NERP and Catcher Defense

Deciding how many SB attempts occur at each arm level is a bit of an art. If you start at the -5 level and work up the increase is exponential and not linear. How, much of a factor can be debated, but this is one where I recommend parsing lots of data to come up with reasonable approximations. I haven't done it in a long time, and it is probably something that should be refreshed and probably improved in my own pricing model.

Nels,

The formula I was using above to arrive at my RC values (before making team adjustments for outs) was simply:

SB*.19-Outs*.0976+Throwing errors*.19

The concept of a throwing error is a level of detail that the ERP and NERP ignores, but for this purpose I assume that it looks and acts like an additional Stolen base.

I will try and process your point about NERP vs ERP every day.

The project though has carded up 100s of cards and the ATG pool. Just missing a simple-to-complex way for catcher defense, but have more of a range in mind now. All ratings are boiled down to NERP in my system and then "EversdeNERP," which is just NERP per $1mil. So I want to finish out this model and then should look at the more contextual things you lay out. I do love me a circular lineup, so definitely try and see the value of more/less "outs".

I can plug in a table for Arm/T-Rating and get those RC values. Thank you. Do RC values translate to NERP, or ERP for your model? Any estimate on how the RC you're providing would translate to NERP? On DiamondDope RC/27 (maybe toootally different) it seems pretty close to 1-1. But that's all pro-rated back to 108 PA, whereas the number of errors and then runs created you provided is to the whole season? So pro-rate that back to 108 PA to get something in line with DD "NERP"?

I hadn't read anything about Estimated Runs Created outside of Berce's article and Diamonddope...just read it from the '85 Bill James Abstract.

The formula I was using above to arrive at my RC values (before making team adjustments for outs) was simply:

SB*.19-Outs*.0976+Throwing errors*.19

The concept of a throwing error is a level of detail that the ERP and NERP ignores, but for this purpose I assume that it looks and acts like an additional Stolen base.

So I could plug in the Catcher SB attempts against at ~65% success rate to the NERP formula?
(TB * .318) + ((BB+HBP-CS-GIDP) * .333) + (H * .25) + (SB * .2) - (AB * .085)

Though, this is scaled to a player, whereas the SB against is for a whole lineup (~x9 as much as 1 player). But still, that's the "worth" of the 1 catcher, so rather than 2.84 NERP for '06 C. Crawford's 58 SB and 9 CS, a +1(T-10) catcher allowing 160 attempts at 65% rate vs a -3(T-3) allowing 80 attempts...? .... wait I'm turned around again, we need run-prevention here, not additional NERP from being a base-stealer.... ????

Should I use RC ----> NERP (per 108/216)

or

plus CS/SB numbers into NERP formula similar to Berce's doc but for DEFENSE of catchers??? idk how to think about that currently, but RC was just about there if RC can translate to NERP/108

***I made my own SB calculations, which look fairly similar to Berce's. I think an OBP multiplier improves validity of just lopping on a bunch 'o NERP for a great AA with a putrid .290 OBP. I set .740 OBPs and above to a 1.2x for their SB NERP, going down .01 of the multiplier (1.19, 1.18, 1.17 etc) for each 10 points of OBP lost. A 1.0x sits at a stellar .540 OBP, all the way to .9x for .440 OBP. From there the multiplier goes down .01 for every 5 points of OBP lost. This bottoms out at a .5x for OBPs below .245

These are the totals it was producing to fit within this kinda range (Berce shows great as negative, I'm just calling it psoitive)
c-1(-4)e0,T-1(pb-0) +4.32 dNERP
c-1(-3)e3,T-4(pb-1) +2.51 dNERP
1(-2)e5, T-6(pb-2) +0.92 dNERP
1(-1)e3, T-17(pb-1) -1.94 dNERP
c-2(-1)e2,T-6(pb-2) +0.17 dNERP
c-2(-4)e3,T-6(pb-1) +1.71 dNERP
c-2(-1)e1,T-11(pb-4) -0.07 dNERP
c-3(0)e1,T-3(pb-1) -1.65 dNERP
c-3(-3)e2,T-6(pb-6) -0.92 dNERP
c-4(+1)e5,T-7(pb-6) -6.3 dNERP

things that felt wrong that your insights "fix" --- will improve as the T-rating tax is tied to arms like you mentioned
c-1(-3)e8,T-18(pb-2) -2.09 dNERP
c-1(-2)e10,T-15(pb-7) -3.08 dNERP
c-1(-2)e13,T-15(pb-14) -4.79 dNERP

