Baseball Savant Infield Outs above average explained

[Tony-OH]

https://www.mlb.com/news/statcast-introduces-outs-above-average-for-infield-defense

Here's the article that explain defensive OOA in case you missed it. Like I dis with the outfielders, in my mind this is the best system available to determine how good or bad an infielder is defensively.

[tntoriole]

10 minutes ago, Tony-OH said:

https://www.mlb.com/news/statcast-introduces-outs-above-average-for-infield-defense

Here's the article that explain defensive OOA in case you missed it. Like I dis with the outfielders, in my mind this is the best system available to determine how good or bad an infielder is defensively.

Does the methodology hold up as well on the bad side of the analysis?  Meaning, I can see where the above average outs ranking can stratify the truly good separated from the mediocre to bad, but within the bad grouping does it hold up?...i.e. both Villar and Richie Martin were under average ratings with Villar at -4 and Martin at -5....does that mean Villar was a “better” ss than Martin or that they were both just under average? 

[Tony-OH]

11 minutes ago, tntoriole said:

Does the methodology hold up as well on the bad side of the analysis?  Meaning, I can see where the above average outs ranking can stratify the truly good separated from the mediocre to bad, but within the bad grouping does it hold up?...i.e. both Villar and Richie Martin were under average ratings with Villar at -4 and Martin at -5....does that mean Villar was a “better” ss than Martin or that they were both just under average? 

From what I can tell, it just means Villar was slightly better than Martin but both were among the worse at SS. If you read through how they come up with the numbers, each play gets a percentage up to 100. 

I also like looking at success rate added. In my mind, if you are -5% success added, that means you would get to 95 balls out of 100 that the average fielder at your position would convert to an out. 

[tntoriole]

9 minutes ago, Tony-OH said:

From what I can tell, it just means Villar was slightly better than Martin but both were among the worse at SS. If you read through how they come up with the numbers, each play gets a percentage up to 100. 

I also like looking at success rate added. In my mind, if you are -5% success added, that means you would get to 95 balls out of 100 that the average fielder at your position would convert to an out. 

Thanks...the good news is Iglesias is markedly better than either at +12 and 6th overall last year among SSs. 

[Frobby]

My one question about their methodology is they take the average sprint speed of the runner into account, rather than his sprint speed on the particular play in question.   Their reason is that:

“A runner's average sprint speed is used in the calculation, rather than "on that play," because a fielder has to plan for a runner's best, even if on some plays a runner jogs, trips, etc.”

Well, do they?    Can’t a fielder often see where the runner is and gauge his throw accordingly?

But I’m not sure it would make much difference if they did it the other way.   

[Just Regular]

http://mlb.mlb.com/fan_forum/podcasts/index.jsp?c_id=mlb&podcast=statcast_podcast

Architect Tom Tango is interviewed by Petriello in the podcast this week, and digs into some of those nuances.  One of his comments was along the lines of...it probably doesn't matter if Billy Hamilton is dogging it, the infielder is just going to be thinking Billy Hamilton.  And vice versa, even if Brian McCann is busting it.

[Frobby]

13 minutes ago, OrioleDog said:

http://mlb.mlb.com/fan_forum/podcasts/index.jsp?c_id=mlb&podcast=statcast_podcast

Architect Tom Tango is interviewed by Petriello in the podcast this week, and digs into some of those nuances.  One of his comments was along the lines of...it probably doesn't matter if Billy Hamilton is dogging it, the infielder is just going to be thinking Billy Hamilton.  And vice versa, even if Brian McCann is busting it.

I’m not sure I buy it, and I wonder if that judgment was evidence-based.    

[Gatoriole]

27 minutes ago, Frobby said:

My one question about their methodology is they take the average sprint speed of the runner into account, rather than his sprint speed on the particular play in question.   Their reason is that:

“A runner's average sprint speed is used in the calculation, rather than "on that play," because a fielder has to plan for a runner's best, even if on some plays a runner jogs, trips, etc.”

