Ok, so what does this represent? The first line is the gross wins, the "game impact" as I called it. It is the number of wins that player is worth. You cannot go below 0. This is akin to Runs Created. This second line is simply the number of wins plus/minus to the ML AVG SS. This number is akin to Linear Weights. The third line is the number of wins over the ML AVG PLAYER, REGARDLESS OF POSITION. This is why Hubies is worth 0, since he is our fictitious player that is let's say the 150th best defensive player out of 300 defensive players in the league (you should probably include backups, etc, but you get the idea).

So, if you insist on using the second line (the LW line) that is fine, as this forces EACH POSITION to come out to a total of zero. HOWEVER, we need a positional adjustment. What David is talking about, and what is used with Total Baseball, is to simply apply the negative of the offensive Linear Weights. That is, if the Off LW for SS is -1.1 wins, then the positional adjustment should be +1.1, so that off+def+positional adjustment adds up to 0. While this is okay if we choose maybe a 10-yr time span to calculate this adjustment value, as I have shown in prior posts, this is just plain wrong for a year-to-year basis.

If we are to stick with the positional adjustments based on the methodology I have proposed, the positional adjustment for SS comes in at +3.7. Admittedly, that is a whopper to accept, and I don't accept it either. What happens is that I gave the range for a SS of .90 to .60 of balls caught in zone. It could very well be that Cecils, the worst fielder in the league, would get to .70 of the balls (or a .20 range). If that is the case, the above table would become the following

Wins	Ozzes	Hubies	Cecils	ML AVG SS
Gross Wins	8.4	4.2	0.0	5.9
Wins over ML AVG SS	2.5	-1.7	-5.9	0
Wins over ML AVG player	4.2	0	-4.2	1.7

Now, based on this assumption, we need to apply a positional adjustment of +1.7 wins to all SS.

So, this is what I am talking about. Let's try to figure out how different levels of players would do at different positions, and we can calculate the positional adjustment value.