tag:blogger.com,1999:blog-12133335.post6741194254488703159..comments2016-10-21T17:00:54.918-04:00Comments on Walk Like a Sabermetrician: Meanderingsphttp://www.blogger.com/profile/18057215403741682609noreply@blogger.comBlogger8125tag:blogger.com,1999:blog-12133335.post-45609462671731929432008-09-21T23:27:00.000-04:002008-09-21T23:27:00.000-04:00I too am a fan of Choo. As I'm sure the Mariner f...I too am a fan of Choo. As I'm sure the Mariner fans will never forget, in 200 two seperate trades sent the Indians' first base platoon of Eduardo Perez and Ben Broussard to Seattle for Shin-Soo Choo and Asdrubal Cabrera. Good one, Bavasi.phttp://www.blogger.com/profile/18057215403741682609noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-66202628023755631772008-09-21T15:16:00.000-04:002008-09-21T15:16:00.000-04:00Speaking of suprise performances for the Indians t...Speaking of suprise performances for the Indians this year, how about Shin-Soo Choo. Talk about flying under the radar. In 88 games he's hitting .310/.400/.559. I had always thought that Choo would be a good player if given the opportunity to play on a regular basis.terpsfan101noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-2537173413652964922008-09-16T17:50:00.000-04:002008-09-16T17:50:00.000-04:00Statistics can be deceptive...and win-loss record ...Statistics can be deceptive...and win-loss record is exempt from this? <BR/><BR/>I never meant to imply that Cliff Lee was a bad pitcher. But I stand by Lee as one of the top pitchers in the league being a one year wonder, at least to this point in his career.<BR/><BR/>Cliff Lee has exceeded the W-L record that you would expect from his runs allowed rate and his run support by about five games over the course of his career, about 1 win/32 starts. Nothing out of the ordinary. Nothing to suggest that his pedestrian runs allowed rates and peripherals give a misleading estimate of his contribution.phttp://www.blogger.com/profile/18057215403741682609noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-15212165577212341212008-09-16T17:17:00.000-04:002008-09-16T17:17:00.000-04:00It just goes to show how deceptive statistics can ...It just goes to show how deceptive statistics can be. <BR/><BR/>If you look at career stats among active pitchers with more than 100 decisions, there's that name "Cliff Lee" again. He has a career record of 76 - 38. His .667 career win percentage ranks him fourth behind Pedro Martinez, Johann Santana and Roy Oswalt.<BR/><BR/>How silly of me not to realize that with your advanced mathematical formulas you can dismiss him as a "one year wonder." Go figure.Roberthttp://www.blogger.com/profile/10824502572268749944noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-89424238074445246022008-09-16T10:09:00.000-04:002008-09-16T10:09:00.000-04:00Sorry, I reversed the sign in the R+/PA formula. I...Sorry, I reversed the sign in the R+/PA formula. It should be <BR/><BR/>R+ = RC - (O/PA - Lg(O/PA))*Lg(R/O)<BR/><BR/>If you make outs at a rate less than the league average, your contribution goes up. Duh, p.phttp://www.blogger.com/profile/18057215403741682609noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-87190805568119428302008-09-16T10:07:00.000-04:002008-09-16T10:07:00.000-04:00Robert, you mean the Cliff Lee who was fourth in t...Robert, you mean the Cliff Lee who was fourth in the league with a whopping 6.46 run support and who finished 19th in Runs Above Replacement, but got Cy Young votes because of his shiny W-L record? I remember him too, and he certainly doesn't make this Lee season any less out of line with his career.<BR/><BR/>Terps, R/O and R/PA are the straightforward versions you could expect--RC/Out and RC/PA, with RC of course being the run estimator of your choice (as long as it's an absolute out value (-.1 type) version).<BR/><BR/>R+/PA was posted at FanHome several years back by a poster named Sibelius. What it does is add the extra runs generated indirectly by the batter as a result of creating more PAs for his teammates to his regular RC. The formula for R+ is:<BR/><BR/>R+ = RC + (O/PA - Lg(O/PA))*Lg(R/O)<BR/>or = RC + (OBA - LgOBA)*Lg(R/O) if you are only considering outs = AB - H<BR/><BR/>Then R+/PA is just that divided by PA. If you use this to find runs above average (R+/PA - Lg(R/PA))*PA, you will get the exact same result as if you used (R/O - Lg(R/O))*O. The rank of players in R+/PA and R/O will not match exactly, but they will be very close.<BR/><BR/>R+/O+ is something that David Smyth came up with, but it really only applies with a Theoretical Team estimator. Basically, the "R+" is the runs the team expects to score as a result of adding the player (which includes the effect of extra PA created), and the "O+" is his PA times the new out rate/PA of his team. It's not really worth worrying about.phttp://www.blogger.com/profile/18057215403741682609noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-22448512812474068212008-09-16T02:38:00.000-04:002008-09-16T02:38:00.000-04:00Patriot,I have seen you express R/O in 2 ways and ...Patriot,<BR/><BR/>I have seen you express R/O in 2 ways and R/PA in 2 ways. I wondered if you could explain the difference between:<BR/><BR/>R/O<BR/>R+/O<BR/><BR/>and<BR/><BR/>R/PA<BR/>R+/PAterpsfan101noreply@blogger.comtag:blogger.com,1999:blog-12133335.post-61470656157959276282008-09-16T01:33:00.000-04:002008-09-16T01:33:00.000-04:00It is refreshing to learn Cliff Lee is a "one year...It is refreshing to learn Cliff Lee is a "one year wonder" with his astonishing record of 22 - 2.<BR/><BR/>Speaking of wonder, I wonder who the Cliff Lee was who was fourth in the Cy Young voting three years ago with a record of 18 - 5.<BR/><BR/>Or was that other Cliff Lee also a "one year wonder."Roberthttp://www.blogger.com/profile/10824502572268749944noreply@blogger.com