﻿ Does anyone see a pattern here? | The Hardball Times

# Does anyone see a pattern here?

The tables below reflect twelve free agent signings this off-season. The columns show the respective players’ 2009 WAR, the average annual value of the contracts they’ve signed this off-season (in millions), and the dollars paid to the player per ’09 win.

 Player ’09 WAR AAV \$/WAR Player 1 1.7 \$6 3.53 Player 2 2.4 \$5.75 2.40 Player 3 2.4 \$5 2.08 Player 4 4.3 \$7.25 1.69 Player 5 6.1 \$9 1.48 Player 6 4.5 \$6.25 1.39 Totals 21.4 \$39.25 1.83

 Player ’09 WAR AAV \$/WAR Player 1 0.3 \$5.5 18.33 Player 2 0.4 \$7 17.50 Player 3 0.7 \$7 10.00 Player 4 0.7 \$5 7.14 Player 5 0.9 \$6 6.67 Player 6 2.0 \$7.25 3.63 Totals 5.0 \$37.5 7.50

The two tables look like they’re comprised of signings from two different eras, like one from this economy and one from those halcyon days when owners showered free agents with money, jewels, and keys to the kingdom. But all twelve signings occurred this off-season. The first table clearly shows contracts that are favorable to their respective teams. Around a year ago, free agents earned on average about \$4.4 million per win above replacement and the worst of those six contracts to the team will pay the player just over three and a half million per ’09 win. The second table shows the best contract being slightly worse than the worst contract from the first table and the worst contract being…well…horrendous. Paying a player who’s only slightly better than a replacement level player \$5.5 million? Why? It makes no sense.

Now I’m going to modify the two tables to reflect projected WAR based on CHONE’S projections for the 2010 season rather than the players’ WAR for ’09.

 Player ’09 WAR AAV \$/WAR Player 1 2.2 \$6 2.73 Player 2 1.7 \$5.75 3.38 Player 3 2.3 \$5 2.17 Player 4 2.3 \$7.25 3.15 Player 5 3.9 \$9 2.31 Player 6 4.0 \$6.25 1.56 Totals 16.4 \$39.25 2.39

 Player ’09 WAR AAV \$/WAR Player 1 0.0 \$5.5 —- Player 2 0.6 \$7 11.66 Player 3 0.7 \$7 10.00 Player 4 0.7 \$5 7.14 Player 5 0.5 \$6 12.00 Player 6 0.7 \$7.25 10.36 Totals 3.2 \$37.5 11.72

It doesn’t look much better for the group in Table 2, does it? So what explains this disparity? I’m now going to make one more change to Table 2, using last year’s numbers. Maybe this will shine a light on what’s going on here.

 **The Astros will pay the guy above the same as the Cubs will pay the guy below…for a lot less production.** (Icon/SMI)

 (Icon/SMI)

 Player ’09 WAR AAV \$/WAR ’09 Saves Player 1 0.3 \$5.5 18.33 37 Player 2 0.4 \$7 17.50 0 Player 3 0.7 \$7 10.00 25 Player 4 0.7 \$5 7.14 3 Player 5 0.9 \$6 6.67 10 Player 6 2.0 \$7.25 3.63 27 Totals 5.0 \$37.5 7.50 102

The six players in table 1, whose average annual values of their salaries are roughly equal to those in table 2, are Mark DeRosa, Nick Johnson, Marlon Byrd, Mike Cameron, Chone Figgins, and Marco Scutaro. The six players in table 2 are Fernando Rodney, Billy Wagner, Jose Valverde, Brandon Lyon, Mike Gonzalez, and Rafael Soriano. The obvious reason why the six players in table 2 will be paid about the same as those in table 1 is the fact that they were signed to close baseball games, to pitch the ninth inning. Why else would the Angels pay Fernando Rodney, basically a replacement level pitcher, nearly the same as the Giants will pay Mark DeRosa or the Red Sox will pay Marco Scutaro?

Yes, I know that Rodney blew just one save last year but I also know his career BB/9 is 4.64 and his career FIP is 4.15. I also know that he was a solid ground ball pitcher last year but his 57.9 ground ball percentage was ten points higher than his career average ground ball rate. I also know that his strikeout rate was the lowest of his career last year. I also know that the average leverage index when Rodney entered the game was just eighth among AL closers. And I also know that the Angels are gambling big bucks on Rodney just a year after doing the exact same thing on Brian Fuentes and seeing it backfire. What’s the definition of insanity again, Mr. Einstein?

