Eliminating Teams Using the Thome Corollary
In the first chapter of his (excellent!) Big Data Baseball, Travis Sawchick describes how now-Pittsburgh Pirates General Manager Neal Huntington used an interesting bit of historical analysis to help inform a controversial roster move. Writes Sawchick:
Using the software [DiamondView], the Indians made key decisions, such as when they elected not to sign aging star Jim Thome to an extension after the 2002 season, in part because of the database, the Cleveland Plain Dealer reported. […] Huntington noted that the Indians found in a payroll analysis that no major league club, dating back to 1985, had won a World Series when committing 15 percent or more of its payroll to one player.
Let’s call that 15% rule The Thome Corollary.
Has The Thome Corollary held up in the years since 2002, when price tags for choice free agents and franchise cornerstones has escalated at a rate far greater than your journeyman’s/rookie’s salary? The answer is: “No, The Thome Corollary has not held up,” or, “Yes, it sure has, with some small tweaks.”
It turns out that seven of the thirteen World Series champions since 2002 have had a player who takes up more than 15% of their team’s Opening Day payroll:
| Year | Champion | Highest-Paid | Highest Salary (in $M) | % of Payroll |
|---|---|---|---|---|
| 2014 | Giants | Matt Cain | 20.8 | 13.9 |
| 2013 | Red Sox | John Lackey | 15.9 | 10.2 |
| 2012 | Giants | Barry Zito | 19.0 | 14.4 |
| 2011 | Cardinals | Matt Holliday | 17.0 | 15.5 |
| 2010 | Giants | Zito | 18.5 | 19.2 |
| 2009 | Yankees | Alex Rodriguez | 33.0 | 16.3 |
| 2008 | Phillies | Pat Burrell | 14.3 | 14.5 |
| 2007 | Red Sox | Manny Ramirez | 17.0 | 11.8 |
| 2006 | Cardinals | Albert Pujols | 14.0 | 15.7 |
| 2005 | White Sox | Paul Konerko | 8.8 | 11.6 |
| 2004 | Red Sox | Ramirez | 22.5 | 17.6 |
| 2003 | Marlins | Ivan Rodriguez | 10.0 | 22.2 |
| 2002 | Angels | Tim Salmon | 9.7 | 15.6 |
So while plenty of teams have “violated” The Thome Corollary, only four did so by more than a percentage point: the 2010 Giants, 2009 Yankees, 2004 Red Sox, and 2003 Marlins.
I feel comfortable moving the 2009 Yankees and 2004 Red Sox into their own category: the mega-spenders. The Yankees were first in the league in salary in 2009, and the Red Sox were second in 2004. Yes, they both violated The Thome Corollary — but not by very much. Furthermore, they did so with excellent players (Rodriguez had 4.0 WAR that year; Ramirez had 3.3 in his) still had significant amounts of money to spread out over the rest of the roster. The lesson still stands, I think, for Huntington and most of the rest of baseball, who are in small or medium markets by comparison.
So what about the 2010 Giants and 2003 Marlins? They weren’t big spenders, and they actually are the two biggest violators of the rule. Basically, if a team violates The Thome Corollary to such a degree, they must do two things in order to also win:
1. Be active at the Trade Deadline. The Marlins paid the eventually hefty price of Adrian Gonzalez to receive Ugueth Urbina at the deadline, and they also acquired Jeff Conine. The 2010 Giants only traded for one excellent player at that deadline — Javier Lopez, who is still with the team — but were generally active, also adding Chris Ray, Mike Fontenot, Jose Guillen. But this is probably not as important as:
2. Have a historically awesome rookie class. The Marlins had a 23-and-under nucleus of Josh Beckett, Dontrelle Willis, and Miguel Cabrera. Giants 23-and-unders included Madison Bumgarner, Pablo Sandoval, and Buster Posey. Uh, pretty good.
Of the seven teams who had one player violating The Thome Corollary, only one of them had two players who were receiving over 15% of the team payroll. And, as you can see below, this is also a team with a historically awesome rookie class:
| Year | Champion | 2nd-Highest-Paid | 2nd-Highest Salary | % of Payroll |
|---|---|---|---|---|
| 2011 | Cardinals | Albert Pujols | 16.0 | 14.6 |
| 2010 | Giants | Aaron Rowand | 13.6 | 14.1 |
| 2009 | Yankees | Derek Jeter | 21.6 | 10.7 |
| 2006 | Cardinals | Scott Rolen | 12.4 | 13.9 |
| 2004 | Red Sox | Pedro Martinez | 17.5 | 13.7 |
| 2003 | Marlins | Luis Castillo | 4.8 | 10.6 |
| 2002 | Angels | Kevin Appier | 9.5 | 15.3 |
Those 2002 Angels saw debuts from John Lackey, Scot Shields, Brendan Donnelly, and Francisco Rodriguez, and two of their team leaders in WAR (David Eckstein, Troy Glaus) were both on rookie contracts.
