SoxProspects News
|
|
|
|
Legal
Forum Ground Rules
The views expressed by the members of this Forum do not necessarily reflect the views of SoxProspects, LLC.
© 2003-2024 SoxProspects, LLC
|
|
|
|
|
Forum Home | Search | My Profile | Messages | Members | Help |
Welcome Guest. Please Login or Register.
4/30-5/2 Red Sox vs. Giants Series Thread
asm18
Veteran
Posts: 912
Member is Online
|
Post by asm18 on Apr 30, 2024 10:08:54 GMT -5
Per Alex Speier (from his and Tom Caron's new podcast "310 to Left"): Vaughn Grissom saw 61 fastballs on his Triple A rehab - he swung and missed twice.
|
|
|
Post by rhswanzey on Apr 30, 2024 11:02:26 GMT -5
ZiPS thinks O'Neill is all the way back to his 2021 self offensively, projecting 143 the rest of the way. The same sort of analysis shows large undervaluation for the other names, and via the other projection systems as well. My point is simple: the systems are more useful for the sort of change analysis SoxStats highlights than they are as reliable projections of performance. Caveat emptor. There were posts throwing those numbers around as if they were some sort of ceiling and they are not.ZiPS is showing us 50th percentile outcomes for any individual player. We don’t see the range of all the percentile outcomes, or what the highest and lowest percentile outcomes are. The projection systems are very much not designed to state “this is what player X will do” or “this is all that player X is capable of”. With a large sample of players, the projection systems have a reasonably high degree of accuracy, but within that sample, it is going to miss on lots of individual players, since it’s impossible to know exactly which ones will exceed or fall short of the system’s median expectation. To your point… we’re really flattening the game by presenting data as if it’s making claims that it’s not.
|
|
manfred
Veteran
Posts: 11,423
Member is Online
|
Post by manfred on Apr 30, 2024 11:04:44 GMT -5
Per Alex Speier (from his and Tom Caron's new podcast "310 to Left"): Vaughn Grissom saw 61 fastballs on his Triple A rehab - he swung and missed twice. When is he coming up?
|
|
|
Post by terriblehondo on Apr 30, 2024 11:20:36 GMT -5
Per Alex Speier (from his and Tom Caron's new podcast "310 to Left"): Vaughn Grissom saw 61 fastballs on his Triple A rehab - he swung and missed twice. Then I guess he is not going to see anything but breaking balls with the Sox. Let's hope he can handle the steady diet of breaking balls.
|
|
|
Post by redsox04071318champs on Apr 30, 2024 11:32:11 GMT -5
I tend to trust my own internal "projections" known as my own opinion and evaluation of a player over system projection. I'm wrong a good percentage of the time, but then again so are the "projections" from all these systems or whatever so why trust them over my own eyes and experience?
Why treat them like its gospel when it's even more wrong than I am?
|
|
|
Post by incandenza on Apr 30, 2024 11:33:44 GMT -5
Per Alex Speier (from his and Tom Caron's new podcast "310 to Left"): Vaughn Grissom saw 61 fastballs on his Triple A rehab - he swung and missed twice. When is he coming up? Supposed to be today.
|
|
|
Post by incandenza on Apr 30, 2024 11:35:33 GMT -5
Everyone who is criticizing projections is criticizing them on the grounds that they are not gospel and can't be taken as reliably predicting the future, which is something that no one who actually references the projections thinks.
|
|
|
Post by Oregon Norm on Apr 30, 2024 11:50:30 GMT -5
There were posts early on comparing projections across teams. Given the error bars associated with those - that range of projections - the comparison is not very useful. Again, change analysis would seem to be the best way to utilize these. I'm also of the notion that it's long past time for all the projection systems and the sites that tout them, such as Fangraphs, to include graphical displays to show that range of possible outcomes. Stacast leads the way here. This isn't the 1950s. There are plenty of tools for doing that.
|
|
|
Post by scottysmalls on Apr 30, 2024 11:55:01 GMT -5
Everyone who is criticizing projections is criticizing them on the grounds that they are not gospel and can't be taken as reliably predicting the future, which is something that no one who actually references the projections thinks. Yeah I’ll also add that some of the specific criticisms that the projections aren’t as good for prospects/young players and when players suddenly change their approach significantly (ex: the Red Sox pitching staff this year) are well understood and acknowledged by people who use the projections.
|
|
|
Post by incandenza on Apr 30, 2024 12:00:29 GMT -5
There were posts early on comparing projections across teams. Given the error bars associated with those - that range of projections - the comparison is not very useful. Again, change analysis would seem to be the best way to utilize these. I'm also of the notion that it's long past time for all the projection systems and the sites that tout them, such as Fangraphs, to include graphical displays to show that range of possible outcomes. Stacast leads the way here. This isn't the 1950s. There are plenty of tools for doing that. What is change analysis? How would you use projections to do it?
