In the past I have found that Joe Peta has great insight on MLB futures. I am having a hard time tracking them down this year. Does anyone know where to find them?
-
-
Hes on twitter, not sure his handle, he is a good baseball guy.Comment -
He took time off to write a golf analytics book . Don’t think he’s doing anything baseball related this year
Twitter handle
@MagicRatSFComment -
One of the few guys I respect about betting MLB. Sportsbettorswin site is another solid source.Comment -
i'm part-way through reading his book............ not as much about baseball betting as i'd hoped. but good writer, good guy and wall street stuff is semi-interesting to me. i feel like he should have written 2 books.
if you want to do his win total picks yourself, this is what i remember quantitatively:
1) compare some hitting ratios to runs scored.......... find the outliers. they won't do the same (good or bad) this year. i willl try to look up what ratios he used.
2) do pythagorean on 2018 win totals....... expected win % = RS^x divided by (RS^x + RA^x).. x = 1.82.... i saw pyth numbers on baseball reference for 2018... not sure if they used 1.82 or 2.00...
somehow combine 1) and 2). they are completely seperate to start.
3) relief pitching regresses to mean.
4) player losses/additions........ obviously quite subjective or soft quant...........
i think 1 and 2 and very obvious additions/deletions will get you to a good number to compare to vegas.Comment -
probably could do #1 for pitching too. i have to think if it makes logical sense. probably doesComment -
someone posted the pecota numbers......... i think they'd have alot of similarities to peta's numbers.... i remember the cubs was way different from vegas and last year.Comment -
looks like peta uses 2010 tampa bay as his model for team whose runs don't match their offensive stats
he looks at BA, OBP, slugging, isopower, walks..... isopower and walks fill a little bit of the gap (should have said tampa was huge outlier in having too many runs i.e. they were amongst top runs team but NOT batting stats)
then i he regresses 3 of them to runs scored. forget which 3 but walks was NOT one of them....... doesn't show his regression or adjust for the 3 indy's being highly correlated.
then it turns out tampa was very clutch hitting with runners on base.... that accounted for the huge positive difference......... i will check 2011 as to betting performance for tampaComment -
+3% ROI on tampa for 2011....... not that great............ although that was only part of peta's analysis.Comment -
i looked into runs scored vs OPS by team. nothing too much out of line there.
part 2 pythagorean..........https://www.baseball-reference.com/l...tandings.shtml
too high last year:
boston 5 wins (pyth expected 5 fewer wins... i used cutoff of 5)
mil 5
miami 5
colorado 6
seattle 12
too low last year:
houston 6 (pyth exp 6 more wins)
clev 7
wash 8
balt 8
LAD 10.
nice thing is that this is very backtestable... you could do whole next season ROI, or just april/mayComment -
He used to work for baseball prospectus too.Comment