So! It looks like a more accurate, updated version of The Thome Corollary goes like this:
A. No team has won the World Series while dedicating at least 18% of the total payroll to one individual player — unless that team has significantly talented young players and is active on the trade market.
B. No team has won the World Series while dedicating at least 16% of the total payroll to two separate players.
With these new requirements in mind, let’s look at all of the players who are earning at least 15% of their team’s Opening Day payroll this season (or, in violation of the original Thome Corollary):
| Team | Player | Salary | % of Payroll |
|---|---|---|---|
| Twins | Joe Mauer | 23.0 | 21.2 |
| Rockies | Troy Tulowitzki | 20.0 | 20.6 |
| Mariners | Felix Hernandez | 24.8 | 20.1 |
| Mets | David Wright | 20.0 | 19.7 |
| Mariners | Robinson Cano | 24.0 | 19.4 |
| Blue Jays | Jose Reyes | 22.0 | 17.4 |
| Indians | Nick Swisher | 15.0 | 17.0 |
| Phillies | Ryan Howard | 25.0 | 17.0 |
| Phillies | Cliff Lee | 25.0 | 17.0 |
| Rangers | Prince Fielder | 24.0 | 16.9 |
| Rockies | Carlos Gonzalez | 16.4 | 16.9 |
| Cubs | Jon Lester | 20.0 | 16.6 |
| Angels | Albert Pujols | 24.0 | 16.4 |
| Tigers | Justin Verlander | 28.0 | 16.2 |
| Phillies | Cole Hamels | 23.5 | 16.0 |
| Cardinals | Adam Wainwright | 19.5 | 15.9 |
| Blue Jays | Mark Buehrle | 20.0 | 15.8 |
| Mets | Curtis Granderson | 16.0 | 15.7 |
| A’s | Scott Kazmir | 13.0 | 15.5 |
| Indians | Michael Bourn | 13.5 | 15.3 |
There are three teams that we can eliminate for paying 16% or more to two players on their roster, none of whom were close to the World Series anyway: the Rockies, Mariners, and Phillies, who have three players over the threshold. This confirms the reputations that these teams all already have — transcendent stars who toil in anonymity, surrounded by near-anonymous teammates.
We’ll have to use our judgment with four other teams, and I’m sure you’ll agree with mine:
–New York Mets: Wright’s injury issues are certainly not making that outsized contract look any better. In addition, the Mets are just a few percentage points away from violating the two-over-16% rule with Granderson’s contract. As impressive as the Mets apparently bottomless young pitching staff is, I think it’s too much to ask any team to overcome both tennants of the New Thome Corollary. Eliminated.
–Toronto Blue Jays: Just like the Mets, the Blue Jays have one contract over 17% (Reyes) and one that is a hair away from 16% (Buehrle). Coupled with an inexperienced young pitching staff, I also declare them eliminated.
–Cleveland Indians: With two salaries at 17.0% and 15.3%, the Indians are definitely not in violation of the New Thome Corollary. As lovely as Francisco Lindor and Trevor Bauer and Danny Salazar are, though, the lack of contributions from both high-salary players make this, I think, too steep a hill to summit. Sorry to do this right after the NBA Finals, but: eliminated.
–Minnesota Twins: Whoda thought that consistent ol’ Joe Mauer would ever own one of the league’s most bloated contracts. While the idea of the Twins as playoff contenders could be laughable by as quickly as August, Byron Buxton is here and Miguel Sano could be shortly on his way. Accordingly, I give them the tiniest sliver of a chance.
Miles Wray contributes sports commentary to McSweeney's Internet Tendency, Ploughshares, The Classical and Hardwood Paroxysm. Follow him on Twitter @mileswray or email him here.
“Tenant” = renter, “tenet” = principle
he said ‘tennant’
Which as everyone knows means “to play a time-traveling doctor”
Still my favorite Doctor. Matt Smith is pretty great, but Tennant takes the cake, or should I say the Banana?
