Strike outs are valuable.
Friday I ran the first post in this series, examining what Fangraphs estimated each Yankee hitter was worth to the team on the season, and then comparing that value to what a player was actually paid in 2011. Today, I’ll be performing the same exercise, only examining the Yankee pitchers.
As before, the rules of this are simple. All the data come from Fangraphs. The value and salary numbers are in millions. The difference between a pitcher’s salary and his value is either the profit he generated for the team, if his value exceeded his salary, or the loss he was responsible for, if the salary was larger than the value.
The above table has shaken out precisely as anyone who read Friday’s post would have predicted. I hypothesized then that would have produced the most profit for the Yankees this season, and he’s sitting right on top of the list, sharing a spot with . Those two pitchers generated the most profit for the Yankees in different ways. Colon came on and gave the Yankees a respectable season as a back end starter, but for, literally, a tenth the cost of a traditional back end starter. He was the definition of a high upside signing for the Yankees this season. Robertson, meanwhile, was a walking strike out. He continued his evolution toward becoming one of the game’s most dominant relievers.
comes in next on the list. (Please note that for some reason Fangraphs lists Nova’s 2011 value as being $12.1 million on his personal Fangraphs page, but only $10.8 million on the team page.) He fits the same mold as Colon. Starters are valuable. Nova had a 97 FIP- this year, which is just a bit better than average, but an average starter is worth at least $9 million a season. The team wins out big if it can pay that starter less than $1 million.
should probably be higher on the list than he is. Fangraphs’ value measure is a direct linear transformation of WAR. Fangraphs WAR for pitchers rewards strike out pitchers and punishes ground ball pitchers. Garcia would probably be higher on the list had he produced the exact same results, but with more strike outs.
The big dog, , comes next. While CC was far and away the most valuable pitcher on the Yankees, and one of the most valuable in all of baseball, he is also one of the most highly paid pitchers in all of baseball, which punishes him a bit on the table above. That said, it should be obvious why CC is going to opt out and go for more money this offseason. He made almost $25 million in 2011, and he was under paid.
The list doesn’t get interesting again until the bottom. It should come as no surprise that sucked the most payroll out of the Yankees this past year. He makes a ton of money and doesn’t give the Yankees a ton of anything good in return. , meanwhile, is vastly over paid. His lousy start to the season and trip to the DL brought down his WAR and value numbers, but even in 2010 Fangraphs said he was only worth $6.2 million. This means that even after his best season he would still be over paid at this salary. By the way, his salary will go up in 2012.
I’m cringing at the thought of the comments I’m going to get because is scraping the bottom of the table. I love Mo as much as the next Yankee fan. That said, it is difficult for even the greatest of relievers to produce a lot of value. Relievers don’t pitch enough innings to accumulate as much WAR as top flight starting pitchers. In the case of an under paid player such as David Robertson this doesn’t matter. It matters, however, for Rivera. The irrefutable truth is that no reliever, not even the best in the game, should be paid much more than $10 million a year, if a team’s front office wants to align that pitcher’s salary to his on-field value from the regular season.
Once again, the recipe for identifying surplus value or excessive loss is the same. Just about any young contributor to a team is going to produce surplus value. Highly paid veterans, on the other hand, will often produce losses because it is difficult for them to justify their salaries, even if they have good seasons.
If you enjoyed this article, please consider sharing it!
“The irrefutable truth is that no reliever, not even the best in the game, should be paid much more than $10 million a year, if we take FIP-based fWAR for relievers and the dollar-to-WAR conversion ratio as an irrefutable gospel with no holes whatsoever, which we shouldn’t, and assume that Rivera isn’t an outlier capable of outpitching his FIP on a consistent, annual basis, which he is.”
FTFY.
To say nothing of the marginal value of a win for a team like the Yankees (or the Red Sox or the Phillies or ~12 other contending teams with big money).
While I appreciate your suggested edits to my post, there are a few holes in your analysis. The most glaring oversight is that Mo’s FIP aligned closely to his ERA this season. His FIP was 2.19 versus an ERA of 1.91. Over the sixty innings he pitched this year that translates into a WAR penalty of about one run. So, using metrics based almost entirely on FIP didn’t dent Mo’s value metrics one bit. He was valued correctly this season.
Second, at no point have I taken any of this as a gospel. It was meant to be indicative of good value players, which in and of itself is something I have already said probably doesn’t interest the Yankees. The facts remain, however, that highly compensated players need to have CC type seasons to out perform their salaries. This in no way makes them bad; it makes them well paid. Mo is a great reliever for the Yankees, but given his high salary he is not generating great value for them, nor is he supposed to.
Finally, the marginal value of a win is LESS for a team like the Yankees or Red Sox than it is, say, for the Blue Jays. Increasing the supply of something makes it less, not more valuable. What was your point?
Haha, OK.
Man I just want to focus on this:
“given his high salary he is not generating great value for them, nor is he supposed to.
Finally, the marginal value of a win is LESS for a team like the Yankees or Red Sox than it is, say, for the Blue Jays. Increasing the supply of something makes it less, not more valuable. What was your point?”
That’s an all-time Fangraphs-head classic right there. Got to love that. Rivera doesnt generate good value for the Yankees, and the Yankees don’t have to concern themselves with the marginal value of a win even though they’re in a division with two other consistent 90 win threats, and even though there are huge looming changes to the way the playoffs are structured. Keep sorting the leaderboards on Fangraphs, champ. What you’re doing is fresh, accurate and not at all boring.
Finally, the marginal value of a win is LESS for a team like the Yankees or Red Sox than it is, say, for the Blue Jays.
Echoing Stephen R (or stephen r, if he prefers it): Baloney. Given that about 95 wins are needed to even begin to enter the conversation about a division title, I don’t think it’s nearly that simple. Think about the different results of a Yankee season with 88 wins versus one with 98 wins.
We’re not arguing over the marginal value of 10 wins. The marginal value of 10 wins is huge. We’re arguing over the marginal value of a single win. From that standpoint, any single win is worth less to a good team than it is to a bubble team, or a bad team.
We know this intuitively. From time to time the Yankees will put a worse pitcher on the mound, or ask a reliever to take one for the team. The Yankees can afford to do that more easily than other teams.
For example, the Yankees won the division this year by six games. They had a large cushion before the value of a single win had playoff implications.
You should include incentive money in this analysis. Freddy Garcia’s contract enabled him to earn up to $3.6 million in incentives based on the number of starts he made. He didn’t earn all of it but I’d guess that he probably earned about $3 million of it, which would bring his cost up to approximately $4.5 million. Bartolo Colon also had incentives in addition to the $900K guaranteed if he made the roster. I don’t know the details, but you are probably also understating his cost by a couple of million. It also wouldn’t surprise me if CC, Mo, and D Rob (all star appearance) earned some bonus money.
My fault for not thinking of incentive money. I’ve never actually taken it into consideration before.
I’ve run a version of this post a few times. The next time I run it I’ll try to include incentive money, because it should count.
There is also no deeper purpose to this series of posts. I’m not trying to somehow suggest that Mariano Rivera is anything less than the greatest reliever ever. He absolutely is. My goal is only to show that it is often surprising which players outperform their contracts versus the players that under perform their contracts. When a player is highly compensated, it is difficult for him to outperform his salary.
Including the incentives might change the order and make Nova the greatest value but it doesn’t change the big picture of which players have a lot of surplus value and which don’t.
I actually agree with you on Mo. All of the elite closers are overpaid. That said, WAR doesn’t include post season contributions and Mo is incredibly valuable to have in the playoffs. I don’t think it makes up the $5.2 million gap but I would “overpay” a little based on his value in the playoffs.
Exactly. My point is only to flush out the big picture.
I would also overpay for Mo. In fact, if I were the Yankees I would over pay for a lot of players, which is precisely what they do. They do this because of playoff contributions, because it is more important for them to have good to great players on the field year-in and year-out than it is for them to find bargains, and because they can afford to.
Good post Mike. I too laughed after reading this and thought to myself, “he’s going to get killed in the comments for Mo being so low”. But everything you said is true, relievers are vastly overpaid. But no one will ever complain about Mo because he still gets the job done.
Joe, thanks for the kind words. I really, really do appreciate them.
I also agree 100%. Baseball over pays established relievers, relative to their regular season, on field performance. In Mariano’s case, the Yankees have demonstrated consistently that they are willing to over pay him. As you correctly point out, he is so good, and so reliable, both in the regular and post seasons, that the Yankees are happy to pay him that surplus.
From that standpoint, any single win is worth less to a good team than it is to a bubble team, or a bad team.
That 91th win was really worth lots less to the Rays & Red Sox?
I didn’t think so.
I had intentionally avoided this subject in my earlier comments, but I’m happy to tackle it now.
No, absolutely not. The 91st win was worth tons of money to both the Rays and the Red Sox.
Valuing wins is not a linear function. Some potential wins are not worth a lot of money, while other wins are worth boat loads of cash. As you correctly point out, one single additional win this season would have been worth millions to the Red Sox. Whereas, any single win in July for them wasn’t worth all that much.
From a statistical stand point, we’re actually discussing two different concepts.
One concept we are discussing is the marginal value of any given win during, say, a 95 win season. On average, a 95 win team is going to the playoffs, so a single win isn’t worth all that much to them. This team could probably afford to have lost up to two or three additional games before putting its playoff spot at risk.
The second concept we are discussing is the value of a specific win, once the distribution of wins and losses throughout the course of the season is known. That changes everything. Once the Red Sox and Rays knew the season was coming down to a single victory, winning that game was worth as much as the 89 to 90 victories that came before it.
We can apply this example to real life. A dollar isn’t particularly valuable to most people. However, if that dollar will be used to buy a winning lottery ticket, and you know this has a high probability of being true, suddenly that same dollar is more precious.
On average, if we are selecting a single victory from the regular season, then that victory tends to be worth less to really good teams than it is to less good teams. This is not the same as looking at a highly leveraged game late in the season, and saying it is also worth less to the good team. That isn’t at all true.
