Star Crashes

October 2, 2013

Prospects? Rookies? Best of the best? Why aren’t the falling stars just as interesting as the rising stars? After spending a thunderstorm of money the last two years, the Los Angeles Angels of Anaheim were supposed to possess one of the greatest trios of hitters in the history of the game: Albert Pujols, Mike Trout, and Josh Hamilton. However, the two that they purchased—Pujols and Hamilton—have been piddly meteorites this year. 

So, let’s do some crashing-star analysis on these two salary burdens. The first thing we should do is analyze their health. I know shoulder problems (Hamilton) can be career-threatening, but Hamilton seems to be recovering and is on a hot streak as I write this. 

I get plantar fasciitis (Pujols) myself, but I’m not a doctor nor a world-class athlete, so I’m not going to suggest how it might affect Pujols. I do know he’s a tough guy who has played through and excelled with many other injuries. He’ll do whatever it takes to get through this and be the best he can be. 

However, age eventually gets to every player. How long players fight it off depends on many things, such as their fitness level, their mechanics, their genetics, etc. This is not my expertise either, so l’m just going to look at players with career trajectories similar to Pujols and Hamilton and see if we can make some guesses about their chances of coming back. 

What got me started on this was a live Baseball Prospectus chat with Ben Lindbergh in which he said he expected Pujols and Hamilton to do better next year than this year. I teased him that he was really going out on a limb with that prediction—and asked if he thought they’d get back up to .900 OPS. Ben was kind enough to admit that it wasn’t his boldest prediction and gave it another try. He put the probability of either hitting .900+ next year at 15%. Later in an e-mail, Ben added that the probability for either to comeback to .900+ in any future year would be about 15% each. Let’s see how that holds up historically. 

First, let’s find a group to compare with Pujols and Hamilton. These two are both in the early half of their 30’s and had multiple impact seasons before flopping this year. After considering WAR, OPS+ and Win Shares, I settled on just plain OPS as a comparison stat. Different stadiums and different eras can have a huge effect on OPS, but I wanted to keep this as simple as possible. OPS is easiest to look up and it doesn’t matter if the player is a DH or a shortstop. If a player can have a few seasons of hitting over .900 OPS by the age of 31 as Pujols and Hamilton have, then I considered them similar for the purposes of this study. 

The ability to hit .900 OPS was the inspiration for this study in the first place. Pujols has had seven seasons hitting over 1.000 OPS. Hamilton has had three-and-a-half seasons over .900—one of which was over 1.000. Pujols is in another echelon as a hitter, but Hamilton has had the more recent .900+ season. Since this study is about the two of them, I left the minimum criteria for inclusion at two seasons of .900 (minimum of 350 plate appearances). The .900 seasons must have come by the age of 31 and before a big decline season. 

The qualifying decline season for this study is hitting .780 OPS or less before the age of 35. (Ages are based on their age on July 1.)  I chose .780 because Pujols hit .767 this year. As I write this, Hamilton is hitting .736. .780 covers it with a little leeway. I chose 34 as the age limit for hitting the low point because Pujols is 33 and Hamilton is one year younger. I’m not looking for exact matches, I need enough data to make this study significant. The results were slotted into a four dimensional grid to find the significance of 1) the player’s age when he collapses, 2) the previous number of .900 season and 3) the era in which he played; all that against 4) the various levels of comebacks. Another marker tracked was the players who had a 1.000+ OPS season. 

Of course, there are complications. If a player didn’t play a significant amount one year, what then? If the player came back and immediately had a season over .780 OPS with at least 350 plate appearances, then I ignored it. Otherwise, I counted it as the start of his "big slip." If a player with a down year comes back to post an OPS over .780, then has another downfall before 35, I count both years and their results. Even if he takes several years to come back (see Eric Davis), I tally them all. An exception to this is when a player stopped playing for military service. I just skip those years. 

The result echelons (#4 above) are based on any subsequent season of: 1) full .900+ comeback (350 PA or more); 2) moderate comeback: .850-.899 OPS (>350 PA); 3) slight rebound: .780 to .849 OPS (and some strong performances under 350 PA); and 4) a burned out former star with an OPS under .780 the rest of his career (or vastly diminished playing time).  

Any active players who are not yet 35 are disqualified. Of the active players disqualified by youth, three have yet to recover: Kevin Youkilis, Jason Bay and Brad Hawpe—now a fourth Mark Teixeira just came down with a washed-up season this year. Four others have had moderate comebacks so far: Ryan Howard, Chase Utley, Adam Dunn, and Travis Hafner. 

Joe Mauer had a qualifying bad season at 28. Derek Jeter had one at 34. They both recovered impressively, but have not returned to the .900 OPS threshold. Hanley Ramirez had three .900 seasons before he was 27, then hit .712. Last year at 28, he wasn’t much better at .759. Now, mid September this year, he is over 1.000, but due to a torn thumb ligament and hamstring issues might not reach 350 plate appearances. 