My favourite Tennant role so far is the Hardened Bitter Cop in season 1 of Broadchurch. It’s a pretty powerful detective show on Netflix. Ignore subsequent seasons.
Hmm, I’m assuming Tennant isn’t in subsequent seasons, hence ignoring them? I’ll have to add it to my list.
Well, it’s like True Detective in that each season is a case, except that for whatever reason, they decided seasons 2-3 should be all the people from the first case reacting to the reveal of the murderer’s identity and being sad. Thus lack of plot, stupid twists.
I’d agree that the second season was not as good as the first, but they haven’t shown the third season yet, so it seems premature to dismiss it.
Actually, he was referring to the lead singer of Pet Shop Boys.
Any way to measure the growth rates of (1) minimum salary (2) arb salaries (3) FA salaries to determine what the 15% SHOULD BE given the relative inflation of FA salaries to minimum salaries?
Love this article. Fun way to look at things. Hope you’re right on the Mets, and way to twist the knife with Cleveland. Exceptional.
Considering the playoffs are a crapshoot, wouldn’t it make more sense to look at percentage of teams that make the playoffs instead of World Series winners?
I’d even take it further than that. Since teams can’t really control the amount of wins required to make the playoffs, it should really be looking at winning percentage vs players making 15% of the payroll.
When the Jays add a few pieces it will push those two guys under this ridiculous rule.
I thought the notion of not spending too high a % of team payroll on any one player stemmed from when Rangers’ owner Tom Hicks was looking for an excuse to trade away a pricey but valuable Alex Rodriguez about 12 years ago.
Isn’t it a bit premature to eliminate teams based on Part A of your New Thome Corollary? I’m not even hiding my allegiances here, either. Yes, the Mets are close to the 2 players over 16% rule, but bottom line, didn’t violate it. Harvey, deGrom, Syndergaard, Matz, Familia, and d’Arnaud constitute as clear a group of “significantly talented young players” as there is in the game today. And since there has been no trade market yet, a crucial piece of your new corollary hasn’t even been tested yet.
Maybe not quite as good as the Mets group, but the Jays have Hutchison, Osuna, Travis and Pillar. Pompey, Norris, Boyd might still factor in this year as well.
They might be active at the deadline as well.
But really, aren’t payrolls a self-imposed limit these days?
This is an excellent example of people taking correlation to imply causation when there’s no reason to do so.
Yep. It would be much more interesting to explore whether there’s any underlying principle that results in the observed correlation.
Besides the underlying principle that there will always be some number to fill in to the statement “no team has won the World Series with a player making more than __% of their payroll?”
Yes, beyond that. You could certainly explore which types of teams tend to commit high percentages of payroll to individual players. Are they high-payroll teams who have money to burn? Low payroll teams who want to hold on to an aging home-grown superstar? Just stupid teams, generally speaking? Does the type of team that does this account for the observed correlation? Etc.
Or is the observed correlation mostly a result of the relationship between salary and experience? I.e., the players who make a lot of money are older than the players who don’t, so the teams who commit that money tend to be relying on declining players? Again, etc.
There are all sorts of angles to explore that might yield some real insight.
And, I’m sure everyone would agree that
Ding ding ding. That said, I think there is something interesting to the question of tying up a lot of hope in one player — both in terms of expected contribution and the associated opportunity cost of spending a large amount of money — as understood against the value gained by having so much production come from just one spot on the roster.
I have one cross-correlation theory. Many teams that know they have bad rosters won’t spend big, they’ll rebuild with a lot of rookie minimum players around. But most teams do have a couple huge free agent contracts lingering even in down years. So the 15% could be an artifact of reduced overall payroll spending on bad teams.
The title made me curious.
The content made me distraught. This is finding a coincidence and supposing it a rule.
If a cup is made out of fishnet one shouldn’t need to pour water into it to know it’ll leak.
All it would take is one well-run organization conserving payroll over several years during a rebuilding cycle, hoarding prospects all the while, to sign the best FA on the market as the enter the peak of their win curve to blow this theory out of the water.
Take the Cubs or Astros, with all these promising young players, both teams having slashed player payroll during rebuilding years. This article implies it would be foolish for them to sign the best FA on the market next offseason to a $20-25M a year deal. On the contrary, if the goal is winning a title, and maximizing your title window, that’s exactly what they should do.
No. Both teams look to be unleashing “historically good” cadres of young (and cheap) players so they fit the Corollary.
As long as the oversize contracts they dish out aren’t overlong and prevent the teams from paying the young’uns down the road, adding in established talent is the proper move for those teams.
