$8.08 4R Kevin Millwood - Bad Value???

I had a post about % vs Lefy hitters and Righty hitters for pitcher. Rather than just one number, people proposed different numbers for LHP and RHP. Makes sense, but the difference is pretty wide. (LHP = 66% vs R, RHP = 44% vs R)

All of this investigation towards splits % started when my pricing engine looked at '99 Kevin Millwood. He's just a bad value my old 50%/50% splits said. Very overpriced and likely an outlier with a z-score of over 2.5. Now with only seeing his favored Righties as a RHP, he's even more overpriced. Thoughts? Any experience with him?

**Basically the new changes hurts guys who are dominant against their side (LHP vs L, RHP vs R like Millwood). For what it's worth, the overall pool seems to fit better with this change, and similar pitchers fare much better than Millwood. But like all SOM cards, he's at a critical price point where his dominance against RHP may be undervalued in such a crude NERP mashup by his "bad" vs L numbers and weak hold, def, etc. My NERP ratings average NERP vs L/R, Pitcher def, Hold, WP/BK, # HR prone and an "uneven" category which dings Millwood for mild (?) 0.38 NERP. That's all in "other NERP" listed below.

10-10-10-10 Veterans '75 DiamondDope #s
Overprice % at 50%/50% vsL/vsR = 18.6% -----this is all built off EversdeNERP, or expected NERP per $mil
Overprice % at 56%/44% vs L/vsR = 21.7%

vsL = .156/.188/.334 --- BR = -7.26 --- NERP = 7.59
VsR = .054/.098/.118 --- BR = -18.04 --- NERP = -1.823

vsL/vsR NERP = 3.45 (should be like 1.5-2.0 for $8mil)
other NERP = 0.73
Overall NERP = 4.18

So, thoughts on Millwood and the new "harsher" % vs L/R??? Baically what are people's opinion on the old -5 BR/-15 BR guy? Danny Darwin is a $7mil 4R who looks terrible but I've always considered him a BAD value. Millwood is interesting with that blistering -18.04 BR vs R and seemingly tolerable/good vs L !!!! '04 Randy is still dominant at -4.1% overpriced, but alot of the pricey R ~4R types are bad values. Is Millwood bad or does his profile just stretch the formula?

He inspired the reassessment and this is where he stands after it.

The formula just really dislikes 4R/5R Right-Handed Pitchers. Frankly, just uneven pitchers. Even at a lessening of the platoon advantage % from 34%/56% to 36%/54%, RHP that kill Righties are just bad "NERP values":

$2.3+mil 4R/5R right-handed SP* = 53 in total, averaging 9.5% overpriced.

6L/6R or more extreme = 28, averaging 8.5% overpriced
6L/6R or more extreme-reverse SP* = 14, averaging 10.8% overpriced.

The 42 3R right-handed SP* only averaged 2.7% overpriced.

*The 14 reverse 6L/6R Reverse SP* DOES include $2.23 Ray Sadecki, and his -0.4% overpriced rate. He wouldn't be included in the other $2.3 mil pools in these quick #s. Without him it goes from 11.7% overprice to 10.8% overpriced

Good read... thanks

One of my favs is the 3R Mike Scott.
Any thoughts on him?
i can live with the HR's since most of them come as solo shots.

I don't follow all these numbers, but Millwood looks pretty good to me.

Good read... thanks

One of my favs is the 3R Mike Scott.
Any thoughts on him?
i can live with the HR's since most of them come as solo shots.

I love that card. But the +7 hold rating is a killer!

Good read... thanks
One of my favs is the 3R Mike Scott.
Any thoughts on him?
i can live with the HR's since most of them come as solo shots.

Hey visick,

Scott is a SP* I don't know if I've ever used. That +7 hold terrifies me but that's potentially foolish dogma. I hope that hasn't worked it's way into my NERP ratings too much, but that along with his being a R-handed 3R/4R like Millwood and many others, gives him an uphill climb.

