Hanley undergoes left shoulder surgery

nvalvo

Member
SoSH Member
Jul 16, 2005
21,484
Rogers Park
Looking at that aging curve, though, Hanley's decline from age 32 to 33 is already an outlier in its severity (about 26 batting runs according to FG; I don't want to bring defense into it because of the apples/oranges issues raised by the 1B-to-DH move). He was only supposed to have declined about 2-3 runs this year. If he actually follows the aggregate trend from age 32 to 34 (about a 6-run drop), that will mean a substantial improvement next year, and a performance that's much closer to his 2016 than his 2017.

In other words, projecting him to be much better next year doesn't require us to assume he's a freak of nature.
Your post makes me realize that we actually need more information. Is that 6-run drop from 32-34 around the amplitude we might reasonably expect to see in the career of a given player, or is it a statistical artifact of the fact that more players are declining than improving? If player A drops by 24 batting runs between 32-33, but players B, C, D, and E stay flat, that would produce the same average as if they lost 4, 5, 6, 7, and 8 runs. (Check my math.)

I guess I would want to know how many players gradually decline versus completely implode before I try to apply these curves to individual cases.

But I don't think this gets Kielty off the hook.
 

Kielty's Last Pitch

New Member
Oct 6, 2017
118
We don't have to reinvent this wheel.


This chart is based on year-to-year pairs: it's an aggregate of about 21,500 player-season-pairs between 1950 and 2008. That is, this chart contains, say, Mo Vaughn's slight improvement from his age 23 to age 24 seasons, his breakout between ages 24 and 25, etc., up to his collapse between ages 34 and 35. All of that is aggregated with everybody else, and this curve is the result; in the article linked above, Lichtman gets into how he weights player seasons to avoid weird seasons, like Sandoval's 2016 when he had 7 PA and a .143 OPS at age 29, mucking up the trend.

There are other methodologies, and some of them find different things. Some place the peak a bit later, around 29. Some say that hitters no longer improve as much in their early 20s, perhaps because teams are doing a better job evaluating when minor leaguers are ready for the Show, and promote fewer players prematurely than they might have in the past. But they all show the same thing for the mid to late 30s.

Are there outlier seasons, and seasons contrary to the trend? Absolutely. But there are outliers in both directions. For every George Brett-style renaissance, there's a Vernon Wells: a player who hangs on because of a long and expensive contract.

My point is that you can use individuals to tell any story. Carl Crawford followed a 7-win season at age 28 with a 0-win season at age 29, and then was worth 3.4 WAR the rest of his five year career. Someone saying he was done after his terrible year in 2011 would have been mostly right. (He had an okay year in 2014, I guess.)

So yes, you're right, some players bounce back from injury in their 30s to reclaim previous levels of play. [edit] But the chart suggests that these players — many of whom are memorable! — are not the norm. Are these individuals a better guide to Hanley's prognosis than the aggregate overall trend of thousands of players?

If anything the aging curves may be too rosy, because most players whose play collapses in their 30s drop out of the sample altogether. I doubt Sandoval's likely to see much MLB playing time in 2018, for example — although I could imagine him DHing in the KBO, maybe.
I appreciate the response and agree with most of what you said. I'm familiar with that chart, and I don't think anybody would dispute that the prime years of a player are generally 24-29. But there are three reasons why Hanley's situation really isn't applicable:

1) He's primarily a DH. Defensive skills are impacted sooner and more dramatically by age than offensive skills. Position players tend to deal with injuries and fatigue more than DH's, which in turn impacts their hitting. I would love to see a similar chart for strictly DH's, but I know that would be a small sample size.

2) Hanley has been an elite hitter throughout his career. Career .291 AVG/.852 OPS. He's got a batting title, multiple Silver Sluggers, a ROY award and an MVP Runnerup. The bulk of the players that comprise that chart are mediocre guys who didn't have the raw talent to achieve any of those accomplishments, let alone extend their careers into their mid-30's.

3) Hanley had a really good season just last year, and was heavily impacted by injury this year. Therefore we have every reason to believe he can repeat 2016 next year if he gets and stays healthy.