Five active players did come back with at least one strong full season. One is only 30 (David Wright) and the other four are over 35 and therefore are included in the study: Carlos Beltran, Todd Helton, Jason Giambi, and Aramis Ramirez. I think it is fair to count these older guys as well as the 39-year-old Jeter, because I had decided on 35 as a reasonable cut-off before noticing that they all had successful comebacks. 

This bodes well for the aged Angels of Anaheim: Pujols and Hamilton. Looking at all 15 active players, three super-.900 comebacks is over double the suggested 15%. The historical comeback rate for sluggers younger than 35 is not so peachy, but we have to acknowledge that sports medicine and personal training has come a long way over the last 120 years. There should be special consideration, however, that some players in PED era may have illegally aided their returns that would be tougher to do now.  

A Zachery Levine article in Baseball Prospectus shows a significant increase in the WARp of players in the over-37 age bracket playing during the ‘90s, then an explosion of value from these players from about 2001 to 2007 as MLB finally made cheating with PEDs far more difficult. Without any accusation here, the players in the study with comebacks during the years 2001 to 2007 that brought themselves back over the .900 level were Reggie Sanders (age 35), Jason Giambi (34), Frank Thomas (35), Jim Thome (35), Carlos Beltran (29), and Juan Gonzalez (33). 

Below we will look at the recovery success of twentieth century players who turned 30 before 1947; post-integration players who turned 30 before 1980; and the rest. 

There were 134 qualifiers with 194 qualifying bad off-seasons. Hopefully, I didn’t miss too many with my semi-manual research method. 

For each era on the table below you can see the percentage of players who recovered to three different echelons as described earlier.

  

Recovery by Era

 

Era \ Pct. Achieved

>= .780 OPS

>= .850 OPS

>= .900 OPS

1901-1946

     63% (30/48)

      29% (14/48)

       15% (7/48)

1947-1980

     61% (35/57)

      26% (15/57)

       11% (6/57)

1981-2013

     65% (58/89)

      54% (48/89)

       31% (28/89)

Total

     63% (of 194)

      40% (of 194)

       21% (of 194)

 

Looking at historical norms, we see Ben’s rough estimate of Pujols or Hamilton returning to .900 hitting wasn’t far off. However, twice as many players from the last quarter century have been able to achieve that level. Even if we exclude all six of the .900+ comebacks of 2001-2007, that brings percentage of full comebacks from the last couple decades of steep downfalls to 25%.

 Let’s see what this table would look like if we only looked at players with multiple .900+ seasons before age 31 and with at least one season over 1.000:

 

Recovery by 1.000 OPS Players

 

Era \ Pct. Achieved

>= .780 OPS

>= .850 OPS

>= .900 OPS

1901-1946

     88% (14/16)

      56% (9/16)

       31% (5/16)

1947-1980

     64% (9/14)

      36% (5/14)

       14% (2/14)

1981-2013

     56% (9/16)

      38% (6/16)

       13% (2/16)

Total

     70% (of 46)

      43% (of 46)

       19% (of 46)

 

There is no significant change in the recovery rates, except that a higher percentage of the early 20th century 1.000 hitters made full recoveries. Man, they must have been tough guys—or, perhaps, we’ve become more fragile. . . or some other explanation such as the most likely one: this is a small sample size. However, it would be unfair to completely ignore these oldtimers. If we take the results from the more recent hitters here with greater weight, then Ben’s assessment becomes bang on. Thirty percent of the 1.000+ OPS players and a frightening 44% (100% - 56%) of the more recent 1.000 hitters don’t even recover enough to play a full season of .780 OPS. Ben’s assertion that Pujols and Hamilton will be better next year appears here to be not as obvious as I chided him for.

 Let’s break this down by the number of seasons over .900 OPS before the player’s very bad year.

 

Recovery by Years of .900 OPS

 

Pct. Achieved

>= .780 OPS

>= .850 OPS

>= .900 OPS

2 years of .900

     56% (50/89)

      29% (26/89)

         9% (8/89)

3 years of .900

     68% (32/47)

      55% (26/47)

       36% (17/47)

4+ years of .900

     71% (41/58)

      43% (25/58)

       28% (16/58)

Total

     63% (of 194)

      40% (of 194)

       21% (of 194)

 

These results might imply that the peak of your hitting doesn’t affect your chances of regaining high impact hitting as much as doing it year after year. This bodes much better for Pujols and Hamilton.

 Yes, the recovery numbers are a little more impressive for players with exactly three years of excellence as compared to four or more. This is likely because being a little further along in your 30’s gives you more chances to accumulate so many .900 seasons—the older you get, the more difficult it is to rebound. Let’s see how much age plays a part in a star hitter’s comeback chances:

 

 Recovery by Age of sub .780 Season (after 2+ .900 seasons)

 

Bad year \ Recov.

