|
Post by Jimmy on Feb 22, 2021 18:06:07 GMT -5
blogs.fangraphs.com/zips-2021-top-100-prospects/I found this super interesting, especially since it compares the ZIPS rankings (done by an algorithm / model, or whatever you want to call it) vs the “manual” rankings done by Fangraphs. Sox have 2 Top 100 guys in the manual version, but 5 in the ZIPS model. Downs 20 (53 on manual list) Casas 67 (54) Seabold 71 (Unranked) Whitlock 77 (Unranked) Mata 88 (Unranked) I think it’s fascinating that a few of Blooms acquisitions are either significantly higher when ran through the model (Downs) or crack the Top 100 without being ranked in the manual version (Seabold, Whitlock). Whitlock is particularly interesting in that he was a Rule 5 pick. Thinking maybe analyses like those done in the ZIPS projection are more in line with how Bloom evaluates talent. Or maybe it’s just one factor but guys who don’t crack the normal Top 100 but light up the ZIPS model are easier to acquire.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,911
|
Post by ericmvan on Feb 22, 2021 23:18:00 GMT -5
blogs.fangraphs.com/zips-2021-top-100-prospects/I found this super interesting, especially since it compares the ZIPS rankings (done by an algorithm / model, or whatever you want to call it) vs the “manual” rankings done by Fangraphs. Sox have 2 Top 100 guys in the manual version, but 5 in the ZIPS model. Downs 20 (53 on manual list) Casas 67 (54) Seabold 71 (Unranked) Whitlock 77 (Unranked) Mata 88 (Unranked) I think it’s fascinating that a few of Blooms acquisitions are either significantly higher when ran through the model (Downs) or crack the Top 100 without being ranked in the manual version (Seabold, Whitlock). Whitlock is particularly interesting in that he was a Rule 5 pick. Thinking maybe analyses like those done in the ZIPS projection are more in line with how Bloom evaluates talent. Or maybe it’s just one factor but guys who don’t crack the normal Top 100 but light up the ZIPS model are easier to acquire. ZiPS here is spotting guys whose performance has been ahead of their scouting reputations. Seabold is a poster child for that, and of course there are a minority of scouts who realize that plus command is a tool underrated by evaluators. But it's a fact that velo (and probably stuff) add very little to a pitch if it's thrown to a hitter's cold zone. Call that the Jamie Moyer Law. "Stuff" within the zone is for getting away with mistakes.
Whitlock's a big surprise, given that his numbers in AA weren't super-impressive.
|
|
|
Post by Jimmy on Feb 22, 2021 23:28:49 GMT -5
blogs.fangraphs.com/zips-2021-top-100-prospects/I found this super interesting, especially since it compares the ZIPS rankings (done by an algorithm / model, or whatever you want to call it) vs the “manual” rankings done by Fangraphs. Sox have 2 Top 100 guys in the manual version, but 5 in the ZIPS model. Downs 20 (53 on manual list) Casas 67 (54) Seabold 71 (Unranked) Whitlock 77 (Unranked) Mata 88 (Unranked) I think it’s fascinating that a few of Blooms acquisitions are either significantly higher when ran through the model (Downs) or crack the Top 100 without being ranked in the manual version (Seabold, Whitlock). Whitlock is particularly interesting in that he was a Rule 5 pick. Thinking maybe analyses like those done in the ZIPS projection are more in line with how Bloom evaluates talent. Or maybe it’s just one factor but guys who don’t crack the normal Top 100 but light up the ZIPS model are easier to acquire. ZiPS here is spotting guys whose performance has been ahead of their scouting reputations. Seabold is a poster child for that, and of course there are a minority of scouts who realize that plus command is a tool underrated by evaluators. But it's a fact that velo (and probably stuff) add very little to a pitch if it's thrown to a hitter's cold zone. Call that the Jamie Moyer Law. "Stuff" within the zone is for getting away with mistakes.
Whitlock's a big surprise, given that his numbers in AA weren't super-impressive.
It’s too bad that their model is proprietary (presumably?) - I’d love to see the inputs driving T100 spots for guys like Whitlock who’s stats don’t really pop off the page.
|
|