2017/18 Hot Stove Thread

[a700hitter]

WARNING - Mini stat lesson below:

 

Many people associate the word 'regression' with negative direction. The reality is that regression can be negative or positive (bad or good), which is why I often use the term 'positive regression', to avoid such confusion, even though the word 'regression' by itself is sufficient.

 

Regression, or regression to the mean, basically is the tendency of data to be closer to the average after an above or below average measurement. So, if a player has a career type year, he is likely to drop off the following year, or regress. Likewise, if a player has a sub-par year, he is likely to have a better year the following year, or regress.

 

I think several of our hitters will regress this season, meaning that they will improve over what they did last season.

The problem with this is that players are not data. They are human. Sometimes player performance just falls off because of something as commonplace and not noteworthy as wear and tear. Their performance will not regress in a positive manner to meet the statistical mean. The application of the statistical notion of positive regression to player performance is nonsense.

[moonslav59]

The problem with this is that players are not data. They are human. Sometimes player performance just falls off because of something as commonplace and not noteworthy as wear and tear. Their performance will not regress in a positive manner to meet the statistical mean. The application of the statistical notion of positive regression to player performance is nonsense.

 

Certainly many players in the history of MLB, especially those in pre-prime, early prime or mid prime, bounce back from a "down year or two" and go on to have "normal" or "better" seasons.

 

Sure, the reason the dipped at an unexpected time on the age/production curve (like so many Sox players last year) could be "wear and tear", but it could also be non-severe, injury related, personal issues or many other reasons.

 

Maybe we're arguing semantics, but I, for one, am expecting more upticks than down ticks or leveling out from these young Sox players:

 

Betts

Bogey

JBJ

Beni

 

I'm expecting 3 out of 4 to have a significantly higher WAR next year.

 

Call it age curve expected, call it random, call it "positive regression" or call it whatever you want, but I'm thinking it happens for the Sox this year.

 

Hell, we may even see upticks from HRam, Moreland and even Pedey, when he returns.

[a700hitter]

Certainly many players in the history of MLB, especially those in pre-prime, early prime or mid prime, bounce back from a "down year or two" and go on to have "normal" or "better" seasons.

 

Sure, the reason the dipped at an unexpected time on the age/production curve (like so many Sox players last year) could be "wear and tear", but it could also be non-severe, injury related, personal issues or many other reasons.

 

Maybe we're arguing semantics, but I, for one, am expecting more upticks than down ticks or leveling out from these young Sox players:

 

Betts

Bogey

JBJ

Beni

 

I'm expecting 3 out of 4 to have a significantly higher WAR next year.

 

Call it age curve expected, call it random, call it "positive regression" or call it whatever you want, but I'm thinking it happens for the Sox this year.

 

Hell, we may even see upticks from HRam, Moreland and even Pedey, when he returns.

I also expect upticks from Bogey and Betts. I expect around the same production from Bradley in his usual very uneven fashion. Beni and Devers are wild cards, IMO. I am not sure what to expect from them. They could have break out years or struggle to make adjustments. I expect a downtick for Vasquez and Moreland. Overall, I see more upticks than down, but not enough to significantly raise our offensive production.

[Bellhorn04]

It's all guesswork.

[oldtimer]

The problem with this is that players are not data. They are human. Sometimes player performance just falls off because of something as commonplace and not noteworthy as wear and tear. Their performance will not regress in a positive manner to meet the statistical mean. The application of the statistical notion of positive regression to player performance is nonsense.

 

In any short term period, that makes sense, however on a long term basis i would expect that statistical information to mirror what the player will do. Age and injury do play a part with long term performance.

[Bellhorn04]

Statistical information can only tell us what happened in the past. With players who have been up and down, you can plug their data into models and come up with projections that are more valid than someone throwing darts at a board. But that's about it.

[cp176]

Statistical information can only tell us what happened in the past. With players who have been up and down, you can plug their data into models and come up with projections that are more valid than someone throwing darts at a board. But that's about it.

 

So I could probably just say that based on their past averages it is possible and maybe even likely that guys like Bogaerts and Betts should have better years obviously depending on their health or do I have to use the new language like positive regression and stuff like that?