>= .780 OPS

>= .850 OPS

>= .900 OPS

Age 25 or 26

              (1/1)

              (1/1)

             (0/1)

Age 27

      67% (4/6)

      33% (2/6)

       17% (1/6)

Age 28

      91% (10/11)

      55% (6/11)

       18% (2/11)

Age 29

      81% (13/16)

      38% (6/16)

       19% (3/16)

Age 30

      67% (16/24)

      54% (13/24)

       33% (8/24)

Age 31

      63% (20/32)

      34% (11/32)

       19% (6/32)

Age 32

      69% (24/35)

      49% (17/35)

       26% (9/35)

Age 33

      62% (24/39)

      33% (13/39)

       23% (9/39)

Age 34

      37% (11/30)

      30% (9/30)

       10% (3/30)

Total

      63% (of 194)

      40% (of 194)

       21% (of 194)

 

It makes sense that age and health would be the key factors in our ability to recover from a decline. Indeed, our bodies take longer to recover from injuries as we age. It is not a smooth curve, however. The thirty-two-year-olds in our study had more successful recoveries than the 31-year-olds. Then there is a big dip in recovering success from washed out seasons of 34 year olds. Again, Hamilton is 32 and Pujols is 33.

 Applying all the info on these charts to Hamilton and Pujols, you would conclude they have just over a 60% of getting back over .780 OPS, a 45% chance of returning to the .850+ level, and nearly a 25% of returning to the .900 level. Of course, it is more important to read their medical prognosis and their dedication to coming back than these crude actuarial charts, but you’ll have to do some spy work for that. 

Fun Facts

Did you notice there was a 25-year-old with a couple of .900 seasons who had a sub .780 season? That was Johnny Bench. He never hit over .900 again, but he did have three more full time seasons over the .850 mark—including one at .889, plus three more over .800. Getting that level of offense from one of the best defensive catchers in history is still great.

 There were two qualifying 27-year-olds who never significantly recovered. One started his career on the Reds three years after Bench’s retirement: Kal Daniels. Through age 26, his five-season slash line was .300/.402/.514. He had 68 stolen bases in 84 attempts (81%) over his first three years before numerous knee operations ended his base stealing feats. It didn’t seem to affect his hitting, though. He hit a .920 OPS out of pitcher-friendly Dodger Stadium in his fifth year. His MLB career was finished two years later at age 28.

The other 27-year-old crashing star played for the Phillies in the late 1920s and early 1930s. Through the age of 26, Don Hurst also had five seasons in the majors producing a slash line almost identical to Daniels’: .313/.396/.516. Note, however, that he played in the Baker Bowl—a hitter’s paradise for the left-handed first-baseman. Hurst held out the following spring and never got his bat back in the swing of things. The next year he was traded for a rookie named Dolf Camilli, who was less than two years younger than Hurst. Camilli went on to have an excellent career earning MVP votes seven times. Hurst played his last game in the majors at the end of the year. 

None of the available web bios (SABR, Baseball-Reference, Baseball Almanac) gave any reasons for Hurst’s or Daniel’s early demises. Both showed below-average ranges for the least demanding positions that they played. Weaker fielders get fewer chances to get their hitting back. 

Have there been any teams such as the Angels who had two hitting stars collapse to this level in the same year? Yes, but, please, wait for Part II. Here’s a tease: there has only been four such occasions: two from the pre-war era and two from this young century.

 
 

COMMENTS (3 Comments, most recent shown first)

hotstatrat
I agree with you both about taking the results here and applying them to 2013 ballplayers as though it is a solid reliable odds maker. There is one sentence in the 2nd paragraph that was changed from "The best thing we can do is analyze their health" to "The first thing . . ." What I was trying to say was that you have to get to the specific reason why the player suddenly stopped hitting so well. Was he aging rapidly or fighting an injury? The more we know about all aspects of the player's health the better we will be at determining his odds of a strong recovery. This article was just to give us a historical perspective.
2:04 AM Oct 5th
 
OldBackstop
I think if you had to pick one narrow sliver topic in which historical data is nearly moot, it would be recovery from bad seasons. Injury recovery/surgery/rehab aside, even in cases of mere mysterious performance dips, the psychological, medical, motivational and (legal) pharmacological capabilities of today are just apples to oranges to the beginning of the century.

Nice work on the research....I think the interesting stories would be the individual ones, how someone came back from injury or alcoholism or a car wreck or whatever....in aggregate, I don't think the data is worthy projecting forward to an individual, except maybe in specific circumstances (e.g.: Tommy John surgery).​
9:09 AM Oct 3rd
 
MarisFan61
I hesitate to say, because I've been sort of killed here every time I've made mention of this factor, but....

Past data may be fairly irrelevant on something like this because of the presence of PED factors, either a possible drop to 'genuine' levels because of stoppage of use, and/or possible PED-related breakdown.

Note: I said "may" and "possible." We don't know. But if we ignore it, we might be missing the 800 lb. elephant in this story.

(Do I think it is probably a factor for the key noted current players and that it renders the past data largely irrelevant? Yes.)

Sabermetrics seems generally disinclined to take much note of this factor, perhaps from reluctance to suspect without clear basis. That's admirable, but it might get in the way of knowledge.
2:10 AM Oct 3rd
 
 
©2019 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy