined with a penalty for those not made, if I'm understanding it correctly.

Thanks.

 

Making it a "counting stat" seems to me a poor choice, tough to evaluate. But what do I know?

[SmellEm]

It's pretty difficult to tell exactly what we can learn from these sorts of stats without at least a few years of data to review.  Advanced defensive stats tend to be pretty noisy.  Their 2017 and 2018 numbers for Polanco are also poor, but nowhere near as bad as last year.  And in 2017 Dozier graded as well above average and then in 2018 he was well below average. Obviously that's possible with the injuries Dozier was dealing with his last year here but that's a pretty significant swing. 

[Thrylos]

Got to wonder the significance of this metric.

 

If for the duration of a season, Sano is allows 5 more PAs than the average third baseman (-5 OAA), which might translate to maybe 1 more run a season, or not.  It is not that much of a deal.

 

The importance of these outs are the situation in which they were given and their effect.

Based on this metric if Gonzalez had the innings that Donaldson had at third base, he would have been twice as good

[jorgenswest]

I don’t think we needed metrics to show us that Arraez, Polanco and Sano were well below average with the glove last year but here is another one with negatives around the infield.

 

As a team the Twins converted were well below average at converting balls in play into outs. That needs to improve. If not it will lead to longer innings, shorter starts and an overworked bullpen.

 

My hope is that improvement comes from within. Polanco is better with healthy legs. Arraez is better with consistent play at 2B. Sano doesn’t drop off. The staff does a better job of shifting their infielders.

 

If the Twins find themselves in the bottom quarter of turning balls in play into outs again they will need to take a serious look at the up the middle infield defense.

[cmoss84]

We also need to sustain our triple plays. Those were nice.

[USAFChief]

So I got motivated and put some research hours into this new stat last night.

 

More details to follow, but my overall thesis: the Twins defense isnt as bad as I feared. They averaged exactly 3 outs per inning, right at (American League) league average.

[Platoon]

 

So I got motivated and put some research hours into this new stat last night.

More details to follow, but my overall thesis: the Twins defense isnt as bad as I feared. They averaged exactly 3 outs per inning, right at (American League) league average.

True, but it often took them 4 chances per inning to do it.

[Doctor Gast]

Before Polanco came up it was always thought he`d take over 2nd in place of Dozier because of his defense at SS. At that time we had Escobar. Dozier was a SS converted to 2nd & he became a gold glove I believe Polanco could really turn it around like Dozier. Now we don`t have anyone that can step in & be that spectacular SS, so that`s why I`ve always suggest to trade for Rockies SS Story or someone like him.

[big dog]

I wonder how much shifting compensates for or overstates weak fielding.  Putting players in better position can make up for a lack of range, so if it is done well (and hitters don't adapt) then our stats ought to be higher.  If we shift and players adapt with 12-bouncers through the empty side, then fielders look worse than they are.

 

I don't know if that's true, but it seems like a potentially important effect, especially since some teams probably shift a lot more than others and some teams probably shift better than others.

[dbminn]

Defense is important. I looked at the new Statcast data for Andrelton Simmons. He had 300 balls hit to him in 2019. He made 16 OAA - that is, he converted 16 more outs than the average SS (not average defender). Doesn't sound like much? An average American league hitter hit .253 in 2019. That's equal to 76 hits per 300 AB. Add 16 hits to his season (92 for 300). That comes out to a batting average of .307. We'd all say that the .300 hitter had a far better season than the average player.

 

(Simmons played only 103 games in 2019. He had 398 plate appearances)

[Squirrel]

Before Polanco came up it was always thought he`d take over 2nd in place of Dozier because of his defense at SS. At that time we had Escobar. Dozier was a SS converted to 2nd & he became a gold glove I believe Polanco could really turn it around like Dozier. Now we don`t have anyone that can step in & be that spectacular SS, so that`s why I`ve always suggest to trade for Rockies SS Story or someone like him.

Adrianza is a far better SS defensively. Far better. He’s probably better defensively at all the infield positions than our starters. Guess that’s why he’s a utility guy. But, we do have someone better at SS ...

[Brock Beauchamp]

I wonder how much shifting compensates for or overstates weak fielding. Putting players in better position can make up for a lack of range, so if it is done well (and hitters don't adapt) then our stats ought to be higher. If we shift and players adapt with 12-bouncers through the empty side, then fielders look worse than they are.

 

I don't know if that's true, but it seems like a potentially important effect, especially since some teams probably shift a lot more than others and some teams probably shift better than others.

Which is why some teams seem to be pursuing “infielders”, not “second basemen” or “third basemen”. We’ve seen some odd sliding of infielders to different positions last season and this offseason and I think it’s a result of teams realizing that position matters far less than the overall skill set because a third baseman will spend time at short and a shortstop will spend time at second and way deep in the hole near third.

 

All the infield positions, which once had somewhat unique needs, are blurring together. Teams need athletes, not specific positions.

[big dog]

 

Which is why some teams seem to be pursuing “infielders”, not “second basemen” or “third basemen”. We’ve seen some odd sliding of infielders to different positions last season and this offseason and I think it’s a result of teams realizing that position matters far less than the overall skill set because a third baseman will spend time at short and a shortstop will spend time at second and way deep in the hole near third.

