THE BOOK--Playing The Percentages In Baseball

Tango, there is a typo:

“Only with Arm Accuracy is there a disagreement, with Dave at 3 and the Mariner fans at 3.1.”

Should be “agreement” of course.

The ratings and run equivalents are as compared to a neutral fielder?  So, for example, if Richie Sexson were a zero in runs and so was Reed (a zero), assuming they played first base and CF, then that would mean that Sexson was a great first baseman and that Reed was a below average CF’er?

If that is the case, this is a really bad way of presenting the data.  At least to the average reader.  It will mislead them and confuse them with respect to how well or poorly a fielder is at his position.

I know you love “neutral” fielding ratings, but you at least have to present the positional adjustments or explain to the readers and fans that these ratings and numbers are all position neutral.  And even then, how do we know what rating is good for each position?

For example, do we (as an average fan or reader) have any idea whether Sexson’s rating is good or bad for a first baseman?  Don’t we want to know that?  Is Ichiro’s rating good for a CF’er or just average?  What about for a RF’er?

Even then, the positional adjustments won’t work perfectly.  Let’s say that you had an outfielder who was exactly average for a RF’er in all respects but his arm is a little above average for all outfielders but a little below average for a RF’er.  His overall rating is going to be a little higher than that of an average RF’er but he is actually going to be a little worse than an average RF’er (because of the arm).

IOW, technically there has to be a different positional adjustment for each category (arm, speed, instinct, etc.) rather than an adjustment for the overall rating.

I should have said “disagreement in optimism”, since in the other 6 Dave had him ahead.  Then the next sentence in that paragraph makes more sense.

***

The intent of the article was only to show the ratings relative to two sources.  The presentation of the Runs-scale was *only* to give the numbers more meaning, much as showing the peripheral numbers of starters and relievers along an ERA scale gives it more meaning, even though we know that the numbers are not comparable to each other.

I could have similarly just looked to see how people ranked players by OBP, converted OBP to runs, shown the list of OBPruns leaders, and not make any statement that a high OBPruns by Utley is more impressive than by Ryan Howard.

That said, I did note:

I should also note that for this article only, I’m treating each of the seven categories the same. Obviously, some traits are more important than others, especially when you consider the position.

I was hoping that that disclaimer would have been sufficient to sidestep the positional and trait-weighting issue, for the particular purpose of the article.