Poll- Would You Make This Trade?

[El Gordo]

It seems to me that how we evaluate defense comes down to what we see. If a manager says I want that guy at SS every day because I have watched a lot of SS play and to my eye he has the best combination of hands range quickness, arm strength, and accuracy of any I have ever seen, then he is relying on what he has seen. As a manager of baseball he is in a better position to make evealuations of a players relative defensive capabilities than most of us because he has seen more. But the +/- sytem is merely doing the same thing in a more systematic and codified way. For the most part when I watch the game, plays seem to be routine and fall in to several basic categories. The unusual ones involving slight differences of spin, velocities, etc. seem few in comparison. And these factors are operating for all players at a position. I don't think that they have that big of an impact on the overal numbers. I would trust the FB system of evaluating defense over the GG method of polling managers. Of course it's not perfect and there is a lot of noise but I think it reflects general trends pretty well IMO.

[SilentJames]

I doubt they are a fluke. The +/- system is pretty accurate when it comes to rating the other top defenders at their positions, so I doubt they are suddenly screwed up when it comes to 2 of our favorite players. The decline may not be as extreme as the numbers are indicating, but I think that the extreme differences between this year at last, indicate a real trend downward.,

One year is not a trend. Paul Blair had five errors in 1969, he then went 3,1,1 over the next three. So this could be just a bad year. It could be an error on Markakis that a lesser fielder would not have gotten to - etc. Errors and fielders are tough sometimes.

We have seen a big improvement in the overall defense already since April, Markakis is included in that. I can't find an efficient way to see how Markakis has been doing since April.

[maudibjr]

I accept. Bear in mind, truly understanding this requires ADVANCED KNOWLEDGE of quantum physics, but you should get the gist.

http://www.mediafire.com/?yoydvl52djm

Nothing happening at work today.

I didn't realize spaceball1 was bigger than the death star

[El Gordo]

All of this makes sense. So, you must believe that Nick has been an awful defensive RF so far this year, right?

I think he has a great arm and that covers up for a decline in range from last season. Whether it is as extreme as the numbers show, I will wait and see as the season plays out. I am looking for improvement. I see that from Jones but I haven't seen that from Nick so far.

[wickedwitch]

What happens if someone else gets to a ball in their zone? Say Jones catches a ball well into Markakis's zone that Markakis can get to. Does Markakis get marked down because of this, because he didn't make a play others made? He certainly wouldn't get credit for it, which would affect his rating negatively.

It seems like a player could be inadvertently penalized for playing next to a good fielder.

But the OF positioning can definitely be improved. Does anyone know if Trembley and his coaches consider how good an OFer is for positioning? If they don't, they should consider spreading Markakis and Jones apart.

ETA: Markakis has only 18 out-of-zone plays, according to the =2009&league_filter[]=1&pos_filter[]=7&pos_filter[]=8&pos_filter[]=9&Submit=Submit"]Hard Ball Times. I assume that Jones's range hurt that.

[PeteCanes]

I didn't realize spaceball1 was bigger than the death star

Who can really say. The poll choices seem to bend and distort time and space. Maybe they caused a massive black hole causing Spaceball 1 to APPEAR larger. Maybe neither of them really exist, or exist at multiple different places of the universe at the same time. All of this is really way out of my league. It boggles the mind.

[olehippi]

Reading through this thread makes me wonder if some other people have been drinking the water from OldFan#5's well.....

[Sports Guy]

I haven't look at the stats but is our staff more of a flyball staff this year vs last year?

If it is and more balls are falling in then last year, could that be what is effecting the numbers for Nick and Adam?

[Sports Guy]

Another question I posed earlier.

According to MLB.com, the Orioles rank 18th with a GO/AO of 1.13. Last year, it was 1.15 and they were 18th. So, no difference it appears.

However, this stat is only outs. It doesn't take into account the hits. I haven't found that info yet. I'm guessing hardball or fangraphs has it.