|
|
|
Post by jmei on Apr 30, 2024 12:30:20 GMT -5
I can't believe we're still arguing about this in the year of our lord 2024. Criticizing the projections for missing on players A, B and C is about as valuable a contribution to internet discourse as criticizing the weatherman for blowing the forecast. It shows a fundamental misunderstanding of statistics and probability that I had empathy for back in the 2002-ish Moneyball days, but now, more than 20 years later, most of which has been spent on this very forum, is just boring and played out reflexive tribalism. I don't even have it in me to try and provide a good faith explainer anymore. I get it, you're tired of the nerds being the smartest guys in the room and making you feel inadequate, so you beat your chests every time a projection system fails (which they do an overwhelming majority of the time). It's just not an interesting discussion anymore. /rant
|
|
|
Post by Oregon Norm on Apr 30, 2024 12:36:30 GMT -5
There were posts early on comparing projections across teams. Given the error bars associated with those - that range of projections - the comparison is not very useful. Again, change analysis would seem to be the best way to utilize these. I'm also of the notion that it's long past time for all the projection systems and the sites that tout them, such as Fangraphs, to include graphical displays to show that range of possible outcomes. Stacast leads the way here. This isn't the 1950s. There are plenty of tools for doing that. What is change analysis? How would you use projections to do it? See SoxStats above: By determining the change in a performance index - the increase in WRC+ in his post. As the season progresses and more data is accumulated, it re-centers the distribution and tightens up the projection. A video of that change would be a nice visual I think.
|
|
|
Post by Oregon Norm on Apr 30, 2024 12:37:45 GMT -5
I can't believe we're still arguing about this in the year of our lord 2024. Criticizing the projections for missing on players A, B and C is about as valuable a contribution to internet discourse as criticizing the weatherman for blowing the forecast. It shows a fundamental misunderstanding of statistics and probability that I had empathy for back in the 2002-ish Moneyball days, but now, more than 20 years later, most of which has been spent on this very forum, is just boring and played out reflexive tribalism. I don't even have it in me to try and provide a good faith explainer anymore. I get it, you're tired of the nerds being the smartest guys in the room and making you feel inadequate, so you beat your chests every time a projection system fails (which they do an overwhelming majority of the time). It's just not an interesting discussion anymore. /rant See above.
|
|
steveofbradenton
Veteran
Watching Spring Training, the FCL, and the Florida State League
Posts: 1,827
|
Post by steveofbradenton on Apr 30, 2024 12:39:44 GMT -5
ZiPS thinks O'Neill is all the way back to his 2021 self offensively, projecting 143 the rest of the way. Yeah, he's murdering the ball with also the best BB% and K% of his career. If the Sox think they can keep him healthy, he seems like a really good candidate to keep around beyond this year and break up our plethora of lefties.
I also would love to see Breslow get him extended for 4 years. He is perfect for Fenway with his power and defense.
|
|
asm18
Veteran
Posts: 912
Member is Online
|
Post by asm18 on Apr 30, 2024 12:41:32 GMT -5
Andrew Bailey's old team is 24th in era. His new team is 1st. Last year his old team was 11th and his new team 22nd. That is a pretty big loss for his old team and a huge gain for his new team. Glad he is ours !!! It's kinda weird because their top two starters Logan Webb and Jordan Hicks have a 2.33 and 1.59 ERA respectively. Their mid-rotation starters and the top of their bullpen (hello Rogers twins!) have been fine. But the back half of their pitching staff has been terrible - and Blake Snell had an 11.57 ERA in like 11 innings before going the injured list.
|
|
|
Post by sxfan on Apr 30, 2024 12:42:16 GMT -5
It is coming today. Interested in seeing if there's Yoshida injury news today. Not expecting him to be in the lineup and you would be down to a 3 man bench of Refsnyder, Dalbec, and Hamilton with one of the catchers at DH if Yoshida isn't on the IL. Logan Webb is one of the top 10 pitchers in the sport. You need all the weapons you can get against him. If Yoshida is hurt IL him and put Valdez at DH.
|
|
|
Post by puzzler on Apr 30, 2024 12:42:30 GMT -5
I can't believe we're still arguing about this in the year of our lord 2024. Criticizing the projections for missing on players A, B and C is about as valuable a contribution to internet discourse as criticizing the weatherman for blowing the forecast. It shows a fundamental misunderstanding of statistics and probability that I had empathy for back in the 2002-ish Moneyball days, but now, more than 20 years later, most of which has been spent on this very forum, is just boring and played out reflexive tribalism. I don't even have it in me to try and provide a good faith explainer anymore. I get it, you're tired of the nerds being the smartest guys in the room and making you feel inadequate, so you beat your chests every time a projection system fails (which they do an overwhelming majority of the time). It's just not an interesting discussion anymore. /rant "There's no justification for not spending $10-15 million extra dollars to get a couple more wins". I don't know how many times this was repeated this offseason, but it was a lot; all the while using projections from ZiPS and Steamer and everywhere else. You even had people specifically saying why they don't like such and such projection. There wouldn't be as many people resistant to learning about statistics or accepting them, if there weren't so many people using them to hammer people over their heads about how right they are. It never starts out as a criticism of the projection, it always starts out with someone misusing them and then people reduce everything to the lowest common denominator and try to pick apart the projection because it is the lowest hanging fruit to say, Jorge Soler.
