I don't believe you can get much lower than the Replacement Level as a ML concept, unless you take a scrawny kid from the stands [always the last one picked chosing sides] and put him in the field. [Slight exaggeration, that!]:And I think that an average fielding ss who hits as well as an average fielding first baseman is a better player. If we give the players defensive value for what they do in the field above a LOWER baseline, we could plausibly do what you are saying. An average shortstop would be making say 40 more plays than a low baseline, while an average first baseman might be making 20 more than the low baseline for first base. If we set a below average baseline for each position, we could eliminate the positional adjustment.
Since we're talking catchers more specifically, I "get" the concept of Replacement; but as applied as a model it gives a distorted sense of value with a very low standard.
And by the way, the positional adjustment is generally in line with the OFFENSIVE differences in production by position, but not completely in line (DHs are -14 while first basemen are -10 even though they hit/run better for example.)[/QUOTE]
The most common WAR methodologies only rate defense relative to "average" at a position. They assume that playing an entire team of replacement level players will produce a team with a 32% winning percentage. They assume that putting a replacement level player at any position in the field will cost an average team about 2 wins overall but do not parse it out over offense or defense (a replacement shortstop will lose a lot of runs on defense, but a replacement first baseman is going to lose a ton on offense).
There are WAR methods that set a replacment level for offense and defense separately so that an average fielder would get a positive score.
In fact a way to look at it is that a first baseman gets -10 for position, while a shortstop gets +7, but they BOTH get about 22 runs above replacement level with about 25% of the contribution coming from defense and 75% from offense. That would mean theoretically that they BOTH are getting about 5.5 defensive runs above replacement for their positions. Still leaves the first baseman at -4.5 and the shortstop at 12.5 versus a neutral replacement fielder, and it implies that a replacement third baseman or centerfielder is a better fielder than an average fielding first baseman.