[moonslav59]

I also expect upticks from Bogey and Betts. I expect around the same production from Bradley in his usual very uneven fashion. Beni and Devers are wild cards, IMO. I am not sure what to expect from them. They could have break out years or struggle to make adjustments. I expect a downtick for Vasquez and Moreland. Overall, I see more upticks than down, but not enough to significantly raise our offensive production.

 

So, couldn't expected "upticks" that would bring several players closer to their prior 2-3 year "norms" be viewed as a return to the expectations or "positive regression"?

[mvp 78]

It's all guesswork.

 

Yup, even if you have mathematical theory behind your belief, it's still a big unknown.

[a700hitter]

So, couldn't expected "upticks" that would bring several players closer to their prior 2-3 year "norms" be viewed as a return to the expectations or "positive regression"?

No, because players are not instruments of statistical probability.

[mvp 78]

No, because players are not instruments of statistical probability.

 

The same goes for posters. Statistically A700 is probably the best and most read poster on TS, but we all really know it's Kimmi.

[a700hitter]

In any short term period, that makes sense, however on a long term basis i would expect that statistical information to mirror what the player will do. Age and injury do play a part with long term performance.

I think it is the opposite. Within a season hit and cold streaks tend to even out, but over the long term players can be on an upward or downward trajectory. When they start on a downward trajectory positive regression not a good bet.

[a700hitter]

The same goes for posters. Statistically A700 is probably the best and most read poster on TS, but we all really know it's Kimmi.

My posts per day have been down dramatically over the last year, so we know what that means— a posting blitz by A700Hitter is about to get under way. #Positive Regression

[moonslav59]

Statistical information can only tell us what happened in the past. With players who have been up and down, you can plug their data into models and come up with projections that are more valid than someone throwing darts at a board. But that's about it.

 

It's probably much better than throwing a dart. Age progression and decline charts are real. Of course, several players bust out of the norm and have great beginning or ends to their careers, but most follow pretty close to the norm with a few blips here and there- possibly related to injuries.

 

This is one reason I felt it was very odd that 9 out of our top 10 returning players from 2016 saw declines in their OPS- some very significant ones. Only Vaz saw an uptick. Since most of our players were pre-prime or prime, and the others were not far beyond peak prime, I felt it was very strange. Yes, the loss of Papi and nagging injuries contributed to some declines, but I'm expecting upticks for a solid majority of our returning players this year.

 

2016 OPS> 2017 OPS (Listed in order or most PAs in 2016 (110+ PAs)/ Player & age

 

.897> .803 Betts 23

.802> .746 Bogey 23

.825> .760 Pedey 32

.835> .726 J.B.Jr 26

.866> .750 HRam 32

.705> .548 BHolt 28

.845> .644 SLeon 27

.850> .709 Young 32

.585> .735 Vazqz 25

.835> .776 ABeni 21

 

Amount of decline/ Player/ Age in 2017

 

201 Leon 28

157 Holt 29

141 Young 33

 

116 HRam 33

109 JBJ 27

94 Betts 24

 

65 Pedey 33

59 Beni 22

56 Bogey 24

 

+150 Vaz 26

 

Nobody was over 33 to start the 2017 season. All but Vaz declined.

 

Okay, throw out the top 3 and Vaz from this list for being reserve (platoon) players, but we still see 6 of 6 declines from players between 22 and 34, including 3 by 94 or more points!

 

All six declined by more than 55 points. It wasn't even close for our 6 best returning players. Add Moreland to the next list and possibly Vaz (if he increases his PAs this season) and we have 8 core players plus Devers (.819 in 2017).

 

I'm pretty certain we see upticks in 5 of those 8 players. I expect Vaz to decline and maybe Pedey. Certainly HRam and Moreland can do worse or the same, but I would bet on all 4 of these guys declining next year. (I actually think HRam will top .790 which is already 40 points higher than 2017.)

 

Of course, the odds are at least one from the other 5 declines, but I'm betting 5 or more improve their OPS. (I'm not counting Devers.)

 

 

[moonslav59]

No, because players are not instruments of statistical probability.

 

Who is saying they are "instruments?"

