Quote:
Originally Posted by StyxNCa
This is totally idiotic. Using the GP system, some very good defensive players get screwed big time in range factor. Ken Berry is one that comes to mind off hand since he was about a 5 IP/G, more or less, type of player. Using the straight GP system makes him look like he had no range where his IP shows he had pretty good range.
That kind of mindset makes me wonder what he's smoking.
I bet he finds the IP info useless since somehow that kind of data won't support something he believes. That's pretty much how numbers people handle data that doesn't support their beliefs (useless, insignificant, irrelevant, etc). The three most used words by numbers people against someone doesn't agree with them.  |
I could think of one reason to support James' view. Not because I want to support James specifically, but because I've used the concept myself.
If a player gets only 2 chances in a game because he came in as a replacement in the 8th inning, you can see how this player is hamstrung if you look at TC/G versus TC/Inn at this granular a level.
But over very large periods of time, if the player's quality of fielding isn't good enough to ever become a starting player (a truly superior fielder should have some level where the manager recognizes the value to the club even if he cannot hit), why consider him an elite fielder?
I'm not even saying I buy that entirely. It also could be a rationalization for Win Shares, since we have granular data for hitters and pitchers for every season for baseball, but we don't have innings played for fielders for the early seasons of baseball.
In one of my leagues, I used that the concept of playing time equals the quality of player for some ranking purposes. It is part of a concept that the quality of a stat is not equal to the quality of its complexity.
The idea is that over several seasons:
- If a player is constantly injured, this reduces his quality in the sense that you can talk about how great C.C. Sabathia is, but quantitatively he's really not much better than Bruce Chen if he's on the DL for the majority of the last two seasons. He's theoretically much better, but not actually much better.
- A team self-selects their best players for playing time. If you have Alex Cora starting at SS for several seasons straight, he must have some level of quality, because the lower the quality of a player, the greater likelihood that he will eventually be platooned, replaced by a backup player, replaced by a minor leaguer or replaced by a free agent.
Now, there will be exceptions, but as a general rule, it works surprisingly well. Not that I've ever seen any system that hasn't had an exception. As a matter of fact, the system had less complaints than an Elias-like system or a Win Shares based system ever had.
Part of that is because it doesn't try to be too fancy and people can understand the concept without a complex algebraic formula. The other part is that it didn't need to be too specific in what it was measuring. It wasn't really trying to measure that Player
x is better than Player
y because his rating was a few percentage points higher, it was used to group players based on their relative worth to their team.
Anyway, I really have digressed.
By the way, Ken Berry played 7.8 IP in the OF per game played over his career. 10184.1 IP / 1311 G (he had 1098 GS).