Agree with you on Crawford, but I don't think you're saying he's an example of age-related decline at 29 ... are you? Especially considering the main reason Theo gave him that contract is because his evaluators determined Crawford was an exceptional athlete who would likely retain his health and athleticism well past his mid-30's.

As for Panda, his drastic decline began in 2012 at Age 25 when his OPS+ dropped 32 points. His numbers declined 3 straight years before even signing with the Sox. Again, hard to argue age-related when he was steadily declining during what was supposed to be his peak years (Age 25-27). But apparently Ben & Company had different numbers that told him to ignore all that, and I'm pretty sure that's one of the reasons why he's no longer here.
 
Last edited:

Cesar Crespo

79
SoSH Member
Dec 22, 2002
21,588
3) Hanley had a really good season just last year, and was heavily impacted by injury this year. Therefore we have every reason to believe he can repeat 2016 next year if he gets and stays healthy.
Even if Hanley does repeat 2016, it doesn't mean he's not in an age related decline and it doesn't mean he's likely to hit as well in 2019 too. He'll be 2 years older and coming off shoulder surgery. The surgery could be a good or bad thing, depending on how he recovers. Regardless, he's on the downside of his career and hasn't really been an elite hitter for a full season since 2009. If you want to argue 2014 is elite, you and I have different definitions of elite. He was really good though. If you want to argue I'm being disingenuous about ignoring 2013, that's ok too but that's going back to when he was 29. In 2018, he'll be 34. I guess to you, that doesn't really mean much though. If you argue his performance drop off from 2011-2017 is all due to injury, I can't really argue against you because I'm sure some of it is due to injury. On the flipside, some of those injuries may have caused permanent production loss. See Adrian Gonzalez.

You also mention guys like Wade Boggs and George Brett who had one year dips in production. Hanley has had 4 off years within a 7 season period. 95 OPS+ in 2011 (92 games), 105 in 2012 (157), 89 in 2015 (105), and a 95 this year in 133. More often than not, he has put up a crappy season since 2010. They aren't really comparable. I've already said one year doesn't really mean much. This isn't a one year dip. And yeah, a 105 isn't "crappy" but relative to his career prior to that, it really is.

I'll agree with you that Hanley's situation is rather unique though, as he's already rebounded twice. You also gave the only other acceptable answer to the question "what caused their decline?" which is injury. As I've said before, I think age and injury are pretty much impossible to separate and the older someone gets, the more injuries they've dealt with, the slower they heal and the more likely they are going to get injured again.
 

Savin Hillbilly

loves the secret sauce
SoSH Member
Jul 10, 2007
18,783
The wrong side of the bridge....
I'll agree with you that Hanley's situation is rather unique though, as he's already rebounded twice.
Agreed. He's an odd duck. What do we think his rational-optimist projection is? I.e., if he's healthy next year and bounces back to whatever a good year would be at this stage of his story arc--not a freak year, just a good year--what does that look like? I'm thinking something like a .270/.340/.460 slash with 50-ish XBH?
 

Kielty's Last Pitch

New Member
Oct 6, 2017
118
Even if Hanley does repeat 2016, it doesn't mean he's not in an age related decline and it doesn't mean he's likely to hit as well in 2019 too. He'll be 2 years older and coming off shoulder surgery. The surgery could be a good or bad thing, depending on how he recovers. Regardless, he's on the downside of his career and hasn't really been an elite hitter for a full season since 2009. If you want to argue 2014 is elite, you and I have different definitions of elite. He was really good though. If you want to argue I'm being disingenuous about ignoring 2013, that's ok too but that's going back to when he was 29. In 2018, he'll be 34. I guess to you, that doesn't really mean much though. If you argue his performance drop off from 2011-2017 is all due to injury, I can't really argue against you because I'm sure some of it is due to injury. On the flipside, some of those injuries may have caused permanent production loss. See Adrian Gonzalez.

You also mention guys like Wade Boggs and George Brett who had one year dips in production. Hanley has had 4 off years within a 7 season period. 95 OPS+ in 2011 (92 games), 105 in 2012 (157), 89 in 2015 (105), and a 95 this year in 133. More often than not, he has put up a crappy season since 2010. They aren't really comparable. I've already said one year doesn't really mean much. This isn't a one year dip. And yeah, a 105 isn't "crappy" but relative to his career prior to that, it really is.