|
|
steveofbradenton
Veteran
Watching Spring Training, the FCL, and the Florida State League
Posts: 1,827
|
Post by steveofbradenton on Apr 30, 2024 12:42:44 GMT -5
Per Alex Speier (from his and Tom Caron's new podcast "310 to Left"): Vaughn Grissom saw 61 fastballs on his Triple A rehab - he swung and missed twice. When is he coming up? Isn't he going to be in the line-up tonight? A whole new right-hand side of the infield with Cooper too?
|
|
|
Post by jmei on Apr 30, 2024 12:59:47 GMT -5
I can't believe we're still arguing about this in the year of our lord 2024. Criticizing the projections for missing on players A, B and C is about as valuable a contribution to internet discourse as criticizing the weatherman for blowing the forecast. It shows a fundamental misunderstanding of statistics and probability that I had empathy for back in the 2002-ish Moneyball days, but now, more than 20 years later, most of which has been spent on this very forum, is just boring and played out reflexive tribalism. I don't even have it in me to try and provide a good faith explainer anymore. I get it, you're tired of the nerds being the smartest guys in the room and making you feel inadequate, so you beat your chests every time a projection system fails (which they do an overwhelming majority of the time). It's just not an interesting discussion anymore. /rant "There's no justification for not spending $10-15 million extra dollars to get a couple more wins". I don't know how many times this was repeated this offseason, but it was a lot; all the while using projections from ZiPS and Steamer and everywhere else. You even had people specifically saying why they don't like such and such projection. There wouldn't be as many people resistant to learning about statistics or accepting them, if there weren't so many people using them to hammer people over their heads about how right they are. It never starts out as a criticism of the projection, it always starts out with someone misusing them and then people reduce everything to the lowest common denominator and try to pick apart the projection because it is the lowest hanging fruit to say, Jorge Soler. I'm totally OK with criticizing misuse of projections. But the braggadocio about projections being wrong is so misguided.
|
|
nomar
Veteran
Posts: 10,921
|
Post by nomar on Apr 30, 2024 13:01:31 GMT -5
What is change analysis? How would you use projections to do it? See SoxStats above: By determining the change in a performance index - the increase in WRC+ in his post. As the season progresses and more data is accumulated, it re-centers the distribution and tightens up the projection. A video of that change would be a nice visual I think. The fact that projections change doesn’t mean that they should be ignored either. Nothing is infallible, and all a projection gives you is one lens to look forwards through.
|
|
|
Post by Oregon Norm on Apr 30, 2024 13:23:40 GMT -5
"There's no justification for not spending $10-15 million extra dollars to get a couple more wins". I don't know how many times this was repeated this offseason, but it was a lot; all the while using projections from ZiPS and Steamer and everywhere else. You even had people specifically saying why they don't like such and such projection. There wouldn't be as many people resistant to learning about statistics or accepting them, if there weren't so many people using them to hammer people over their heads about how right they are. It never starts out as a criticism of the projection, it always starts out with someone misusing them and then people reduce everything to the lowest common denominator and try to pick apart the projection because it is the lowest hanging fruit to say, Jorge Soler. I'm totally OK with criticizing misuse of projections. But the braggadocio about projections being wrong is so misguided. Given the wildly varying estimates of those systems, it's essential that they be seen for what they are, not for what people want them to be. And there's no way to slice it or dice it: Steamer's numbers are nearly useless for proposed roster decisions. The systems using nearest neighbor analysis perform better, but they also have a dynamic that needs to be understood. That is my point.
|
|
|
Post by philsbosoxfan on Apr 30, 2024 14:05:53 GMT -5
It won't be very long before a lot of projection and analytical tools take a back seat replaced by scouting in the form of electronic tracking.
|
|
asm18
Veteran
Posts: 912
Member is Online
|
Post by asm18 on Apr 30, 2024 14:13:43 GMT -5
Promoted to Boston, or kidnapped?
|
|
asm18
Veteran
Posts: 912
Member is Online
|
Post by asm18 on Apr 30, 2024 14:33:27 GMT -5
Promoted to Boston, or kidnapped? Okay, so I guess Vaughn Grissom has indeed been kidnapped: Duran CF, Devers 3B, O’Neill DH, Abreu RF, Refsnyder LF, Valdez 2B, Cooper 1B, McGuire C, Rafaela SS
|
|
|
Post by patford on Apr 30, 2024 14:37:26 GMT -5
"There's no justification for not spending $10-15 million extra dollars to get a couple more wins". I don't know how many times this was repeated this offseason, but it was a lot; all the while using projections from ZiPS and Steamer and everywhere else. You even had people specifically saying why they don't like such and such projection. There wouldn't be as many people resistant to learning about statistics or accepting them, if there weren't so many people using them to hammer people over their heads about how right they are. It never starts out as a criticism of the projection, it always starts out with someone misusing them and then people reduce everything to the lowest common denominator and try to pick apart the projection because it is the lowest hanging fruit to say, Jorge Soler. I'm totally OK with criticizing misuse of projections. But the braggadocio about projections being wrong is so misguided. It's absolutely "a thing" that all across baseball there are fans of every team who would rather be right than see a player they disliked succeed.
|
|
|