 

Most people reach their primes after 24 or so. It's not dehumanizing people to expect improvement as they gain experience, confidence and strength and then decline as their bodies begin to break down.

[jacksonianmarch]

And we are talking about a return to a career season. That’s not realistic IMO

[mvp 78]

My posts per day have been down dramatically over the last year, so we know what that means— a posting blitz by A700Hitter is about to get under way. #Positive Regression

 

Watch out. Moon may pass you in the summer of 2029!

[mvp 78]

And we are talking about a return to a career season. That’s not realistic IMO

 

For A700? It's possible.

[jacksonianmarch]

I know people are assuming Mookie is a sacred beast and he’s damn good, but expecting a “regression” to MVP levels is cherry picking. Years 1 and 3 have almost identical WARs. Year 2 was an MVP season. How do you know that’s his mean?

 

JBJ has had near full season WARs of 0.5, 5 and 2.3 with a half season of 2.5 sprinkled in. Where’s he “regressing” to? You have no idea who he is.

 

Bogaerts has had 4 seasons in the bigs with WARs of 0.4, 4.4, 4.9 and 3.2. His mythical power has totaled 50HR the past 4 seasons. Where’s he “regressing” to? The idea of these guys returning to career seasons is cherry picking.

[mvp 78]

I know people are assuming Mookie is a sacred beast and he’s damn good, but expecting a “regression” to MVP levels is cherry picking. Years 1 and 3 have almost identical WARs. Year 2 was an MVP season. How do you know that’s his mean?

 

JBJ has had near full season WARs of 0.5, 5 and 2.3 with a half season of 2.5 sprinkled in. Where’s he “regressing” to? You have no idea who he is.

 

Bogaerts has had 4 seasons in the bigs with WARs of 0.4, 4.4, 4.9 and 3.2. His mythical power has totaled 50HR the past 4 seasons. Where’s he “regressing” to? The idea of these guys returning to career seasons is cherry picking.

 

2015 OWAR 23.5

2016 OWAR 41.9

2017 OWAR 15.8

 

I think it's reasonable to think that maybe 2017 was just a down year. Since the whole offense struggled, it could have pulled Betts down a little. I don't believe the MVP season is the "mean", but it may be somewhere between 2016 and 2017.

[moonslav59]

And we are talking about a return to a career season. That’s not realistic IMO

 

No, just improve over 2017.

 

Many declined over 2016 numbers as well:

 

.832 JBJ

.820 Betts

.797 Pedey

.776 Bogey

.727 Holt

 

HRam was at .817 in 2014.

 

I'm not asking for career bests, but actually, we should expect one of two from our 5 or 6 players between 21 and 28 to have their best seasons (or close to it).

[moonslav59]

I know people are assuming Mookie is a sacred beast and he’s damn good, but expecting a “regression” to MVP levels is cherry picking. Years 1 and 3 have almost identical WARs. Year 2 was an MVP season. How do you know that’s his mean?

 

JBJ has had near full season WARs of 0.5, 5 and 2.3 with a half season of 2.5 sprinkled in. Where’s he “regressing” to? You have no idea who he is.

 

Bogaerts has had 4 seasons in the bigs with WARs of 0.4, 4.4, 4.9 and 3.2. His mythical power has totaled 50HR the past 4 seasons. Where’s he “regressing” to? The idea of these guys returning to career seasons is cherry picking.

 

Counting rookie seasons as part of the "norm" is not realistic.

 

Betts, Bogey and JBJ all had worse seasons than 2016 or 2017. All are just entering prime years.

 

Pedey is at the end of prime. Expecting a worse 2017 than 2015 and 2016 in not unrealistic, but he's still young enough to have an uptick here and there, assuming good health.

 

HRam has traditionally been up and down. It's not unrealistic to think he could hit .850 in 2018.

 

Yes, JBJ, Beni and Devers are too hard to identify what their "norm" is, so it's even harder to expect a "regression" to something so nebulous.

[mvp 78]

@alexspeier

 

With today’s deadline to exchange arbitration figures, it’ll be a pretty chaotic day for the Red Sox, who have 10 unsigned arb-eligible players, including Betts, Bogaerts, Bradley, Pomeranz.

