Baseball for the Thinking Fan

Login | Register | Feedback

You are here > Home > Primate Studies > Discussion
Primate Studies
— Where BTF's Members Investigate the Grand Old Game

Friday, December 06, 2002

Thome or not Thome

Dan examines Jim Thome’s new deal with the Phillies.

He’s one of the best hitters in baseball. He led the American League in OPS in 2002. He finished second in our Baseball Primer AL MVP Award voting. Yet the thought of giving him a 6-year, 85 million dollar contract has left some of us with a Thome-ache.

But that’s what the Philadelphia Phillies did, signing Jim Thome to a long-term contract that will surely be the biggest of the offseason. Now, the question everyone wants to answer is whether the Phillies will get their money’s worth. Of course, we can’t know the answer to that until six years have gone by, but we can guess. And no matter how sophisticated the analysis involved, the result is still a guess; the methodology is selected subjectively. This article will be just as subjective as the rest. In case that bothers anyone, I’ve thrown in some numbers and charts to make this look more like science..

The preferred method for this kind of prediction is to see what other players have done, whether they’re all hitters at a certain age or only those deemed to be similar. Either way, there’s a big problem that I call the Canseco Conundrum. The Canseco Conundrum is simple: even twins are nothing alike. Or, as Police Chief Clancy Wiggum might say, “People are just like snowflakes. They’re both very pretty.”.

Nonetheless, comparing Thome to other players is worthwhile. Simple aging patterns tell us a lot, but they’re just an average progression. We know that many or maybe most hitters don’t do what the average hitter does. So in an effort to narrow down the possibilities, one popular technique is to observe players deemed similar to the subject—in this case, Jim Thome..

Similarity scores are the quickest way to find similar players. They were created by Bill James and are available at They’re based on a variety of criteria, including measures like slugging and on-base percentages as well as at bats and games played. Do these give us the best results for projecting future performance? Well, that leads us off into more subjective areas..

The first question I’d ask in the case of Thome is “What makes Jim Thome Jim Thome?” Where does Thome’s value come from? What separates him from the pack?.

The most obvious answer is his power. So when I compare Thome to similar players, I’m interested in players who share that attribute. To measure Jim Thome’s power, we’ll need to look at a couple statistics that isolate power from other facets of his style of play. And is there a better way to isolate power than with the stat known as isolated power? The answer to that is a resounding yes..

Why I’m Not a Big Fan of Isolated Power

Isolated power is a nice little statistic. It’s very convenient to use. For those of you unfamiliar with it, it’s arrived at by subtracting batting average from slugging percentage. This is a perfectly natural thing to do; slugging percentage is a composite of frequency and magnitude. Batting average is frequency. If we want magnitude, we can just subtract that frequency right on out of there and get something left over that sure looks like magnitude. And it is pretty darn close..

But for the kind of comparison we’re looking for here, close isn’t going to cut it. Sadly, isolated power has a little frequency left in it, like the tissue that’s stuck to the bottom of the garbage can when you empty it. Here’s a glaring example:.

           AB    HR   1B    SO    AVG    SLG    ISO   TB/H     HR/CON
Player 1   20     4    4    12   .400  1.000   .600   5.000     .500
Player 2   20     2    0    18   .100   .400   .300   8.000    1.000

Assume that whenever each of these players connects, they get a hit, and that the above is truly representative of their ability. Player 2, who has much more power but less ability to connect, is correctly identified as the more powerful player by TB/H and HR/Contact, but not by isolated power. The problem is that frequency residue showing up in a big way..

Now, those are extreme cases, but when we look at Thome’s power, we’re looking at an extreme case. And it shows up even in less extreme cases. For example, look at these correlations for all players in 2002 who had 300 or more plate appearances:.

                  ISO TB/H HR/CON
Correl with AVG: .359 .055 .195

Certainly, there’s an element of power driving batting average there, but isolated power is far above TB/H, which completely weeds out frequency. As a sidenote, HR/Contact has a higher correlation with average for the simple reason that when a player makes contact, the more often the result is a home run, the less often it will be an out. Total bases per hit, by not considering outs, avoids that issue completely..

And when you look at how these statistics break down, it’s even more clear..

(H/AB) x (TB/H) = (TB/AB) is the same as saying AVG (Frequency) x TB/H (Magnitude) = SLG..

Isolated power is (TB-H)/AB. By using at bats as the denominator, frequency is inherently included..

Oh, yeah, Jim Thome. So when we look for hitters with the raw power of Jim Thome, we’ll look at their TB/H and their HR/Contact..

Let’s make up something arbitrary called a Truly Spectacular Power Season (TSPS). In order to record a TSPS, a player must be in the top 50 all time in both TB/H and HR/Contact while amassing 500 plate appearances. Such a thing has happened 32 times, and 20 different players have done it. These are the six players that have done it multiple times:.

Player                TSPSs          Years
Mark McGwire                5      1987, 1992, 1996, 1998, 1999
Babe Ruth                   4      1920, 1921, 1927, 1928
Barry Bonds                 3      2000, 2001, 2002
Sammy Sosa                  2      1999, 2001
Willie Stargell             2      1971, 1973
Jim Thome                   2      2001, 2002

Only two of Thome’s similar players through age 31 ever accomplished the feat. They were Reggie Jackson and Harmon Killebrew. That group looks like pretty good company—even from ages 32-37..

The point isn’t that Thome is Harmon Killebrew or Mark McGwire. The point is that he also isn’t Albert Belle, Mo Vaughn, or Frank Thomas. He’s Jim Thome, and we don’t know what he’s going to do over the next six years. We don’t even know the probabilities. We can guess, but there’s a million different ways to get to a best guess..

Based on that, is it reasonable for the Phillies to give 85 million dollars to him? Maybe. We don’t know how likely it is that Thome will be that valuable for the next six years. But we do know how valuable he’d be to the Phillies if they didn’t sign him. What would they have done with the money? Could they have done something better?.

Consider also that the Phillies had a lot more money than that to play with this offseason, considering their deal with David Bell and their offer to Tom Glavine. And consider that if you can afford one of the greatest hitters around, you can try to build a supporting cast for not a lot of money. Thome’s distance from replacement value is enormous—maybe more than the sum of lesser players who can be signed for the same aggregate amount of money..

Does Alex Rodriguez’s contract prevent the Rangers from putting a competitive team around him? Does Oakland’s lower revenue prevent them from building a competitive team at all? No—we know that neither is the case, reality has already contradicted those ideas. What does Philadelphia lose, even in the worst case? They lose wiggle room. If Thome becomes suddenly worthless, then they have a smaller margin of error to work with. Is this signing worth that risk? Is the chance of Thome staying just as productive, or almost as productive, or somewhat as productive—is that worth the risk? Well, that’s for the Phillies to decide, and I don’t see the evidence that their decision—to sacrifice their wiggle room (the value of which should not be underestimated) for the chance of Thome’s potential greatness—is inherently wrong..

We have tendencies—aging patterns, old and young players’ skills, similar batter progressions—but tendencies aren’t even close to certainties. They constantly contradict each other, and they constantly run counter to chunks of reality. And even a risk that has a less than twenty-five percent chance of succeeding can be worth taking, depending on the circumstances..

If you ask me, if you think you can afford the wiggle room, Jim Thome is a risk worth taking.


Dan 'The Boy' Werr Posted: December 06, 2002 at 05:00 AM | 5 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Dan 'The Boy' Werr Posted: December 07, 2002 at 01:08 AM (#607520)
You're right, Dr. Math. Hey, I said it was subjective...

Anyway, I inadvertently doubled them both; they should be 2.5 and 4. Same effect.
   2. Mike Emeigh Posted: December 10, 2002 at 01:09 AM (#607543)
The Indians have probably the best core of youth in MLB. By 2005, the Indians should be very competitive.

The Indians have a fair number of youngsters who *might* develop into good major league players - but so did the Pirates around 1996, and the Bucs are still waiting. Victor Martinez will probably be a very good player, as will Hafner, but we need to wait and see about guys like Broussard, Crisp, and Phillips - I wouldn't bet the ranch on any one of them, as they all have holes in their offensive games. They have a lot of young pitching talent, but converting young pitching talent into good major league pitchers is one of the hardest things to do, and the Indians haven't been able to do it consistently yet. If everything goes right, the Indians could be competitive again by 2005 - but everything never goes right, and the key will be if the Tribe can handle it when things go wrong.

-- MWE
   3. bob mong Posted: December 13, 2002 at 01:09 AM (#607588)
Let's assume that Philly shouldn't be worried that the other three "modern" players (McGwire, Bonds, and Sosa) that have had 2+ Truly Spectacular Power Seasons just MIGHT have benefited from chemical enhancement.

