Boy, when Zub sinks his teeth into something good things come out...

Posted by zubinsum on 10/18/2011 4:14:00 PM (view original):

Posted by Trentonjoe on 10/18/2011 2:12:00 PM (view original):
Well, if you're offering.....

Lets say those are the # stats and both have hr/9# of .5.  

I assume the higher the hr/9#, the more the low OAV pitcher is better right?  (there has to be a better way to word that but you get the point right?)

I am not sure.  My think that you'd allow fewer homeruns (with a lower oavg) but each of those home runs would be worth more runs.

but if they have the same WHIP#, isn't the chance there will be a runner on base essentially equal?

I think BB/9 inflates faster than oavg in OLs.

Posted by winnetka1 on 10/18/2011 6:45:00 PM (view original):
Boy, when Zub sinks his teeth into something good things come out...

Sounds like butter coming out of the chicken Kiev at the old Russian Tea Room in New York.  Sorry, guess I'm still hungry tonight.  Reverie over now.

I believe a .260 OAV produces about 1 hit more per 9 innings than a .230. Compare that to 2 walks per 9 innings. If i had a man on second in the 9th inning of a close game, I know I would want a low OAV rather than a low BB/9 from my pitcher.

To put it another way, pitchers will sometimes give up an intentional walk. I have never heard of a pitcher giving up an intentional hit.

But Bilfert is right. When the WHIPs are the same, the lower ERC# is the better pitcher. That stat computes the HR allowed and gives a better overall prediction of how the pitcher will perform in  the sim.

If we knew both pitchers will have the WHIP, its a no-brainer.  The lower oavg pitcher will be better.  However, I am not sure they will have the same WHIP.  I think BB/9 will tick up more than oavg.

All other things being equal, BB/9 will tick up more. But stadiums and opposing lineups can affect that.

Posted by Trentonjoe on 10/13/2011 7:26:00 PM (view original):
Assuming  a neutral park with average defense, would you rather have a pitcher with low OAV or BB/9?

 I don't know if the #'s work right but let's say:

Guy A:  
WHIP 1.3
OAV .260
BB/9 2.0

Guy B:
WHIP 1.3
OAV  .230
BB/9 4

Okay, I have an answer:

The scenario that you describe is impossible.

If pitcher B's WHIP is 1.3 and his oavg is .260, his BB/9 will be about 2.6.  There is no way his BB/9 will go as low as 2.0.

Here are my projections:

Please note that while I am confident in the ERA projections based on oavg, BB/9, HR/9, I am not confident in how an oavg# or BB/9# projects.

	B	C	SimB	SimC
oav	.260	.230	.295	.267
bb/9	2.70	4.00	3.38	5.10
k/9	5.00	5.00	5.00	5.00
hr/9	0.50	0.50	1.00	1.00
hbp/9	0.27	0.27	0.27	0.27
h/9	9.12	7.77	10.78	9.40
whip	1.31	1.31	1.57	1.61
pa/9	38.87	38.90	41.16	41.58
t1b/9	10.03	10.23	11.66	12.25
k/pa	0.13	0.13	0.12	0.12
gidp/9	0.43	0.37	0.52	0.45
sb/9	0.95	0.97	1.11	1.16
cs/9	0.34	0.35	0.40	0.42
tb/9	10.91	9.52	14.11	12.69
sh/9	0.39	0.40	0.45	0.48
sf/9	0.18	0.18	0.21	0.22
ERA	3.64	3.33	5.25	5.04

I looked up players similar to those # stats earlier quickly.  Can't remember who they were, I'm pretty sure C was best matched by a Kris Benson and B by a Sam Gray.  The C guy cost about $25.6k/IP, B cost $23.4 IIRC.  Those are certainly close.  So B costs about 8.5% less to have a 4% lower ERA.  Whether or not that's a good tradeoff depends on what baselines you're using for price and performance, but there's certainly the argument that B is more cost effective.

Put a different way, if we fix the OAV# within 6 points of .260, HR/9# between .45 and .55, K/9# between 4.5 and 5.5, and $/IP between $25k and $26k 5 of the 7 results have WHIP# of 1.25 or lower.  The 2 outliers have OPSs of .649 and .689, which likely explains some of their apparently inflated prices.  The best WHIP# of the group belongs to 2003 Mark Mulder (who, incidentally, has a 0 OPS) at 1.17.  This is dramatically better than the 1.3 of Gray.  So for the same price you get a guy who gives up more hits but allows significantly fewer total baserunners.

Thanks guys...that is some interesting stuff.

Here is a simplified version of my calculation-- I call this simplified zub's era (sZERA).  In the more complex version, I use K/9 and customize an estimate of gidp/9, sb/9, cs/9, sh/9 and sf/9.  Again, the estimate of ERA based on expected h/9, hr/9, bb/9 is pretty accurate.  The estimation of h/9 hr/9, bb/9 probably only work with low oavg, low bb/9 pitchers.  

---

h/9  = h/9# * 1.1574
dp/g = 0.6900
cs/9 = 0.3757
sb/9 = 0.9850
bb/9 = bb/9# * 1.3084
hr/9 = hr/9# * 1.2827
sf/9 = 0.0422
sh/9 = 0.0922
tb/hbip = 1.21
ibb/9 = 0.255

A = (h/9 + bb/9 +hbp/9 -cs/9 -gidp/9)
B = [(h/9 - hr/9) * tb/hbip] + (4 * hr/9 +0.26) * (bb/9 - ibb/9 + hbp/9) + [0.52 (sh/9 + sf/9 +sb/9)]
C = 27 - cs/9 - gidp/9

sZERA = A*B / C