tag:blogger.com,1999:blog-12133335.post113591026987460614..comments2024-02-13T06:56:14.486-05:00Comments on Walk Like a Sabermetrician: Win Shares Walkthrough, pt. 5phttp://www.blogger.com/profile/18057215403741682609noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-12133335.post-83996593413821813612016-03-16T23:03:29.823-04:002016-03-16T23:03:29.823-04:00Do you know how to get holds from retrosheet?Do you know how to get holds from retrosheet?Anonymoushttps://www.blogger.com/profile/03164411672717968788noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-1136776443196246182006-01-08T22:14:00.000-05:002006-01-08T22:14:00.000-05:00Your point about changing the zero-line for starte...Your point about changing the zero-line for starters and relievers is well taken and something I did not consider. We know from empirical studies that the "replacement" or "freely available" or whatever you call it line for relievers is lower(in terms of ERA) then it is for starters, so it would make sense to incorporate that into WS.<BR/><BR/>The problem is that Bill's zero lines are not based on a concept of "replacement level", but rather on the mathematical consequences of his W% estimator(ie 152% equals no marginal runs allowed equals zero value). So making a change in this area would kind of violate the "spirit" of Win Shares. I think you mentioned in one of our other posts or on your blog about pondering how many changes you could make and still call it Win Shares. This would seem to be a pretty big one in that area to me--not that I am a "Win Shares purist" or anything. If it causes the system to work better, I'm all for it, whatever it is. But IMHO it would be a sgnificant departure from the WS approach.<BR/><BR/>BTW, thanks for reading/commenting/linking/etc.phttps://www.blogger.com/profile/18057215403741682609noreply@blogger.com