Well, do they?    Can’t a fielder often see where the runner is and gauge his throw accordingly?

But I’m not sure it would make much difference if they did it the other way.   

I'm not sure how much difference it makes either, but I agree with your point.

Also, if a fielder has to "plan for a runner's best," shouldn't they use the player's maximum sprint speed rather than average sprint speed?

[Frobby]

9 minutes ago, Gatoriole said:

I'm not sure how much difference it makes either, but I agree with your point.

Also, if a fielder has to "plan for a runner's best," shouldn't they use the player's maximum sprint speed rather than average sprint speed?

Arguably.   To be clear, the sprint speed they use is on “competitive” plays, as further described here:  https://baseballsavant.mlb.com/sprint_speed_leaderboard

[Gatoriole]

45 minutes ago, Frobby said:

Arguably.   To be clear, the sprint speed they use is on “competitive” plays, as further described here:  https://baseballsavant.mlb.com/sprint_speed_leaderboard

Yeah, the distinction I make probably makes little to no difference then. But I still agree with your point. We've all seen defenders, particularly infielders, who took as much times as they had based on how fast the runner is running on that particular play. JJ Hardy comes to mind.

[Tony-OH]

1 hour ago, Frobby said:

I’m not sure I buy it, and I wonder if that judgment was evidence-based.    

At the end of the day it's based off whether an out was recorded? Why would Hamiliton be dogging down the line on tough play? Not sure what your issue here is. It makes total sense to use the avg sprint speed because on the close plays, where it matters, there going to be very little "doggin it."

[Tony-OH]

7 minutes ago, Gatoriole said:

Yeah, the distinction I make probably makes little to no difference then. But I still agree with your point. We've all seen defenders, particularly infielders, who took as much times as they had based on how fast the runner is running on that particular play. JJ Hardy comes to mind.

But it doesn't matter. If a guy is dogging it down the line and the infielder sees it and tossed it over, they still get the out. The difficulty of the play would be measured by the average sprint speed because that is the actual difficulty. I would still argue a difficult play would not see a runner dog it down the line.

 

[Gatoriole]

19 minutes ago, Tony-OH said:

But it doesn't matter. If a guy is dogging it down the line and the infielder sees it and tossed it over, they still get the out. The difficulty of the play would be measured by the average sprint speed because that is the actual difficulty. I would still argue a difficult play would not see a runner dog it down the line.

 

I misunderstood the import of the runner's average speed. I understood it that if the runner ran slower than his average and was thrown out, but if he had run his average speed and would have been safe, then it was held against the fielder. 

I looked at the linked article and it looks like the speed is used to determine the likelihood that the defender will make the out. That makes more sense.

Nonetheless, I'm still not sure the average speed rather than speed on the play should be factored in. If the runner runs faster than his average speed, and the fielder doesn't make the out, the formula hurts the fielder based on how difficult the play was expected to be rather than how difficult the play actually was. Interesting stuff though.

[Tony-OH]

20 minutes ago, Gatoriole said:

I misunderstood the import of the runner's average speed. I understood it that if the runner ran slower than his average and was thrown out, but if he had run his average speed and would have been safe, then it was held against the fielder. 

I looked at the linked article and it looks like the speed is used to determine the likelihood that the defender will make the out. That makes more sense.

Nonetheless, I'm still not sure the average speed rather than speed on the play should be factored in. If the runner runs faster than his average speed, and the fielder doesn't make the out, the formula hurts the fielder based on how difficult the play was expected to be rather than how difficult the play actually was. Interesting stuff though.

I just don't think fielders take their time because they see the runner running at different speeds. I think fielders play each play as quickly as they can with some variances of knowing the average speed of the runner. In other words, fielders will rush a play when Hamilton is running compared to when Pujols is running.

[HowAboutThat]

YakYakYakMartinisterribleYakYakYak

 

and that makes me kind of sad. I was hoping he was better than he appears to be.