Let’s compare Brandon Lyon to Marlon Byrd, two men receiving the exact same contracts. Lyon will be the Astros’ closer while Byrd will man center field for the Cubs. Last year Lyon pitched to 314 batters in 78.2 innings. Byrd came to the plate 599 times AND made 341 plays in the field. Byrd affected games more on defense, even if he hadn’t taken one plate appearance for the Rangers in ’09, than Lyon did pitching for the Tigers. It’s not that Lyon’s a bad pitcher, he isn’t; it’s that by pitching only in the ninth inning, he only affects one-ninth of the game every two to three games. Byrd will affect most every inning of most every game for the Cubs this season.

Now, Byrd certainly isn’t a great player. But he is a league average defensive player at a premium defensive position as well as being about a league average offensive player. You could do a lot worse in center field than having a league average center fielder, particularly when you’re paying him the same amount as a division rival is paying someone else to pitch to 300 hitters all season.

While the economy seems to have hit some teams hard, and they’re making decisions accordingly, others are still overvaluing closers. None of these guys is Mariano Rivera or even Joe Nathan. The only one of these contracts that appears to even come close to being a beneficial contract to the team is the one doled out to Rafael Soriano. The others are examples of how some teams continue to overvalue saves. Most of them are pretty good middle relievers who just don’t do enough throughout the course of a season to earn six or seven million dollar contracts when above average shortstops and center fielders are getting the same amount of money.

Inline Feedbacks
frank brownsob
13 years ago

don’t we think that ninth inning outs have more value than previous inning outs. fyi cards vs lad july 18th tyhru 21st at busch

MikeS
13 years ago

Maybe WAR is not the best way to evaluate relievers, especially closers.  Maybe leverage index or WPA would be better.  I dunno, just asking.

For instance: Striking out the nine hitter to open the sixth in a 4 run game?  OK.  Whiffing Joe Mauer with a man on third and 1 out in the ninth of a one run game?  A little bigger.  I think that would give the same WAR, but slightly different WPA.

Lucas A.
13 years ago

I agree.  Since relief pitching is so situational, I think that WPA is the way to go.

Lucas A.
13 years ago

In addition to an ERA-type metric.

Chuck Brownson
13 years ago

WPA is a pretty good tool for evaluating relievers but it makes it tough to compare relievers to everyone else.  Additionally, while it allows us to compare 1 reliever to another, there’s no verifiable way to value 1 WPA.  Is it the same as 1 WAR?

As for outs in the 9th as opposed to other innings, outs in the 9th are more valuable than outs in other innings, thus the emphasis on WPA.  However, Sean Smith has written some about chaining relievers—the idea that, if a reliever goes down to injury or ineffectiveness, he’s not replaced by someone from the minors, as a first baseman might be.  He’s replaced by the setup guy, who is then replaced by the 7th inning guy and so forth.  The point is, if Rodney (for example) isn’t signed to be the closer, it’s not like he’s replaced by the 12th guy in the pen.  He’s replaced by Fuentes.

The significance of that is that nearly every team has a reliever who is getting paid very little but would cost the team only a couple more runs over the course of the season.  For example, the Tigers will pay Valverde \$7 million this year but will pay Zach Miner and Ryan Perry less than a million.  According to Chone’s projections, Valverde projects to be just 2 runs better than Miner and 6 runs better than Perry over the course of the season.  Couldn’t then those extra \$6 million be better used?  Using Miner as a closer may cost them an extra save or 2 over the season but the extra \$6 million toward a starting pitcher or position player is worth 15 to 20 runs, if spent wisely.  It’s simply a poor use of resources to spend too much money on a closer.

WY
13 years ago

I can’t believe I keep seeing this same article over and over (at Fangraphs, Beyond the Box Score, and now here), yet no one seems to be able to draw the obvious conclusion: WAR clearly doesn’t work in accounting for the market price for relievers.

And comparing Brandon Lyon to Marlon Byrd is just completely apples and oranges. Lazy thinking here.

