Fangraphs: Twins may have weakened a weakness

[diehardtwinsfan]

Honestly, I don't trust defensive metrics, and when they say they take 3 seasons to even be reliable (which if I remember correctly is how UZR works), I'm even more skeptical.  I do think that the Twins could have looked for better defensive options than Hunter, as the team needed a LF.  It's pretty clear they think Rosario or maybe even Hicks can be the answer there, as I would have expected a  2 or 3 year deal for someone given that there's very little else in the pipeline for the high minors (though Harrison and Walker should be auditioning in AA for 2015). 

 

Like it or not, Hunter is a 1 year filler who I'm guessing they expect to teach some of the younger guys a thing or two about how to play OF defense, even if he is physically incapable of playing it well anymore.  Not sure I like that, but that's what it is. 

 

I think the frustrating thing here is that the Twins may end up being .500 this season, and OF defense is going to be a big reason why they are just pretenders and not contenders.   I might be a tad optimistic, but I think this team is much closer than the record last year indicated.

[jorgenswest]

Honestly, I don't trust defensive metrics, and when they say they take 3 seasons to even be reliable.

That's reasonable as long as the same mistrust is placed on batting average, ERA and other commonly used measures. Numbers that can be trusted and used more confidently in comparing a skill or looking for a change in skill are strike out rate, walk rate, ground ball rate and fly ball rate.

 

Projection systems use at least 3 years to project offensive or pitching performance. The same would be said about projecting defense.

[Kwak]

We don't need three years of statistics to conclude that the Twins OF is poor by ML standards--eyeballs will suffice.

[diehardtwinsfan]

That's reasonable as long as the same mistrust is placed on batting average, ERA and other commonly used measures. Numbers that can be trusted and used more confidently in comparing a skill or looking for a change in skill are strike out rate, walk rate, ground ball rate and fly ball rate.

 

Projection systems use at least 3 years to project offensive or pitching performance. The same would be said about projecting defense.

 

Average can fluctuate a bit over a career as there's not a huge difference in terms of the number of hits between a .300 hitter and a .270 hitter, but you'll see a guy struggle in his first couple of seasons and then eventually settle in on some consistency, (baring injury). 

 

You don't see that progression with defensive metrics at all. 

[TheLeviathan]

That's reasonable as long as the same mistrust is placed on batting average, ERA and other commonly used measures. Numbers that can be trusted and used more confidently in comparing a skill or looking for a change in skill are strike out rate, walk rate, ground ball rate and fly ball rate.

 

Projection systems use at least 3 years to project offensive or pitching performance. The same would be said about projecting defense.

 

Frankly nothing is 100% trustworthy for projections because you're basically getting out your crystal ball.

 

The problem with defensive metrics is they aren't reliable even as an evaluative tool for what has actually already taken place on the field.  ERA, BA, and other commonly used measures are.

[Parker Hageman]

I feel like I cover this topic...hmm.

 

http://twinsdaily.com/_/minnesota-twins-news/torii-hunter-in-the-outfield-just-how-bad-can-it-be-r3230

[stringer bell]

That's reasonable as long as the same mistrust is placed on batting average, ERA and other commonly used measures. Numbers that can be trusted and used more confidently in comparing a skill or looking for a change in skill are strike out rate, walk rate, ground ball rate and fly ball rate.

 

Projection systems use at least 3 years to project offensive or pitching performance. The same would be said about projecting defense.

To me, the difference between the measures of ERA or BA versus fielding metrics is that I don't believe fielding metrics even tell us how a player performed in the time measured. That is if a batter hits over .300, it is certain that he gets a hit at least three times out of 10 official at-bats whereas if somebody has poor fielding metrics in a year, we can't be absolutely sure that he was good in the field during the time measured.

 

When predicting future performance, we know that a good ERA or BA doesn't tell us for sure that what the future will hold.

[DaveW]

Defense statistics are fine but shouldn't be the only thing looked at. They vary way too much year to year, but after you have a large sample size I think it helps paint a picture weighted equally with the "eye" test.

 

The defense will be fine moving forward. I still think hicks ends up being at least average in CF, Hunter honestly can't be as bad as his "numbers" and Arcia...well....he is still young at least? Hope for improvement?

[kab21]

It's also important to remember that going forward after this season the Twins likely (hopefully) have an OF defense with Buxton, Rosario and Arcia.  That's actually a pretty good defense and certainly much better than last year.

[Mike Sixel]

This was a topic about if the Twins plan to have a terrible OF defense, and a fly ball staff, was a good plan or not.

 

Somehow, it turned into a discussion about the value of defensive metrics. Can we ever have a discussion about defense that isn't an argument about the stats? Or, do we always want to distract the conversation away from the Twins' plans, and onto something else?