THE BOOK--Playing The Percentages In Baseball

Tuesday, February 16, 2010

Is there any meaning to Poz’s arbitrary list of ranking players?

Poz said this:

I should tell you that a big part of the system involves the ranking of players. I ranked each team’s top 3 starters, closer and everyday players on a 20-80 scouting scale, with 80 being Pujols and, well, here’s a little cheat sheet:
80 — Albert Pujols
75-79 — The best of the best not named Pujols. Only a handful of players in baseball in this group. Joe Mauer is a 78.
70-74 — Great players. Chase Utley is a 73.
65-69 — All-Star caliber players. Kevin Youkilis is a 68.
60-64 — Very good players. Raul Ibanez is a 61.
55-59 — I’d say 55 is about an average every day player. I have Cleveland’s Jhonny Peralta as a 55.
50-54 — Now, we’re getting into below average territory, but these guys can be very useful. Mark Teahen is a 53.
45-49 — The worst every day players. For various reasons, I did not allow any everyday player to score lower than 45. So, yes, Yuniesky Betancourt was a 45 for me.

I won’t bore you with all the ratings, but I do list off the players who ranked highest at each position in each division*. I then used those ratings to calculate some predictions. Like I say, complicated and ridiculous. But it’s February. And, really, something has to get us through the winter months.

Now, it is obvious the spirit in which Poz wrote this.  Not only is it obvious, he explicitly says it.  But, does it actually tell us more than what Poz himself cautions against?

First off, I should tell you that I love the 20-80 scouting scale.  It is, in some way, a precursor to wins above average.  Think of the average MLB as a 50 (.500).  Think of the one of the best pitchers around as a 70 (.700), and the best possible pitcher ever as an 80 (.800 pitcher).  An MLB player on the bubble a 40 (.400), and a triple A player that might, might, see MLB action as a 30 (.300) and a player with no hope of even riding the bench as a 20 (.200).  It’s all so very lovely, it doesn’t ask for much precision, just a sense of accuracy.

He’s got the worst (every day) player in the league as 10 units below the average everyday player, while a typical star player is 13 units above the average regular, and the absolute limit is 25 units above.  Basically, PozUnits * 2 = runs.

Poz doesn’t show it, but I would guess that if you add up all the PozUnits by team and divide by 5, and compare the results to the average, and add 81, that gives you PozWins.

So, no, I don’t think what he did was complicated and ridiculous.  Indeed, what he did was the exact opposite of that: simple and insightful.  It forces everyone to rank the regulars of every team, not just your own team.  It forces everyone to rank players along some system, rather than just ordinal rankings or worse.  What he doesn’t do is look at the team’s bench and half the pitching staff, reasoning that all those players are basically, a crapshoot: there’s not much to choose there.  I don’t know that I’d go that far.  I think he should have gone to a 4th starter, and to a setup guy.  But, we’re probably splitting hairs at this point.

What Poz did was the first step.  That’s not ridiculous.  The ridiculous thing would be to give your predictions and not even do the first step.