I guess none of you guys took high school geometry (or remember pieces of it from 25 years ago). A corollary is a proposition that follows from a previously proven one. It’d be more accurate to call it the “Thome Hypothesis.” Other than that, interesting.
Yeah, but I also took marketing and the Thome Corollary sounds cooler. #branding
I thought it was an episode of the Big Bang Theory 🙁
Seems sort of a shame to lay it all on Thome, a nice guy who managed about 25 bWAR after he was too old and deteriorated to be worth re-signing.
So what was the team in 1985 (or earlier) that used more than 15% of their payroll on one player?
This feels to me like just post hoc finagling.
Blue Jays, Citing ‘Thome Corollary’, Quadruple Salaries of Hutchison, Estrada
Mets have insurance on David Wright’s contract where they will collect 75% back of the salary for whatever time he misses (http://www.si.com/mlb/2015/05/27/david-wright-contract-new-york-mets-injury-insurance). They will get back at least $5 million of his salary if he returned tomorrow.
Could invest that on the current roster (especially if he comes back) to augment the roster. Most likely a big bat in the outfield (I dont know what they do with Granderson or Cuddyer but they could use the $ to put them out to pasture somewhere). Alternative is Tulo.
Great article, very good read. I’ll be sure to share it.
To a large extent, the 15% rule is a warning that a purely “stars-n-scrubs” roster is a very risky approach for teams with genuine aspirations, especially when it comes down to dishing out big contracts to free agents (which tend to b e older players).
The irony that the team the first publicized the guideline is now one of its more prominent victims is actually kind of sad; they literally knew they were taking on a huge baseball risk just for sake of a bit of PR to distract from ownership’s windfall, selling 20 years of TV rights.
As for pre-85 examples?
No need to leave town: just look up the Wayne Garland 10-year contract.
This is exactly right. The 15% rule assumes that resigning Thome would be a waste of money UNLESS the team also spent money on a surrounding cast which would bring his salary under the 15% of payroll. It makes perfect sense that no team with *only* a stud can win a World Series title. I think the underlying test here is how much support from 2-WAR players is that star player receiving? If he’s making more than 15% of the roster then he’s not getting support. If he’s making less then you’ve got the depth required to make a postseason run. Most teams have a star capable of earning 15% of the lowest payroll team’s salary. The teams that fill out their roster win World Series and thus pay that star less than 15% of their teams payroll.
That is how the ARod deal played out for Texas.
He delivered MVP numbers every year but the team got nowhere for lack of good pitching. They didn’t develop it cheaply in-house and the one attempt to buy one blew up in their faces. After that, the budget couldn’t cover their needs so the only solution was to punt his contract.
Should Fielder really be on this list? Isn’t Texas only paying ~$18 million of his salary?
So the Zito and Ramirez contracts both violate and provide support for this rule?
The really odd part about the whole thing is that the Indians did make a really aggressive offer to keep Thome. He just took an even bigger offer from the Phillies. But right up until he decided to go to Philly the Indians were intent on keeping him.
Let’s take a look at the Mariners again. Lumping them in with the Phillies and the Rockies is lazy at best. The Mariners have Seth Smith on his way to a 3-4 WAR season making $6-7 million, a 4-5 WAR 3B making less than $5M, an All-Star & Cy Young caliber pitcher who is coming back in two weeks with no spot in the rotation making only $7M. Happ is on course for a 3.5 WAR season making only $7M
Then you have the young talent, Brad Miller is a top 5 SS right now and due for some offensive progression, while guys having better years to date are playing over their head based on BABIP. Miller is making the minimum and is on course for a 4 WAR season. Walker, Paxton, Elias, Montgomery, Carson Smith are all making the minimum and are on course for 1 WAR or more, even considering the horrible start to the year for Walker, the late call up for Elias and Montgomery, as well as the injury to Paxton, and that Smith is a reliever. Then you add in Wilhelmsen making $1.4M and Lowe less than $1M, who are both on course for 1 WAR. That’s 9 out of 12 pitchers being worth north of 1 WAR each.
If Trumbo can play like he did in Arizona this year and Jackson can play like he has the last month or so, then Cano and Zunino can rebound to what they were last year, which is not unreasonable, the Mariners would be one of the 5 best teams in baseball regardless of payroll. It’s just simply irresponsible to write off a player like Cano with his long history of success or to think that other proven players won’t step up and play like they have. As for Zunino, last year was like the 25% of his true talent, so it’s not unreasonable to think he could raise his numbers to do that. In May he had a .775 OPS. Today the Mariners go for the sweep against the Astros a team that fits the mold of the Thome Corollary, yet who have played way better than they really are talent wise. Oh and did the Indians win a World Series after not signing Thome? So the only thing that corollated was them not signing Thome and them not winning a world series.