All the infield positions, which once had somewhat unique needs, are blurring together. Teams need athletes, not specific positions.

Polanco is certainly an athlete, so hopefully improved shifting will help him.  That will only compensate for range, though.  I guess "bobbles" and throwing errors aren't much improved by shifting, other than perhaps you are more likely to be in better position to field and throw the ball.

 

Perhaps in a few years the big delays in baseball won't be mound visits or batters stepping out- it will be teams switching players like Adrianza to opposite sides of the infield as the count on the batter changes.  Wow, I don't look forward to that.

[Number3]

jorgenswest in part posted:

 

"I don’t think we needed metrics to show us that Arraez, Polanco and Sano were well below average with the glove last year but here is another one with negatives around the infield.

As a team the Twins converted were well below average at converting balls in play into outs. That needs to improve. If not it will lead to longer innings, shorter starts and an overworked bullpen."

 

Two interesting points. The first regarding metrics as opposed to the "eye test" I fully agree with. Show me a practice session one time and I don't need to read some obscure stat. Main thing is to make the routine plays nearly 100% of the time. The ones that make Sports Center's" top 10 plays are made by all of the players one time or another. I remember Polanco makiong a play against the Os that Jim Palmer said was one of the best plays he had ever seen.

The balls in play converted into outs stat is probably such a small differentiator that its impact on a single game is meaningless. Also, does a solid no doubter base hit count as one of those "balls in play" or are pitchers not accountable for this stat? Watch an infield practice session and I will make a judgement; forget the geek stats by people who are trying to find a nitch in a sport that they have probably never played.

[dbminn]

 

Thanks.

 

Making it a "counting stat" seems to me a poor choice, tough to evaluate. But what do I know?

 

They also show percentage success rate v. "expected", which is an average for all players at that position.

[Mike Sixel]

Geek stats by people that never played. Sigh. The best teams are run by those people you insult. Do people really think that measuring things is bad, wrong, evil?

[mikelink45]

This does not do it for me - defensive measurements just cannot gather in all the variables.  We complain about our infield at a time when everyone is trying to hit fly balls and no one is stealing bases, DP are way down and the shifts mean that four players play all over the field. 

 

Get your four best guys in their and let your eyeballs tell you if they are making the plays you want while hitting the way you need.

 

At this moment OF defense might be the most important.

[Linus]

I like Jorge Polanco. He is not a major league SS which is not a slam. It’s really hard to play SS at that level and be good. Hence the high number of players that start there and end elsewhere. Arreaz is about the same at second. The Twins simply can’t have these guys up the middle for the next 6 years. I wanted Iglesias who just signed for peanuts. That would have been a real difference

[Richie the Rally Goat]

Geek stats by people that never played. Sigh. The best teams are run by those people you insult. Do people really think that measuring things is bad, wrong, evil?

the evolution of defensive metrics is trending in the right direction.

 

I’m encouraged by the possibilities of measuring range, reaction and optimum routes to gain more objective determination of what a “good defensive player” and “good defensive play” really are.

 

I still don’t think the fielding side of metrics is at the same level as hitting, but it’s catching up.

[Mike Sixel]

the evolution of defensive metrics is trending in the right direction.

 

I’m encouraged by the possibilities of measuring range, reaction and optimum routes to gain more objective determination of what a “good defensive player” and “good defensive play” really are.

 

I still don’t think the fielding side of metrics is at the same level as hitting, but it’s catching up.

Agreed.

[Brock Beauchamp]

jorgenswest in part posted:

 

"I don’t think we needed metrics to show us that Arraez, Polanco and Sano were well below average with the glove last year but here is another one with negatives around the infield.

As a team the Twins converted were well below average at converting balls in play into outs. That needs to improve. If not it will lead to longer innings, shorter starts and an overworked bullpen."

 

Two interesting points. The first regarding metrics as opposed to the "eye test" I fully agree with. Show me a practice session one time and I don't need to read some obscure stat. Main thing is to make the routine plays nearly 100% of the time. The ones that make Sports Center's" top 10 plays are made by all of the players one time or another. I remember Polanco makiong a play against the Os that Jim Palmer said was one of the best plays he had ever seen.

The balls in play converted into outs stat is probably such a small differentiator that its impact on a single game is meaningless. Also, does a solid no doubter base hit count as one of those "balls in play" or are pitchers not accountable for this stat? Watch an infield practice session and I will make a judgement; forget the geek stats by people who are trying to find a nitch in a sport that they have probably never played.

It fascinates me (and makes me sad) that this attitude exists in a world where a man who never played professional ball took the two most beleaguered franchises in American sports and brought them each a championship.

 

And he did it using math and a bunch of nerds in rooms with computers.

 

Maybe the problem isn’t these new-fangled metrics but your refusal to accept their efficacy because “reasons”.

[Darius]

Sure, the infield defense isn’t stellar. But, before moving people around we need to think the overall implications.

 

Pursuing someone like Iglesias seems like a good idea. The defense would undoubtedly by better. But, how many runs does that take away from the offense? It may end up being a net loss.

 

Then, there is the clubhouse factor that has been a hot topic surrounding Cruz and Hunter around here. Does it have a detrimental effect on clubhouse morale?