A's Trade 3B Donaldson to Toronto

[TheLeviathan]

Clutch has nothing to do with it - they were 1-11 in their first 12 one run games.  Care to guess how many unearned runs scored in those 11 losses?

 

It's not about being "clutch" fielders, it's about giving away runs and outs in a lot of close games, which only magnifies the problems with the bullpen.

 

Fangraphs agrees by the way.

[Otto von Ballpark]

I didn't disagree that their defense is poor. But unearned runs are included in pythag. All runs are included in pythag.

 

Are they disproportionately bad defensively in 1-run games?

[TheLeviathan]

 

I didn't disagree that their defense is poor. But unearned runs are included in pythag. All runs are included in pythag.

Are they disproportionately bad defensively in 1-run games?

 

They wouldn't have to be, one run games reduce the margin of error which makes defensive problems more exacerbated.  The fangraphs article makes my point more thoroughly than I care to spend time doing again.

[Otto von Ballpark]

The Fangraphs article makes the same assertion as you, but doesn't really explain. Lots of errors should lead to lots of extra runs allowed, which would affect pythag.

 

Put another way, were they more likely to lose a game 6-5 due to errors that they should have won 5-4, than to win a game 5-4 that they should have won 5-2 without errors? Errors unlike bad relievers are more or less evenly distributed.

[TheLeviathan]

The article grasps what you seem not to - bad defense effects almost every game and generally reduces your margin for error.  Which certainly contributes to being bad in 1-run games.

 

Looking at the aggregate of runs allowed isn't going to tell the whole story and never will, that's why it should be a quick and dirty guide to point you to look at the team's key deficiencies.  In Oakland's case that is an impressive ability to rack up errors and a bullpen that appears particularly good at giving up just the wrong number of runs.

[Otto von Ballpark]

"Bad defense affects almost every game"

 

That is exactly my point. Bad defense will cause you to lose some close games by 1 run -- but it will also make 1 run losses into 2 run losses, and 2 run wins into 1 run wins. It really shouldn't affect your overall record in 1 run games that much, any more than it affects your record overall... unless you are unlucky about their distribution.

 

The article doesn't address that at all. It shows how bad their defense is, makes the same claim as you, but doesn't show the connection.

[Otto von Ballpark]

Bullpen I understand, their appearances get concentrated into late innings and close games. And every tie or lead that the bullpen blows in the 9th inning or later is more likely to end up with a 1 run differential.

 

For bad defense to have a similar effect, its occurrence would have to be skewed to those situations. (Similar to hitting and RISP situations, etc.)

[TheLeviathan]

I'm going to just stick with the crazy idea that bad defense helps you be a bad team.  

[Otto von Ballpark]

I agree with that, obviously. It was the connection to 1-run games that was at issue.

[TheLeviathan]

Except you don't agree with that.  You cited their pythagorean record in response to my suggestion that their awful error total and general bottom 5 ranking in most defensive metrics was another major reason for their problems.

 

So you somehow agree that their defense is hurting them....but that it's not hurting them?  Everytime I said "Their defense is bad and that's hurting them" you retorted with "yeah, but run differential"

 

You've totally lost me on your point.  Mine is simple - they have a bad bullpen and a bad defense.  That explains a lot of why they're bad.

 

[Otto von Ballpark]

Not trying to be difficult, really.

 

But overall bad defense is going to affect your overall run differential, just like overall bad hitting or overall bad pitching.

 

Obviously pythag is jus an estimate, but you can't explain away a large pythag discrepancy by saying bad overall hitting/pitching/defense. It has to be situational, which usually means bullpen, RISP, and/or plain old luck.

 

So I am saying they can have a bad defense, even a very bad defense, but bad luck may still be the larger factor. Especially if bad defense gave them a bunch of 1 run losses, but did not apparently move the margins much in their other games (which would in total show up in their overall run differential).

[TheLeviathan]

I didn't explain it all with defense - it is the bullpen and the defense.  Defense, as we Twins fans should be well versed in lately, has a way of creating plenty of "bad luck".

 

Whatever bizarre way you want to say the defense is hurting them, the point is that their defense is awful and much of the cause of that was directly caused by Beane's trades and the players he targeted in return.

