Spencer Kerch

Belmont University Grad | Sports Analytics | CSU Applied Stats Masters Student

[Spencer Kerch @sjkerch]

Just presented my senior poster at BURS, my schools research symposium! Thanks to @totally_t_bomb for his great website and data, and @JonRothstein for the title inspiration. I'd love any feedback or questions. The project is at https://t.co/sKENjX5cv5

[Spencer Kerch @sjkerch]

@Stat_Ron 100% agree

[Spencer Kerch @sjkerch]

@benbbaldwin @_DylanGrafius @nflfastR What’s the reasoning behind this change?

[Mike Petriello @mike_petriello]

Billy Beane still rocks. I'm 100% convinced some team is going to take the wrong lessons from Milwaukee and lose 100 games trying it.

https://t.co/jesXqhqxCR

[Spencer Kerch @sjkerch]

@Matt_barlowe I take it there is no adjustment for predicted starting qb?

[Spencer Kerch @sjkerch]

@tangotiger @NateSilver538 I think if they showed how it faired over time, it would show it converged with his model as it got closer to the day. Def misleading a bit

[Spencer Kerch @sjkerch]

@ARKNDJL How does it work that Dak and Trevor have the same WAA but larger difference in WAR

[Spencer Kerch @sjkerch]

@CDonnick1 🥞 @StatsbyLopez

[Spencer Kerch @sjkerch]

Cignetti has seemingly forgot he has two timeouts in back-to-back situations. I’m struggling to understand what he’s doing.

[Spencer Kerch @sjkerch]

@sunshinevvn Facts.

[Spencer Kerch @sjkerch]

@wwfryer @Stat_Ron Is there a study or writing you can link to I’d love to read it

[Spencer Kerch @sjkerch]

@BLSMaS_NFL Fire. Makes sense controlling for qb makes it more marginal. I wonder if there needs to be a decrease in pa% as pass rate increases or if that’s just what we see.

[Spencer Kerch @sjkerch]

@AtoZ_Payton Also the trades for Micah or Simmons are a bit different than the ones for sauce or Russel which are treated as equal in this sample as well.

[Spencer Kerch @sjkerch]

@LeeSharpeNFL @SethBurn It makes it significantly better than the Mack trade and maybe even the garret trade. I know la didn’t use 2 firsts but verse is almost more valuable than a first

[Spencer Kerch @sjkerch]

@KentWeyrauch Agreed. Feel like extrapolating to a run pass split that’s never happened causes issues that are being ignored

[Spencer Kerch @sjkerch]

@Kirsten_Tanis1 @grimposting Pretty solid arm punt and gives a good deep ball receiver a shot imo 3-21 not a good spot either way

[Ron Yurko @Stat_Ron]

Two things can be true:
1) Shanahan has historically been too conservative in 4th down situations
2) Analytics twitter is overconfident in their win probability estimates
REPORTING UNCERTAINTY MATTERS https://t.co/ZRhXaTSMgh

[Spencer Kerch @sjkerch]

@startkyIeorton I think Kevin is correct in his reasoning but I always thought you should have had stronger priors on Williams over Daniel than you normally would for a 1 vs 2 overall qb. Even at the end of y1 if 5’s mean outcome is slightly higher 18 could have higher ceiling

[Brendan Sugrue @BrendanSugrue]

Caleb Williams finishes the day 30/36 for 305 yards and a TD with a completion percentage of 83.3%.

It's the highest completion percentage for a Bears quarterback ever with at least 35 pass attempts.

[Spencer Kerch @sjkerch]

@BenjaminSolak Goat

[Spencer Kerch @sjkerch]

@SarahMeyohas I think it’s more that it is a bad look for Elizabeth and as she says she can’t really benefit from all the money because of her legal situation. It seems that she was having Billy do it so that since their money is seperate they can’t lose the money. But you make a great point.