I'll agree with you that Hanley's situation is rather unique though, as he's already rebounded twice. You also gave the only other acceptable answer to the question "what caused their decline?" which is injury. As I've said before, I think age and injury are pretty much impossible to separate and the older someone gets, the more injuries they've dealt with, the slower they heal and the more likely they are going to get injured again.
I guess we'll have to agree to disagree on whether or not a repeat of 2016 would be a good thing for Hanley. I also think 2013's 5.4 WAR, which was good for 12th-highest in the NL, is elite despite the number of games missed ... but again, I respect your opinion and your belief that WAR isn't the one and only stat that determines a player's value. So we'll wait and see how he responds to the surgery, and whether his numbers spike back up next season. It kinda sounds like you don't want his option to vest, even if it means having a good 2018 season. Without knowing who would replace him, and at what cost, I personally can't form an opinion either way on 2019. In general I tend to stay away from predictions, because more often than not they don't come to fruition.
 

streeter88

Member
SoSH Member
Apr 2, 2006
1,807
Melbourne, Australia
I guess we'll have to agree to disagree on whether or not a repeat of 2016 would be a good thing for Hanley. I also think 2013's 5.4 WAR, which was good for 12th-highest in the NL, is elite despite the number of games missed ... but again, I respect your opinion and your belief that WAR isn't the one and only stat that determines a player's value. So we'll wait and see how he responds to the surgery, and whether his numbers spike back up next season. It kinda sounds like you don't want his option to vest, even if it means having a good 2018 season. Without knowing who would replace him, and at what cost, I personally can't form an opinion either way on 2019. In general I tend to stay away from predictions, because more often than not they don't come to fruition.
This is an interesting post. The general feeling on the board - and I have been guilty of it too - is to not want Hanley's option to vest because he is a very expensive DH / occasional 1B with not very reliable production, who is frequently injured and they fear the potential further drop off in his results in 2018 and beyond. The discussion above has been all about the aging curve vs. Hanley's own propensity for injury and bounce back seasons, in other words.

I know you said you shy away from predictions, and I would agree that we do not know enough to be able to predict which way Hanley might produce next year. We could almost simplify it into a 2 x 2 matrix where the axes are OPS and PA with the midpoints at 525 PA (or 120 games played if you prefer) and league average OPS for a DH / 1B. Let's blend those OPS figures 50/50 (.737 for DH and .833 for 1B) to get an average of .785 OPS for an average DH/1B in 2017 - and Hanley had average durability and OPS a bit low this past year with 550 PA (133 games) and .750 OPS. But to be honest, I think most posters would want Hanley's OPS to exceed .800 for it to be classed as a good year.

So, in my little 2 x 2 matrix, if Hanley exceeds 525 PA / .800 OPS, we would class that as an above average year both in production and health. Below that -- especially in OPS, and I suspect SOSHers would hope his option didn't vest (he needs 497 PAs in 2018 to exceed his target 1050 in 2 years), so DDski can spend the $22M on someone more productive for 2019.

upload_2017-10-19_17-53-51.png


To boil it right down, how many posters think that Hanley is likely to exceed those marks in 2018 to put him in the upper right Healthy Slugger quadrant, vs. again be injured or not produce?
 

absintheofmalaise

too many flowers
Dope
SoSH Member
Mar 16, 2005
23,336
The gran facenda
Ian York wrote an article on an aging study he did back in 2016. What makes this one a bit different was he looked at the debut age of the players and how they aged.Ramirez' debut age fits into the parameters.
 

Kielty's Last Pitch

New Member
Oct 6, 2017
118
This is an interesting post. The general feeling on the board - and I have been guilty of it too - is to not want Hanley's option to vest because he is a very expensive DH / occasional 1B with not very reliable production, who is frequently injured and they fear the potential further drop off in his results in 2018 and beyond. The discussion above has been all about the aging curve vs. Hanley's own propensity for injury and bounce back seasons, in other words.

