This might fall under "Small sample size" but I've ran pretty much the same staff for the last 3 seasons in Mantle.   ERA 3.57, 3.69 and 4.09 this season.    I made some defensive changes in order to try to produce more offense and, using my PPPS, I'm losing about 25-30 outs at all three positions.

The first player, highlighted, is the guy I'm using this season.    The total is the rest of the players over the last 2 seasons(plus this 40 games).

Since I'm playing .500 ball after two pretty successful seasons, the experiment may end soon.   I just can't figure out which bat I want to bench.

I didn't realize that you had picked up Joey "Boo Boo" Cox after he left the Apollos.

Yeah you did.  We talked about him last season.   Your first words were "Good luck with that injury-prone bastard" or something to that effect.

I'm old.  I forget things.

In my first 9 seasons playing HBD I've always pretty much ignored defense. Maybe it would be a good idea to change that. 

Bump for cbriese

So, if you have gone to the trouble of creating PPPS for each of you infielders, what factors would you use to arrive at a "Runs Saved" stat that correlates with the RC stat on offense?

The other worry I have: you weight putouts, assists, and +/- plays the same. But I assume that a plus play really counts twice in this formula, because on a single plus play, a lesser fielder would have recorded not only one less plus play, but also one less assist/putout. Minus plays are similarly dual-weighted. And I don't necessarily think that is a bad thing.

My concern is that often I have RF numbers between platooning players where Player A might be half a play per game better than Player B, but that the number of +/- plays do not correlate the RF.

I have in the past multiplied RF by a modified FPCT (factoring in plus plays, minus plays, and errors) as a way to compare players,  but I am not sure that is right either.

I'd figure up my three year average first.   By position.    For instance, at 2B, it's 479.3.   Prieto, at the time, was 448.9 for a difference of 30.4.   Conventional math says that's just 10 innings worth of outs.   If you give up 4 runs per game, that's just 4.1 runs per season.    I'm sure that's not right as it doesn't factor in extra pitching fatigue or the fact that's is 4 outs per inning 30.4 times per season.    So I just account it for half a run per out.   My rough estimation is Prieto would cost me about 15 runs per season.   That's probably a little high but it's just a rough number.  

FWIW, the previous two seasons, I gave up 646 and 631 runs.   At the A/S break, I'm on pace to give up 672 this year.    I'm working on the defensive numbers now so I'll come back with position by position in a bit.

My bad.  I take putouts out of the equation for infielders.

Out of curiosity, why do you remove the putouts from the equation?

It's just catching the ball.    It cuts out some line drive catches but, if you leave it in, the SS/2B numbers are greatly inflated due to DP attempts.   And the 1B would have the highest number while actually making the fewest number of plays.

But many of the plus plays are lineouts or popouts - catching the ball. There is no way to compare defense across positions, so I just leave the putouts in there and accept that a 1B will make 10-12 per game.

Then you are counting them twice.   As a putout and as a plus play.

Again, I just want a rough idea of which player is producing better results.    I started doing it because I didn't know if range or arm strength or glove or accuracy would provide better results at which position.    You could tell that Player A made less errors but had less plus plays or that Player B got to more balls but really didn't provide much else.    I'm not really into debating if my way or your way or some other way is better.    My way serves my purpose.