# Fielding Win Shares

On a more serious topic. In my first post here, I talked about developing a Win Shares type system that was much simpler than James’ and made more sense. I’ve already talked about how to do so for pitching, with pitching runs created. The hitting part of it is obvious (just use runs created). But what about fielding?
What James does, in simplest terms, is devise a system where each position has a certain number of “claim points” that players can get buy making plays. The system by which these claim points are assigned for each position is complicated, poorly explained, and pretty much pre-determines what each player will get based on the position they play. This, of course, is wrong, and the system is, well, stupid.
But I do (I think) understand what James was trying to do. Or at least what he should have been trying to do. Win Shares are about absolute value, so players should be credited for the absolute number of plays they make, not how good they are compared to the average fielder at their position. Unfortunately, James ends up with a system that is a strange mishmash of the two, a system that neither accurately rates fielders, nor fully assigns absolute value. My goal in my system is solely to assign absolute value.
The way I do so is simple. Each player is assigned “plays” made, which are measured as such:
OF = PO+A
C,2B,SS,3B = A
1B = PO + A – (Other infielders A)*Innings/Team_Innings
P = PO + A – (First basemens A)*Innings/Team_Innings
The reasoning for each is simple. With OF, I’m just accounting for their absolute value, every chance they convert, they’re given credit for. With infielders, I did use some discretion: infielders’ PO have so little to do with them, that I feel it is unnecessary to give them any credit, even though I am measuring absolute value. For example, in using assists, without adjusting for pitcher or batter handedness, I’m giving players’ credit for converting their chances into outs, even if they have more chances. But, with an infield PO, almost every time, more than one player will have a chance, so why should the player who always waves everyone off get credit for something someone else would have done had he not? It just doesn’t make sense to me, to be honest. With first basemen, I’m taking James’ idea about independent putouts. In other words, we’re looking at how many plays a first basemen has made independent of the other infielders. For pitchers (for whom James doesn’t even count fielding), I’m using James’ idea again, except, since most plays for pitchers are interactions with first basemen, I use first basemen’s assists rather than infield assists.
I could expand on that last paragraph (heck, I could make a book out of it), but I don’t really want to. I think it’s all pretty simple, obvious, and self-explanatory. Now, what you’ll end up seeing is that where with James, he pretty much dictates that shortstops have to get 18% of the credit, or so, it happens naturally within my system, because historically, shortstops are going to make roughly 18% of the “plays.” This system is much less arbitrary, and, in my opinion, it makes more sense. But how do I assign fielding win shares? Simple.
I add up the team plays made, and then divide each individual player’s plays made by the team’s total plays made. Then, as I’ve already split the team’s defensive win shares between pitchers and fielders (using the same system I use to split defensive runs created between pitchers and fielders, which is where pitcher’s runs created come from), I simply multiply each player’s % of the team’s plays made by the team’s fielding win shares, and thus I have individual fielding win shares. Later today, I’ll put it all together, including releasing all my formulas, etc.
Edit: One more thing. Using this system, we can compare players with different positions fairly, unlike a +/- baseline system that needs a positional adjustment. More so, using this system, we can derive the positional adjustment for baseline systems. I’ll do so in the coming days/weeks.