Looks like the FB% is the same and LD% was actually higher last year...So, there goes that theory.

Its very interesting...Like you said, if these guys are that much worse, why is that the case and like you, I don't see them being way worse...It does seem like Jones isn't getting to as many balls over his head...Nick was shaky at times early in the season but it wasn't like it was several plays or anything like that.

[RShack]

How does somebody decide if Nick missed a play that at least one other player didn't miss?

The system doesn't care about anything except where the ball went, how it was hit, and the outcome of the play. It doesn't care where the fielder was standing, and it doesn't care whether he caught it with or without drama. All it cares about is where the ball went, how it was hit, and whether it got turned into an out or not. For each play, that is what gets input into the database. From interrogating the data for all the plays for all the RF's, the db spits out the +/- results for Nick.

To me, the idea makes perfect sense. I think the one key thing we don't know is about how many attributes and gradations there are for rating how the ball was hit. I don't think we know what the answer is. If I was the guy running that operation, and had money coming in based on its utility, what I would want is computational analysis of video images to make the rating both accurate and meaningful. However, I think that's not viable without multiple camera angles. Last I heard, they used trained observers to rate the ball. Using trained observers is not the problem that some seem to think it is, but we don't know how many ball-scoring attributes and gradations they have to work with. We also don't know how many they really need to produce meaningful results. I think exactly how they do it, and where they think the imperfections might be, is a trade secret.

[Old#5fan]

One year is not a trend. Paul Blair had five errors in 1969, he then went 3,1,1 over the next three. So this could be just a bad year. It could be an error on Markakis that a lesser fielder would not have gotten to - etc. Errors and fielders are tough sometimes.

We have seen a big improvement in the overall defense already since April, Markakis is included in that. I can't find an efficient way to see how Markakis has been doing since April.

Or is it not possible that you (an a host of others here) have simply vastly overated Markakis to begin with? To me at least, not only is this possible it is more likely as true as anything I have ever observed since reading and posting here.

[Lucky Jim]

A few things...

> Reference the question of what happens if Player A catches a ball in Player B's zone, and how does it get counted. The people who operate these systems are extremely smart, and know an awful lot about baseball. I think it is safe to say that they have thought of and addressed this question. It should also know that they go so far as to make park adjustments. It is possible that they are making a different adjustment this season for RF at OP@CY. I'd love to see Nick's H/A splits.

> One thing I didn't see anyone mention (I only looked at the last page), is that the ratings can also be affected by the performance of other players. What I mean is that Nick's 2008 rating was based on how he compared to other players who played RF in 2008. His 2009 rating is based on how he compares to other players who have played RF in 2009. If the cumulative numbers for all other RFers improved this season, and Nick stayed the same or regressed slightly, then that would make a huge difference.

> Reference the question of whether Orioles pitchers are giving up more fly balls this year. They've faced 2801 batters and allowed 840 flyballs. Last year's staff allowed 746 flyballs for every 2801 batters faced. At their present pace they will allow in excess of 200 more flyballs than last year's staff. That very well could have an impact. I'm a strong believer that a pitcher plays a big role in the quality of defense behind him (probably more so than most "DIPS-centric" people). Maybe some of the putrid stuff some members of this year's staff have called pitching has had a negative impact on 2110. If this is the case, then we should see their numbers improve as the season moves on.

> Unlike some, I have a great deal of confidence in UZR and TFB. I buy into most of what they're saying. I'd still love to see how they compare to some of the more advanced metrics that many teams use. For that matter I'd love to see how the publically available UZR compares to the more advanced system that its creator surely uses.

Thanks, 1970. Your faith in these systems has - from the beginning - been a source of confidence for me that they're legitimate.

These systems were designed to counter the real problems of bias that occur in trying to eyeball fielding. As a result, I'm highly skeptical of attempts to disregard them because folks think they see things differently.

None of these stats are a reason to say that Nick (or Adam) is - as a matter of rule or law - a poor fielder. They just point out reasons for concern.