Quote:
Originally Posted by David Watts
Having a blast, but I do think the Negro League pitchers need to be toned down some as a whole. Brown, McDonald, Mendez are sure fire hall of famers. Hilton Smith is well on his way to the Hall. Lankford, would be a sure thing, but his career may be too short to qualify. Then again, I'm by far not a historian, so maybe these guys really would dominate. Question, is recalc run on these players, or are they operating on a more development engine basis?
Oh and Rube Waddell whiffed 393 batters in 1967. He's on pace for 365 in 1968.
Again, having a blast.
|
So, the MLEs that were the basis for these guys (done by Eric Chalek) had a ton of work put into them to contextualize things like WAA ... that OOTP doesn't directly care about. In e-mail exchanges with Garlon, Eric expressed more confidence in the WAA analysis and resulting ERAs and such than the traditional stats. In fact, he latest version of the MLEs did not publish any of the traditional stats like HR/BB/K that OOTP needs. I'm reworking the MLEs for a revised mod as we speak and while still in a draft form, my model is far more focused on adding context to these traditional stats that feed OOTP while ensuring the results are consistent with the WAA, WAR and other more analytical metrics derived by Eric. So, the next version will be done much less in the abstract and much more with the OOTP-specific translation in mind.
As to recalc, they recalc in RD until they run out of MLE stats and then the development engine takes over with RAH off. Without calling up the MLE files themselves and just looking at my dashboard in Excel for ease, Eric's Lankford had a career K/BB ratio of 1.51 (compared to my initial redo of 1.18). Eric's K/9 rate looks like a spreadsheet error as it is 5.21 which is, for the era, very high. The engine gets a hold of that, TCR and age don't knock it down... well, there ya go.
As for Mendez, Eric had a career K/BB of 1.83 and mine came in at 1.37 for his career - still great but not obscene. Webster McDonald was at 1.41 career K/BB with Eric's calculations and my new one looks to be 1.21. For Dave Brown, whose career was cut short by legal issues, Eric had him at a career K/BB of 1.30 but mine only came in at 1.09. As for Hilton Smith, he is a legit HOF'er in real life so I had no issue when my version clocked at 1.34 K/BB versus a career 1.18 from Eric. They are also compilers since Eric had them debut early (as they did) and stick around forever (because they did).
The interesting thing is after pushing 291,866 IP of MLEs through the new model, it has their combined aggregated K/BB at 1.08 .... which is nearly exactly what Eric's is (1.06) for the same group of players. My K/9 is 3.07 and his 3.04 and my BB/9 is 2.84 and his is 2.87. Mine just distribute differently and (so far) imho, better.