[Carpin]

2015 OWAR 23.5

2016 OWAR 41.9

2017 OWAR 15.8

 

I think it's reasonable to think that maybe 2017 was just a down year. Since the whole offense struggled, it could have pulled Betts down a little. I don't believe the MVP season is the "mean", but it may be somewhere between 2016 and 2017.

 

I agree, his numbers should be somewhere between 2015 & 2016, I think the 2017 is "out of character" for him, still to early to call a "career mean" with only three full years in the majors. The guy has talent, I wouldn't be surprised if he pulls a "2016 type of season" again, but chances are he will be closer to the numbers he had in 2015.

[oldtimer]

Counting rookie seasons as part of the "norm" is not realistic.

 

Betts, Bogey and JBJ all had worse seasons than 2016 or 2017. All are just entering prime years.

 

Pedey is at the end of prime. Expecting a worse 2017 than 2015 and 2016 in not unrealistic, but he's still young enough to have an uptick here and there, assuming good health.

 

HRam has traditionally been up and down. It's not unrealistic to think he could hit .850 in 2018.

 

Yes, JBJ, Beni and Devers are too hard to identify what their "norm" is, so it's even harder to expect a "regression" to something so nebulous.

 

We do have a change in conditions for these players as well with a new manager and hitting coach. It remains to be seen what if any effect that will have but I am guessing it will be positive.

[a700hitter]

Who is saying they are "instruments?"

 

Most people reach their primes after 24 or so. It's not dehumanizing people to expect improvement as they gain experience, confidence and strength and then decline as their bodies begin to break down.

True, but it doesn’t follow any rules of statistical science. That’s all that I am saying.

[a700hitter]

I know people are assuming Mookie is a sacred beast and he’s damn good, but expecting a “regression” to MVP levels is cherry picking. Years 1 and 3 have almost identical WARs. Year 2 was an MVP season. How do you know that’s his mean?

 

JBJ has had near full season WARs of 0.5, 5 and 2.3 with a half season of 2.5 sprinkled in. Where’s he “regressing” to? You have no idea who he is.

 

Bogaerts has had 4 seasons in the bigs with WARs of 0.4, 4.4, 4.9 and 3.2. His mythical power has totaled 50HR the past 4 seasons. Where’s he “regressing” to? The idea of these guys returning to career seasons is cherry picking.

It is extremely crate for a player to win more than one MVP or have more than one MVP type season. Players that can do that have a good shot at having hall of fame careers. I think Mookie is an excellent player, but I am not sure yet that he is HOF caliber. It isn’t reasonable to expect him to build on his 2016 performance. That could be his career season.

[baseball jones]

I know people are assuming Mookie is a sacred beast and he’s damn good, but expecting a “regression” to MVP levels is cherry picking. Years 1 and 3 have almost identical WARs. Year 2 was an MVP season. How do you know that’s his mean?

 

JBJ has had near full season WARs of 0.5, 5 and 2.3 with a half season of 2.5 sprinkled in. Where’s he “regressing” to? You have no idea who he is.

 

Bogaerts has had 4 seasons in the bigs with WARs of 0.4, 4.4, 4.9 and 3.2. His mythical power has totaled 50HR the past 4 seasons. Where’s he “regressing” to? The idea of these guys returning to career seasons is cherry picking.

 

its not regression with young guys.its busting out.they have taken their lumps and now they understand a lot of things they didnt before.the great things we have seen from these young guys IS what they are.now they will make it happen or "bust out"

 

thats the best part of having such a young team,we havent seen their best yet!!

[moonslav59]

True, but it doesn’t follow any rules of statistical science. That’s all that I am saying.

 

I can agree with that, but projections made with the age curve as part of projections will likely be way more accurate.

[jacksonianmarch]

its not regression with young guys.its busting out.they have taken their lumps and now they understand a lot of things they didnt before.the great things we have seen from these young guys IS what they are.now they will make it happen or "bust out"

 

thats the best part of having such a young team,we havent seen their best yet!!

 

So expecting Mookie, JBJ, and Bogaerts to improve is just "busting out" but expecting Severino, Judge and Sanchez to improve is sheer homerism?