Scottwood
13 years ago

In looking at the last 3 years, there are 5 relievers in the top 10 of WPA for pitchers.  And, 4 of the top 6 are relievers.  Now, a late inning reliever may not have the same “value” as a top tier starter but I don’t see why we need to break every player down by WAR.  It works well for starters and position players and the role of the closer is probably overrated as it is presented in the media.  But, its beaten to death so much that it is probably underrated by sabermetrics.  Teams are not paying for a reliever’s WAR.  They are paying for their potential WPA and one can see why when you look at the win probability stats.

Even when you just look at team WPA added by relievers, you can see the importance of the bullpen.  The Yankees had a team WPA of 9.60 last season.  5 teams had a WPA of 5 or higher.

At some point we need to stop saying how overrated the closer position is and how overrated relievers are and look at the impact they have in terms of win probability.

Chuck Brownson
13 years ago

No one said that bullpens weren’t important, only that relievers are a lot more fungible than many teams believe.  The fact is that Rodney, Lyon, Valverde, and these other guys could have been replaced by other guys that earn a lot less money, thereby spending that money to get a much bigger impact.  If the Tigers want to spend an extra \$6 million a year to gain 3-4 runs, it’s up to them, but those \$6 million could be put to better use elsewhere.

cpebbles
13 years ago

The problem with using WPA or some other leverage-based metric is that it assumes that the increased leverage of a closer’s innings is an innate part of his skill set and not simply the result of managerial decisions.

Most closers are the best relievers in their pen and so some of the difference between WAR and WPA can be attributed to their skill, but the majority of it is strictly a result of modern-day bullpen usage.

Scottwood
13 years ago

CHONE’s projections also have Mariano Rivera as just 1 run better than David Robertson and just 7 better than Alfredo Aceves.  Rivera is not going to be worth just 1.3 wins above a replacement level reliever.  He’s going to be worth much more when you factor in his leverage.  Teams are paying relievers to perform well in those high leverage situations.  They don’t really pay starters or position players to specifically play well in those high leverage situations.  Relievers are a completely different role and that is why WAR can work well for starters or position players and not for relievers.

Lahr
13 years ago

Economic theory suggests that scarce items should have higher prices. Hence, diamond, which are not really needed, are prices higher than water, which cannot live without. So perhaps pitchers with the proper mentality and stuff to be a successful closer are simply very scarce.

GreggB
13 years ago

The shortstop who boots a ball in the fourth and lets in two runs is just as responsible for a one-run loss as the closer who gives up a walk and a homer in the ninth.  It just doesn’t FEEL that way to us, because in the ninth we have emotionally counted on the win and we want someone to “save” it.
Eliminate the emotional element, and the three outs a closer is responsible for is no more critical to wins and losses than any other three outs.  The only difference is closers they don’t pitch in blow-outs.

MikeS
13 years ago

“It’s simply a poor use of resources to spend too much money on a closer.”

No argument there.  Like many people, I’ve noticed that most closers are fungible and overrated.  Big save totals are flashy, but not a necessarily a sign of quality.  The problem is the market.  As soon as you make Perry or Miner “the closer” they are going to want closer money.  Maybe they know they can’t get Rivera money, but they know they can get better deals than they got as set up men.  Baseball has created a specialist that is similar to football’s quarterback in this way.  Even a bad starting quarterback makes great money since starting quarterbacks get paid based not just on talent, but also on being “the starting quarterback.”

David Cameron
13 years ago

Pitcher WAR includes leverage.  I guess I should write a post about this, considering how often people assume that we just left that out for some strange reason.

Lahr
13 years ago

What I am suggesting is that while there are many closers, very few are deemed successful at their jobs. Hence, comparing closers to the “average”  (median or mean) closer is not a good metric. Like all-around, high-performing second basemen, shortstops or catchers, there are so few that WARP or other such measures founded on an average metric under-estimate the true value of players at these positions. Presumably GMs perceive some tipping point or threshold of blown-save percentage beyond which they are unwilling to pay the premium demanded by a prospective closer. For example, last year it may have been that only Mariano Rivera, Jonathan Papelbon, David Aardsma, Joe Nathan, Joakim Soria, Fernando Rodney, Andrew Bailey, Heath Bell, and Houston Street were decent when closing (note: not all had the job of closing all year). Supposing only nine possibly consistent closers can be had, then it makes sense that when one of them is available on the market that their price would go through the roof. Moreover it then yields spillover value to the Matt Thorntons and Matt Guerriers of this world, pitchers who may have closer potential given their stellar performances as regular releivers, although not as closers. In fact, when not a closer role Rodney stank it up, which actually supports the aura he has gained as a hard-throwing closer. (Like Ryan Madson, Thornton also did a rather poor job in the closer role, however.)