After all, each of those three were having careers which were consistent with their advancing years, and then experienced anomalous jumps in power. Thome, on the other hand... uh oh... peaked in '96, had a consistent mild decline through 2000, and then posted two monster years.

Faced with a guaranteed six year contract, and with baseball testing for Steroids for the first time, I suspect that most players would NOT be using steroids in the future...

Ummmm...that is a pretty simplistic analysis. Let's look at Thome's career a little more closely.

His first full season, at age 23, he posts a 125 OPS+
Unsurprisingly, he takes a big step forward at age 24, posting a 158 OPS+.
This he follows with 3 more years of essentially the same production (OPS+'s of 166, 155, and 153). Not surprising, considering that those are the typical "prime" years (25-27).
Then, unexpectedly, considering his age, at the age of 28 he declines (ever so slightly), posting an OPS+ of 142.
Then, at age 29, perhaps even more surprisingly, he doesn't reverse this slight decline, but continues it, posting his worst (131 OPS+) ever (in a full season).

I think that those two years, his declining age 28 and 29 years, should be seen as the fluke years of his career (to this point).

To continue the narrative, he then bounced back to his previous level of performance (ages 24-27, average OPS+ 158) with an OPS+ of 169, at age 30. He continued that bounce on into last year with what looks to be a career year, posting an OPS+ of 191 at age 31.

Is that a career pattern consistent with steroid use? Maybe...but I doubt it. The list of players who peaked at age 31 is long and prestigious:

Mike Schmidt, peak at age 31 (OPS+ 199, 1981)
Willie McCovey, peak at age 31 (OPS+ 211, 1969)
Felipe Alou, peak at age 31 (OPS+ 143, 1966)
Rod Carew, peak at age 31 (OPS+ 178, 1977)
Carl Furillo, peak at age 31 (OPS+ 147, 1953)
Charlie Gehringer, peak at age 31 (OPS+ 149, 1934)
Willie Stargell, peak at age 31 (OPS+ 187, 1971) - interestingly, he matched this two years later, at age 33.
Minnie Minoso, peak at age 31 (OPS+ 155, 1954)
Bill Terry, peak at age 31 (OPS+ 158, 1930)
Jackie Jensen, peak at age 31 (OPS+ 147, 1958)

All of these players were prior to the "Steroid Era," I would presume.
   4. Marc Posted: December 15, 2002 at 01:10 AM (#607601)
Thome's career trajectory is absolutely typical. That players decline from their prime is a tautology, and that most players prime at 27-29 is well known. Thome peaked a couple years early. No big deal.

Not only did Thome bounce back from the typical early decline years, so do most other players of any real value. Players decline when they lose some skill, the recover when they put experience and intelligence to work to compensate, which they don't do right away. This is absolutely typical.

As to why Thome left Cleveland (no jokes, please), I heard that he said he wanted to win. I hadn't heard that it was the sixth year, per se. He wanted to win. So what is it about the Phillies that would lead Thome to correlate Phillies and winning? Does anybody really think the Phillies are on the rise, other than payroll?
   5. Dan 'The Boy' Werr Posted: December 19, 2002 at 01:11 AM (#607686)

The stats aren't to measure value; they're specifically to measure power and nothing else. They won't necessarily tell you what players you want. They are far from including all the information you want in evaluating a player's value.

Their purpose is to identify where a player's value comes from in order to evaluate and compare players better. You can look at a hitter's slugging percentage and get a better idea of value, but it's handy to know if that comes from batting average or power. Two players can have very similar value stats (OBP, SLG, OPS, etc.), and be compared on those grounds, but the comparison might be faulty if the value comes in completely different ways.

My point in showing Thome's TB/H and HR/Contact is to show a specific attribute where his value comes from. Because when you look at what he'll do in the future, and compare him to other players, you want to know his specific abilities as much as possible. The stats are to measure power, not value, as a player with less power can be more valuable.


Suffice to say I disagree with your premise that players would demand more from teams with high-paid players than from other teams. Texas is able to take on plenty of payroll besides A-Rod and I've never seen any evidence that players demand more from Texas because of A-Rod's contract. Are you suggesting that a player would turn down a Texas team making the best offer, and take a smaller offer from another team because their other players aren't paid as much? Even if there were such players, I don't think all players are that way.

You must be Registered and Logged In to post comments.



<< Back to main

BBTF Partner

Dynasty League Baseball

Support BBTF


Thanks to
Martin Hemner
for his generous support.


You must be logged in to view your Bookmarks.


Page rendered in 0.1962 seconds
41 querie(s) executed