I know you said you shy away from predictions, and I would agree that we do not know enough to be able to predict which way Hanley might produce next year. We could almost simplify it into a 2 x 2 matrix where the axes are OPS and PA with the midpoints at 525 PA (or 120 games played if you prefer) and league average OPS for a DH / 1B. Let's blend those OPS figures 50/50 (.737 for DH and .833 for 1B) to get an average of .785 OPS for an average DH/1B in 2017 - and Hanley had average durability and OPS a bit low this past year with 550 PA (133 games) and .750 OPS. But to be honest, I think most posters would want Hanley's OPS to exceed .800 for it to be classed as a good year.

So, in my little 2 x 2 matrix, if Hanley exceeds 525 PA / .800 OPS, we would class that as an above average year both in production and health. Below that -- especially in OPS, and I suspect SOSHers would hope his option didn't vest (he needs 497 PAs in 2018 to exceed his target 1050 in 2 years), so DDski can spend the $22M on someone more productive for 2019.

View attachment 18033


To boil it right down, how many posters think that Hanley is likely to exceed those marks in 2018 to put him in the upper right Healthy Slugger quadrant, vs. again be injured or not produce?
Good stuff, thank you.
 

Cesar Crespo

79
SoSH Member
Dec 22, 2002
21,588
Agreed. He's an odd duck. What do we think his rational-optimist projection is? I.e., if he's healthy next year and bounces back to whatever a good year would be at this stage of his story arc--not a freak year, just a good year--what does that look like? I'm thinking something like a .270/.340/.460 slash with 50-ish XBH?
That's pretty much in line with what he did this year, except the higher batting average. Seems about right.

So, in my little 2 x 2 matrix, if Hanley exceeds 525 PA / .800 OPS, we would class that as an above average year both in production and health. Below that -- especially in OPS, and I suspect SOSHers would hope his option didn't vest (he needs 497 PAs in 2018 to exceed his target 1050 in 2 years), so DDski can spend the $22M on someone more productive for 2019.
I think everyone on this board would be happy if Hanley put up the the .270/.340/.460 line mentioned above, yet some would still be incredibly disappointed his option vested. The money could be put to better use. It's an interesting subject.
 

EricFeczko

Member
SoSH Member
Apr 26, 2014
4,823
Your post makes me realize that we actually need more information. Is that 6-run drop from 32-34 around the amplitude we might reasonably expect to see in the career of a given player, or is it a statistical artifact of the fact that more players are declining than improving? If player A drops by 24 batting runs between 32-33, but players B, C, D, and E stay flat, that would produce the same average as if they lost 4, 5, 6, 7, and 8 runs. (Check my math.)

I guess I would want to know how many players gradually decline versus completely implode before I try to apply these curves to individual cases.

But I don't think this gets Kielty off the hook.
Right.
What the plot above informs is our expectation for the population of players.
However, we know that aging in this population is heterogeneous, which means this plot is less useful for inferring how any individual player may age.

One way to get a handle on the question is by identifying players that age similarly. A quick and dirty way would be to use the correlations between every pair of set of player trajectories (player A vs. player B, not within player). You could then try to identify clusters, and find the one most similar to the individual you are interested in. Unfortunately, such a method does not establish whether said clusters are real...

An alternative to a data driven approach is to split players into groups based on hypothesis driven measures and see whether fitted trajectories differ. For example, one may expect players with high BMI to age differently from players with low BMI.

EDIT: Iayork's analysis is a great example of the latter. Is debut age associated with differences in player performance trajectory?
 

iayork

Member
SoSH Member
Apr 6, 2006
639
EDIT: Iayork's analysis is a great example of the latter. Is debut age associated with differences in player performance trajectory?
An older article (The Effect of Debut Age on a MLB Player’s Offensive Peak and Decline) has some charts that give a better idea of the variability. For example, this chart

Age-and-Peak-IMG-2.jpg
shows OPS as a percent of peak performance by age, and you can easily see a pretty clear trend, but the outliers are also pretty obvious. For any individual player, you have pretty wide confidence intervals around the expectations.