This is the point. The replacement pool for CF, LF, RF is identical... it's one pool.

Yes it's one pool but there are a bunch of different people in the pool, with different attributes - this is the point.

Let's say there's a guy called Mo Ramirez, Manny's light-hitting cousin. He's league average with the bat, and -2 wins with the glove in left field. Then there's Cito Beltran, Carlos's light-hitting cousin. He's -4 wins with the bat, and +1.5 wins with the glove in left field. Both are in the replacement-player pool for OFs. And we know that if we put a guy from LF into CF he gets twice as many chances in the field so his fielding rate stats double in value. He doesn't get any extra ABs, of course. Well, in LF Mo is a -2 win player, while Cito is -3 wins. But in CF Mo is a -4 win player, while Cito is -1 win. So if we need to replace our LF we'll call up Mo, but if we need to replace our CF we'll call up Cito.

So although the pool is identical the replacement level is different for CF and LF.

Doubling the value of your fielding rate stats is for demonstration purposes only, obviously. You can make it whatever factor you like. But the point is that it is multiplicative.

Dan, I think I should be the one thanking you! I have learned a tremendous amount from your research and writing, both about how to properly value baseball players throughout time and the history of the game. I would like to add my voice to those encouraging you to take your research to a more public forum when you are satisfied with it. Among the "casual sabermetric" crowd--the people who read a little Bill James, a little Baseball Prospectus, but aren't total obsessives like us--you hear WARP3 and WS and all sorts of other silly stats getting thrown around all the time. I think it would do a lot of good for general baseball discourse to have an Uber-stat as good as your WARP to refer to. My one piece of advice, should you choose to do so, is to find a different name for it; otherwise, you are just begging for confusion.

I don't know Tango, position-switching in general makes me nervous. Related to a point that AlouGoodbye made about SS and DH, I would expect it to be badly intransitive. I would expect the SS to 1B gap you would see by going SS to CF to LF to 1B to be very different from what you would see by simply swapping all 1B and SS (I am now trying to picture Prince Fielder doing an Ozzie Smith-style backflip...) The reason for that is, at most positions, you see a biased sample of players going between positions, and the more steps you need, the more bias that gets introduced. In general, an empirical based replacement level seems preferable.

Having said that, in the CF to LF/RF situation, position switching does seem more compelling; I don't have a settled opinion on the matter. I'll think about it more.

Alou, definitely if you consider that the CF gets 33% more chances, then we need to include that in the translation. It's not a straight additive translation necessarily. It could be, but we don't know.

Blackadder: right, baby steps. Let's focus on OF first. Then you can do IF. Then OF/1B and IF/1B. Then OF/IF. We have the Fans Scouting Report so we know the kind of players that move around.

The replacement pool for CF, LF, RF is identical... it's one pool. Managers may inefficiently allocate it (or the non-fluidity of player movement between teams may not make it very stable year-to-year), but it's the same pool.

I don't know about that. I think the same statement could be made about MI/OF. There's always lots of talk about moving Jeter to CF or any speedy players, and the Mets routinely play IF in the OF (Easley, Tatis, Anderson). I think the pool for CFs is smaller than for LF/RF. Kruk or Luzinski just aren't getting to play CF, and ther eare lots of those guys (I think).

A Luzinski type could effectively replace a Cf. If your cf went down, and Greg was the best overall of you had, then he goes into a corner and one of your starting corners goes to center. With middle if the problem is it's a one way switch. You might consider moving your ss to center but almost no cf will move to short.

I thought CF was harder, not just because there were more total chances, but that the chances themselves tend to be harder to handle? I know that this is true between 2B and SS, which apparently are quite close in chances even though SS is a lot harder.

Average NL chances per inning (2008):
Yes, that includes IF line drives, but those are a small, small number.

So, what would baseball's equivalent to Orr be? Maybe a catcher or shortstop putting up 6 consecutive MVP-type seasons in his first 8 years.

ARod. Bench.

I do use my own numbers in my NY Times columns, Blackadder. Thanks for the kind words. This is a gigantic project, but I will finish it someday. :)

Chris: I notice that right fielders handle more chances than left fielders.

I would have thought it would be the other way around.

Has it ever been thus?

right fielders handle more chances than left fielders

Somewhere once I heard this explained: a majority of batters are right-handed, so a majority of base hits go to LF. But a majority of sliced flyball outs go to RF. Hence a RF will make more POs, and have more total chances, on average, even if a LF touches the baseball more often.

Somewhere once I heard this explained: a majority of batters are right-handed, so a majority of base hits go to LF. But a majority of sliced flyball outs go to RF. Hence a RF will make more POs, and have more total chances, on average, even if a LF touches the baseball more often.

This is exactly correct.

-- MWE