 

But I'm done, they have a bad defense and a bad bullpen.  Those are reasons 1A and 1B, whatever order you want to put them in, for their crappy play this year.

[Otto von Ballpark]

Put another way: their run differential suggests 53 pythag wins. Simply with better defense, they might have prevented more runs, to suggest, say, 56 pythag wins.

 

Still doesn't explain the gap between the 53 estimated and 44 actual wins. Unless most of the bad defense was concentrated mostly in close games... which sounds like situational fielding, which in turn sounds like luck.

[jimmer]

Athletics are tied for 11th in DRS (5th in AL) with 11 DRS.  We are tied for 25th (tied for 12th in AL)  with -16 DRS

[Otto von Ballpark]

I haven't looked closely at the defensive stats, but I do recall that Lawrie fared much better by DRS than UZR this year, and last I remember Torii Hunter was the reverse...

[jimmer]

 

I haven't looked closely at the defensive stats, but I do recall that Lawrie fared much better by DRS than UZR this year, and last I remember Torii Hunter was the reverse...

Defensive Runs Saved uses a rolling one-year basis for the Plus/Minus system, while UZR uses several years of data to determine each play’s difficulty level. 

 

So DRS uses the most current level of defensive talent at positions to set the plus/minus level and UZR takes into account the talent level over several years to include players who may have retired (or switched positions or are no longer in the game).  

 

It's why I like DRS better.  Why compare a current players defense to a guy who retired two years ago at age 38 who was bringing down the average talent level?

[USAFChief]

Defensive Runs Saved uses a rolling one-year basis for the Plus/Minus system, while UZR uses several years of data to determine each play’s difficulty level. 

 

So DRS uses the most current level of defensive talent at positions to set the plus/minus level and UZR takes into account the talent level over several years to include players who may have retired (or switched positions or are no longer in the game).  

 

It's why I like DRS better.  Why compare a current players defense to a guy who retired two years ago at age 38 who was bringing down the average talent level?

Most proponents of defensive metrics will say three years of data are needed to get the better picture of ability. Why would using one year as a baseline be better than using several?

[jimmer]

 

 

 

Most proponents of defensive metrics will say three years of data are needed to get the better picture of ability. Why would using one year as a baseline be better than using several?

I already explained why.  Because it uses the abilities of more current players not players who may have retired, only played one season, have since left the game, moved to other positions, etc.  And I explained why I think that is better. I want to know how a player is doing against players who are more likely still playing the same position.

 

I'm not going to be involved in derailing another post on the validity or application of metrics.  Most moderators have made it clear not to do that, no matter how many times a trolling post tries to turn the discussion that way.

 

I pointed out Oakland's DRS rating because people were saying how bad Oakland's defense was and I showed one stat that we may use to say maybe not so much.  Then I tried to explain why sometimes there's a difference between UZR and DRS. That's it.  People can take it or leave it, doesn't matter to me.  Said my piece and now their debate can continue. But they have to be basing their belief Oakland's defense has been horrendous on something.

[Brock Beauchamp]

 

Most proponents of defensive metrics will say three years of data are needed to get the better picture of ability. Why would using one year as a baseline be better than using several?

It's an interesting question but the several years of data required is to evaluate a single player. Over the aggregate, one year should be adequate to rate the current defensive ability of MLB.

 

I think both metrics are flawed but have their place in today's game. They're a good counterpoint to what we watch on a daily basis.

[jimmer]

Brock since all stats, traditional and metric, give us true talent level with three years of data, I'm waiting for them to decide to only give out awards every three years instead of yearly.  MVP, Gold Gloves, Silver Sluggers, Hank Aaron, Cy Young, Fielding Bible awards, etc. Makes sense, no? Since all we care about it true talent level of a player as opposed to what a player did any given year. Pretty much have to stop doing the rookie of the year award completely, because it says year in it :-)

 

Should we do that with that for the manager award too?  Might be hard:-)

[drjim]

I'll give Beane credit, he has done very solid work. Didn't mess around, moved the guys that he needed to move, and got solid returns.

 

Do think this will lead to a little bit of a longer downturn than usual but they are reloading.