Giles was ranked second for right fielders and Gerut was fifth for center fielders, ahead of Cameron at 10 and Edmonds at next to last. And Maddux was second for pitchers.

I do follow sabermetrics though, to the point where I have a better understanding of how stats work than your average broadcaster.

That said, and here’s more copping out, I probably know the least about sabermetric methods of defensive analysis.

But I have to say that in terms of subjectivity, I believe using errors and fielding percentage, which is derived from errors, is pretty high up there. There are two reasons I think this.

1. Errors don’t consider the first, say 80% of the defensive play. One example: If a fast fielder reaches a far ball and drops it, it’s an error. But if a slow fielder doesn’t reach the same ball it isn’t one.

2. Even once a player reaches the ball, the awarding of an error is up to the judgement of the official scorekeeper.

I agree it sucks to use the “while it’s not perfect” for any ananlysis, but there are also varying degrees of imperfection. I think errors are a lot less perfect than the sabermetric methods. I believe looking at the correlation charts in the second link above confrims this. Each defensive system is independently computed, and a fair amount of correlation exists between them. That shows you these systems are at least on the right track if they can independently come up with similar results.

I’ve never seen something like that done for errors or total chances, but I have a hard time believing there would be such a strong correlation.

As for your example of an inaccurate fan rating of Jody Gerut, my guess is it’s a result of fewer opportunities for fans to rate him. He only played in 90 games this year, he played no in major league games his previous three years, and he played for a small market team in last place. Fewer data points mean a less accurate result.

And finally, absolutely confirmation bias is an issue I think everyone has to overcome. I know I catch myself falling prey to it a lot, which is one of the reasons I enjoy sharing different points of view with others. I encourage anyone to call me out on anything, and if you do I will listen and always consider other opinions.

1st – In watching more than 600 games this season on DVR, I am constantly looking for some specific items like distance covered by an outfielder to make a catch from a subjective standpoint and then I calculate exactly how far that fielder had to go and where they started running from based on triangulating from given points using the ballparks measurements and baseballs standard measurements. Gerut had 3 of the 10 longest runs that resulted in a catch in those 600+ games. Yes I know there were 4860 games this year, but that is a pretty god sample size. That he ranked 27th of the outfielders listed is totally out of my observations, measurements and his basic stats.

2nd – As a coach, teacher and a trainer in business, the one thing I have to keep in mind constantly is –

Confirmation Bias.

If I believe something, I can consciously or subconsciously devise a rationale that will “prove” my point, even when I am wrong. A very human condition.

And the better we get at mathematics, the easier it is to “prove” a point in the way that fits our hypothesis.

One faulty subjective assumption can still LOOK good in a mathematical equation, while still being totally off base.

That is my problem with all the Sabermetric models that anyone has to start off by saying “while it is not perfect” while they are using that model. If that is where you have start, then its probable that you or the model are not being objective. The model is probably based on a faulty, subjective assumption.

Everything else is based on opinion of how a person has done.

After 10 years of playing the game in high school, college and professional ball and another decade working in the game at all those levels as a coach or scout I have found that no one, and I mean no one, has the same opinion of what constitutes good defense, but long term success at a position has almost always correlated very well with guys that are at or near the tops of those three stats.

If someone can show me otherwise in a sabermetric model and make me understand how its not completely subjective, I am certainly willing and eager to listen. I personally love the evolution of the game since Bill James and the other geeks that became SABR got their hands on stats and shook up the stodgy baseball community.

So if there is a fielding metric that is not based on total subjectivity, explain it in a future post.

Here are a couple articles comparing the fans’ scouting report with other defensive metrics, on which the fans scored pretty well. Which stats are you referring to?

Even including guys that had 100 innings or less, most of these rankings have no basis in what the stats actually indicate.