GREAT STUFF. Just looking at the numbers above, a GREAT defensive catcher can be worth +10 NERP vs. a bad defensive catcher. Point being a Benji Molina is "worth" as much as many $3.0 Mill catchers when you factor in his defense.

This thread scared me away from John Henry's +5 arm.
So I cut him to go with Don Padgett (C4(+3)e6 T8 PB3).

Some people never learn.

This is probably the best most useful thread posted to these boards since I started playing. Pelletier-level stuff and I can't wait to dig in when I have some time.

Here's what I have for basic "Berce" dNERP for Range-Errors for Catchers:

Range
E 1----- 2----- 3----- 4----- 5
0 -1.08 0.38 1.85 3.23 4.15
1 -0.88 0.58 2.03 3.41 4.32
2 -0.74 0.71 2.16 3.52 4.44
3 -0.53 0.91 2.35 3.70 4.61
4 -0.35 1.08 2.51 3.86 4.76
5 -0.15 1.28 2.70 4.04 4.94
6 0.06 1.47 2.89 4.22 5.11
7 0.20 1.60 3.01 4.34 5.23
8 0.40 1.80 3.20 4.52 5.40
9 0.54 1.93 3.32 4.64 5.52
10 0.75 2.13 3.51 4.82 5.69
11 0.96 2.33 3.70 5.00 5.86
12 1.13 2.50 3.86 5.16 6.02
13 1.34 2.70 4.05 5.34 6.19
14 1.48 2.83 4.18 5.46 6.31
15 1.69 3.03 4.37 5.64 6.48
16 1.82 3.16 4.49 5.76 6.60

*This is inverted to what I described earlier. Negative scores ADD to a catcher's overall NERP and Positive scores SUBTRACT from a catcher's overall NERP


The formula for OTHER POSITIONS dNERP is:
(TB * .318) – (DP * .333) + (OB * .25) - (AB * .085)

Infielders get the DP boost, I think catchers would (?) too

AB = 27
TB = [(chance of a one-base error + (2 * chance of a two-base error)) * 27] + (chance of a range hit * 27)
DP = 27 * 18.75% * (chance of a G1) * (1 – (chance of an error)] would it work this way? Only 1s and 2s have G1s
OB = (chance of an error * 27) + (chance of no error * chance of a range hit * 54)

***chance of range hit was the most unknown. Catcher results break down into W/S, W/G, P/P, P/F, PO, FO, G1, G2 and G3. It's hard to say what is a "range hit." It's not quite the same but W/S and W/G are weird/rare plays that are consistent across all C-ranges (5% for W/S, 15% for W/G). I'm counting the P/P and P/F as the "chance of range hits". Would anything else? For range 1s, 15% of x-rolls are this, 2s its 30%, 3s its 45%, 4s its 60% and 5s its 70%. These rolls can be bad or "hits" in that they can be passed balls based upon the pitchers WP rating-----unsure here, but the other results are more "outs" than chances for hits? ***I'm taking 60% of the %s mentioned above because it felt high otherwise and often the Pitchers WP ratings will bail the catcher out? Unsure here

So a potential ....
-range/error rating like this
-throwing rating (runs created to NERP?)
-pb rating

and that's Catcher dNERP?

This thread scared me away from John Henry's +5 arm.
So I cut him to go with Don Padgett (C4(+3)e6 T8 PB3). Some people never learn.

Ha. I never get the nerve to use Padgett but after this thread...maybe I still won't...

This is probably the best most useful thread posted to these boards since I started playing. Pelletier-level stuff and I can't wait to dig in when I have some time.

Agreed--well done