The Scouting Report, By the Fans, For the Fans

Part 3 - 1B Report (2002 - 2003)

© Tangotiger

Background

If you are new to this project, you should read about it first.

The Preliminary Results

Thank you to all those who participated! I've received almost 500 ballots (and am still accepting them until Opening Day). For this particular article, let's look at the results of 1B, as you Fans see them.

The Fans Runs category is my own weighting system of the 7 categories, with extra weight placed to Acceleration, Speed, and Hands, and less weight on the three throwing categories. UZR162 is the player's UZR + DParm per 162 Games, from 2002-2003.

The Fans' love affair with J.T. Snow is well-captured here, along with their admiration for Nick Johnson. They certainly don't care for Julio Franco.

A good application of use here is John Olerud and Phil Nevin. UZR sees them as close to even, but the Fans see them as complete opposites. Since sample data is just that... a sample... we need to look for the signal inside all that noise. The best way to do that is to regress the sample data towards the population mean. Normally, this would be the population of 1B, and their mean would be zero. However, now we know (or think we know) something more about Nevin and Olerud. Therefore, their sample UZR should be regressed, somewhat, towards the Fans' evaluation. (How much is yet to be determined.) At a current first guess, I would regress about one-third of the way towards the Fans. This would put Olerud at +3 and Nevin at -6. J.T. Snow would go from a UZR of -4 to an overall of +3.

Pay special attention to the last line. It's a correlation (r) between each category and UZR.

Notes

To qualify for these lists, a player had to have at least 5 ballots for each category, as well as 40 games in each of 2002 and 2003.

I converted the 5-point scale (Great, Good, Average, Fair, Poor) into a scale from 0 to 100, with 50 as the average for all fielders. One standard deviation was set to 20 (meaning that two-thirds of all players' traits will fall within the band of 30 to 70; if a player is higher than 70, then he falls in the top 16% in that category for the league).