Nice post. Where exactly did you buy those rose-tinted glasses?
Add Nuño to the list as well, making 10 pitchers out of 12 who are on course for 1 WAR and he would be another guy making the league minimum.
What is the true correlation with current teams and the Thome Corollary? Well, it’s simple. Back then teams didn’t focus on getting prospects to the team quickly, so it was all about buying a winner, rather than building a winner. Now 8 of 12 teams broke the mold because all teams are starting to value their prospects more. The correlation is championships won buy teams focusing on younger players, just like the NFL, where effective players are pricing themselves out of jobs with so many young players ready to step in and replace them.
Worst statistics in a FG article since that piece about the O’s and random deviation about a year back.
With all due respect, this article represents poor analysis. That said, understanding the fallacy here is equally as valuable as if we had good analysis because oftentimes knowing what not to do is equally as valuable as knowing what to do. The Thome Corollary is false. The “updated” corollary is equally false. Correlation (w/r/t % of payroll) does not imply causation (winning a World Series). Moreover, creating multiple exceptions to a “rule” based upon the observed results since then, is self-serving and not predictive at all.
The “new” corollary:
A. No team has won the World Series while dedicating at least 18% of the total payroll to one individual player — unless that team has significantly talented young players and is active on the trade market.
B. No team has won the World Series while dedicating at least 16% of the total payroll to two separate players.
This is like making up nonsense. Why not say, no team since 1985 has won the World Series without having a 2-WAR player? That’s equally valid. The article and corollary is all about one thing: roster construction. The fact is that many teams do not have one player earning over 15% because they have several similarly-priced players spread across their 25-man roster…whether they be highly-paid players on a high payroll team or average-paid players on an average payroll team. (Yes, we can list exceptions to this all day long.) Of course a team can dedicate a lot of payroll to one player if they have “significantly talented young players” because then any high-priced free agent will eat up a large percentage of the overall payroll since the other players are all on minimum contracts. The total budget that a team has to work with plays an important role here as well as getting the most value out of each dollar allocated to the roster and from each roster spot.
Based on the new “rules,” the author eliminated the Mariners who, while violating the rule last year as well, performed quite well. Eliminating the Twins and Mets at this early juncture seems premature. However, it is completely wrong to eliminate the Blue Jays. (Disclaimer, I am a Yankees fan, so no bias on this statement here.) The Blue Jays have the BEST run differential in all of baseball at +78 (which should have them at a much better record) and sit 3 games back in the highly competitive AL East and have a GM that is known for bold trades. Are we going to eliminate them because they have (rule-violating) Reyes and Buehrle? If anything, such a team should be applauded for its young talent (after trading away so much of their farm) and the vets they kept on team-friendly contracts (Bautista and Encarnacion) and talent they recently traded for (Josh Donaldson and Devon Travis) and the Russell Martin signing. I could see them going all-in (with their upcoming free agents) and trading for both Hamels and Papelbon (or going after Cueto). But, the author has declared them “eliminated.”
The teams that violate the Thome Corollary (or the “new” corollary) simply do NOT (as the author incorrectly states) have “transcendent stars who toil in anonymity, surrounded by near-anonymous teammates.” But, yes they have to have talented young players in order to succeed. Thank you for stating the obvious there.
Here’s my “new” Thome corollary: spend boatloads of money fairly evenly (or at least in the areas providing the greatest impact…unless you buy certain free agents at discount rates due to market inefficiencies) or just have lots of talented young players (mixed in with veteran players as needed)… or just look at percentages and create fictitious rules 🙂
Frankly, Huntington used the 15% rule to govern free agent spending by a small market team. He knew that such capital utilization would leave holes on his team that were not already filled by young players nor could be filled by other free agents (due to a lack of funds). In determining the motivation of different actors in the market, we must remember that it is all relative. Huntington did not uncover an absolute edict to govern all player acquisitions. (He also used a small sample size – 18 world series winners – which then failed to hold true 4 times in the following 13 years.) Moreover, one must remember that yesterday’s market inefficiency is today’s market overvaluation.
Also, to the author: please learn what “corollary” means. It’s not the same as “conjecture.”