Scottwood
13 years ago

Pitcher WAR includes leverage but it takes into account bullpen chaining and does not give them full credit for their leverage.  When a team is signing a closer they are paying him for his full leverage.

Scottwood
13 years ago

How would the free agency reliever signings look of this year and the past couple years if we used a ERA or RA WAR and then gave the reliever full credit for his leverage?  We could use a 3 year weighted ERA or RA WAR if need be.  I imagine that a lot of those signings would start to look a lot better from the team’s perspective.

Using bullpen chaining may be the best thing to do when comparing relievers of one team to relievers from another team.  But, when teams sign a closer, they are specifically paying that guy to pitch in the highest leverage situations.

Nick Steiner
13 years ago

Pitcher WAR on FanGraphs also uses FIP, which will underrate relievers as they will generally under-perform their FIP’s.

Train
13 years ago

Just my poorly worded two cents, but I don’t understand what the issue with this analysis is. What is being measured and compared here is value to the team, not degree of influence on the outcome of the game due to role. Yes, closers have a greater effect on the outcome of a game than many position players – and thus why they are paid more on average vs their true skill – but this is not the point of the analysis as I see it. The point is that this is a bad metric to use in determining how much a team should pay a closer. If you pay Rodney a boatload of money and justify it by saying it’s because his role is so critical you are making a mistake. You can fin another relief pitcher with a better skill set that is coming of a year of low or no save totals , pay him a lot less and have a better closer on your hands. This is the formula the A;s have been using for a decade and successfully. In 2009 the A’s were better off not paying for someone like a Kerry Wood who has lots of time spent in critical situations but really isn’t very good and instead giving the ball to Bailey.

Jon
13 years ago

When calculating WAR for “closers”, is it Wins Above Replacement Closer, Wins Above Replacement Reliever, or Wins Above Replacement Pitcher?  A replacement level “closer” would ostensibly be better than a replacement level reliever, and you would expect a replacement level reliever to have lesser skills but better stats than a replacement level starter (and therefore replacement level pitchers as a whole).

Ryan
13 years ago

Lahr,

Diamonds, like closers, aren’t scarce.  Diamonds are just compressed carbon that is incredibly well marketed and shiny.  Closers are easy to find, but saves are shiny.  So dumb GM’s, like sappy women, go crazy over closers.

pounded clown
13 years ago

As to the closer issue.  A reoccuring theme in the ‘09 post season was blown saves…The Phillies hurt Huston Street, and Broxton.  The Yankees took Nathan, Fuentes and Lidge to the toolshed.  Ryan Franklin did his best Icarus, and the Angels smote Papelbon.  In steps Rivera and suddenly his freakish consistency is a stark reminder of how valuable a save is in the post season.  Perhaps these post season closer performances were still fresh in the minds of GMs during the Winter meetings.  I’m sure it caught the attention of the agents as well.

Chuck Brownson
13 years ago

The point of the post was about how a team should allocate its resources.  The point is that Orlando Hudson would help the Tigers much more than Jose Valverde does for less money.  Yes, as soon as someone is anointed the role of “closer” he’s going to want “closer money” but that doesn’t mean that a team has to pay him that much, particularly to those who are in their first 6 years in the league.

The Astros gave Brandon Lyon the exact same contract that the Cubs gave Marlon Byrd and Byrd is going to help the Cubs much more than Lyon will help the Astros.  The fact that Lyon will get “saves” inflated that contract.  And the Astros fell for it.

The point is that teams, by and large, are better off allocating their resources toward position players and starting pitchers.  They’ll give up a few runs over the course of a season by going w/ Ryan Perry or someone instead of Jose Valverde (though the Angels would be better off w/ any one of the 4-5 relievers who could fill the closer’s role instead of Fernando Rodney) but by giving that money to someone who pitches 200 innings instead of 70, or plays 145 games instead of 65, they’ll be better off in the long run.  Mike Cameron helps the Red Sox much more than Valverde helps the Tigers and Marco Scutaro helps them much more than Mike Gonzalez helps the Orioles.  It’s just a poor allocation of resources, whether they want or expect that money or not.