NERP Values - Assuming 46% of PAs vs R (Negative is good for P)
BR Bal: -14.43 BR (.104/.120/.132)
NERP Bal: 1.24 NERP
fNERP (2e11): 0.20
END NERP: -0.79 ---> the better the pitcher, the more boost/tax for END. $11mil Nola is 0.58 whereas Ed Walsh is -1.24. Crappy S6*s get about 0.3 tax
Ks NERP: -0.65 ----> this seems low, but uses a weird curve to moderate outliers' boost.
Hold/BK/WP NERP (+7hold): 0.784
SP Bal NERP: 0.205
BP # Tax: 0.284

"Total" NERP: 1.27

This scores him at 8.3% overpriced, which as commented above is a fate too many Righty-dominating Righties face. Still, at $8.31, my guy '84 Gooden scores -6.3%, so either my system shares my biases or there is hope for this type. That 8.3% overpriced means he "should" be priced at about $8.19 compared to his $8.93, whereas '84 Gooden should be $8.84 compared to his $8.31.

Idk, it's a bit off as Scott's Balanced BR is -14.43 whereas '84 Gooden's is -13.94. So really close. Gooden's Bal NERP from Diamond Dope is ~0.16 NERP better than Scotts, so there's that.... Plus, Scott getting dinged for 0.784 Hold/BK/WP is brutal. ---> the system does has dimishing rewards/punishment for SP with Holds, which helps Scott a bit:

% off of 100% for each +/- hold from 0 (because catcher's can counteract Pitcher's hold)
+/-1 = 3.1%
+/-2 = 6.2%
+/-3 = 9.3%
+/-4 = 12.4%
+/-5 = 15.5%
+/-6 = 18.6%
+/-7 = 21.7% ----- This is Scott. From just hold, if it was a flat rate per hold, he'd be 0.931 penalized, but is instead 0.756. Still maybe too puntative.

With no Hold/BK/WP factored in, he'd still be seen as about 1.8% overpriced.

Fantastic discussion.
nels--are you saying there is some kind of boost for strikeouts?

If so, what is the rationale for not lumping strikeouts as an identical category as popouts, lineouts, etc? Note some non-strikeout pitchers have just as many non-productive outs as strikeout pitchers.

Of course, there is also the effect of deep vs shallow fly balls, ground ball distribution etc, but I'd guess those are roundoff error in the final tally.

Fantastic discussion.
nels--are you saying there is some kind of boost for strikeouts?

If so, what is the rationale for not lumping strikeouts as an identical category as popouts, lineouts, etc? Note some non-strikeout pitchers have just as many non-productive outs as strikeout pitchers.

Of course, there is also the effect of deep vs shallow fly balls, ground ball distribution etc, but I'd guess those are roundoff error in the final tally.

Thanks for the comment and pointing out such important stuff. Yeah that's a huge methodology flaw. I have dug in to hitters more and certainly first. Ported a lot of those ideas to pitchers and have carded fewer, so not too late I suppose.

I have an Excel program with conditional formatting for OBPs like SI, 2B, 3B, HR, BB or HBP to show up as red and log the numbers per 108 PAs like DiamondDope. It has inputs for defense, speed, clutch etc. Pitcher scores too. I've always been aware of gbBs for guys like M. Wills and Ichiro being unaccounted for. I have box for Ks and gbAs. gbAs go directly into NERP ala Diamonddope and Ks is a separate field like speed of Hold etc.

So no, nothing for popouts, gbBs, flyBs etc. Thanks for re-bringing this to my attention. I'll see if I can or want this level of detail. I've got some 1,400 new hitters priced and 700 pitchers. I think the flaw is more pronounced in pitchers so that'd be a bit easier to rectify but still a lot of re-backtracking. If anyone has a data-set of this for the ATG pool, I'd totally get started! *Diamonddope Pool has no such values* ---> Diamonddope also doesn't seem to load on Chrome anymore, but does on Firefox. I am so grateful for Adrian and Diamonddope. Favorite website of All-Time.