Well, yes. That’s my point, and not one acknowledged in your post. Rivera is held out of situations where the win is less (or averagely meaningful) and put into situations where the win is more meaningful (e.g. Torre would *not* use him for more than one inning in the regular season, but quite regularly use him two innings in the post-season). Your analysis did not include that.
“Relievers don’t pitch enough innings to accumulate as much WAR as top flight starting pitchers.”
The fault lies not in Mo but in how is used (or under used) by the Yankees. At age 41 (42?) he is obviously not going to be used for many multi-inning appearances but he could be used in higher leverage situations for a similar amount of innings and produce more value than he does by being reserved for “save” opportunities that might never happen or could be handled by lesser pitchers.
Is there any way to work leverage into this analysis or is it already accounted for in the WAR formula?
That’s a great point. The Yankees themselves limit Mo’s ability to accumulate lots of WAR because they limit his usage. For example, if Mo pitched as many innings as CC did this year, at the same value, he would have been worth nearly 10 WAR, which is crazy good.
I don’t know if WAR takes leverage into consideration. I doubt it considers it directly, the war B-Ref’s leverage index does, but for relievers it probably uses a combination of the value of strike outs and the penalty associated with letting inherited runners score to capture leverage and performance in those situations.
feliciano was a hugh mistake i shocked they past on scott downs the guy is great.angels got him right..sori , could never have gotten that money if i was calling the shots..marte what can you say./. MO is paid right. AJ became a disaster after we signed him, but i liked the signing at the time..now buyers remorse..CC is right..all the rest were bargain basement..all deserve more money..hey igawa is finally gone..
sori could not shine downs shoes
I would like to add a note about getting caught up value. It’s really not what the Yanks are all about. They would rather have a marginal profit on 5 CC Sabathia’s than turn a big gain on a back end starter like Colon. As a franchise they’re all about stars, big winners and going for it all every year. In a perfect world you’d like to do both, but doing great on pre-arb players is where small market teams like the Rays and Royals live.
That being said, love this series MJR. Great stuff.
Glad you like the series Steve, and I agree 100%. The Yankees don’t need to extract value from their players. I think the team would happily lose $3 million on all its players because it means the big salary guys are getting it done on the field. I just find this info interesting to look at, especially in the offseason.
And just to go on from this, the argument from above would be that Colon was more valuable than Sabathia last year. Is that the conclusion we want to be arguing?
Not exactly. CC’s WAR still blew away Colon’s. His NET Value was higher, but again I’m not sure the Yanks obsess over stuff like that.
Considering they left Colon off of the postseason roster, I think that answers your question. There is no way to sanely argue Colon being more valuable than CC. Both did their jobs this year, but they both had vastly different jobs. CC is suppose to lead a staff, Colon was just suppose to not throw up on himself. That alone defines each ones value.
I am not arguing that Colon is more valuable that CC. Sabathia was worth $28 million on the field, while Colon was worth just $11.4 million on the field. CC was one of the three most valuable pitchers in baseball. Bartolo was average. I’m just pointing out that for an average pitcher, the Yankees got very good value from him. I think I can safely speak for the Yankees when I say they would rather field an entire staff of CC’s and pay them like CC, than an entire staff of Colon’s, even if it meant paying them like Colon.
Can you explain to me why the WAR/Value conversion ratio changes from player to player? I’ve listed the conversion ratios below, working back from your data which you presumably mined from Fangraphs. The numbers below indicate the supposed value of each WAR for each player accruing positive value.
Colon – $3.93M
Robertson – $3.93M
Nova – $4.0M
Garcia – $4.4M
Sabathia – $4.03M
Logan – $3.86M
Wade – $4.25M
Noesi – $4.33M
Ayala – $4.5M
Chamberlain – $3.75M
Hughes – $4.0M
Laffey – $3M
I may be way off-base here, but I can’t figure why each WAR isn’t valued similarly, particularly amongst pitchers, and particularly amongst the same subset of pitcher type (starter v. reliever). Why is a WAR from Sabathia worth $400,000 less than a WAR from Garcia? And why is a WAR from Ayala worth $700,000 less than a WAR from Logan?
Ah, I see the problem. Your stated numbers in the Value column don’t match up with the current numbers on each player’s respective Fangraphs player page. In most cases, you’ve understated their Value, thus skewing the conversion ratios. The conversion ratios appear consistent when the correct Value numbers are given. Here are the first 4, just as an example –
Colon – $12.9M (you listed $11.4M).
Robertson – $12.4M (you listed $11.0M).
Nova – $12.1M (you listed $10.8M).
Garcia – $9.9M (you listed $8.8M).
Sabathia – $32.2M (you listed $28.6).
Not sure where you’re getting your data, would love to see an explanation though.
I took the data from Fangraphs team page for pitchers. Apparently there is a major discrepancy between what is listed on the team pitchers page and what is listed on a player’s individual page. I didn’t catch this until after I ran the post. I explained as much for Ivan Nova, who was the only pitcher I caught it for. I don’t know why the data would be different on the team and player page. I pulled from the team page because it is easier to do.
I can’t say I really blame you. No idea why Fangraphs would list the values differently, seems like a major oversight on their part. Figured you had pulled the data from somewhere, since the $/WAR ratios weren’t consistent (wondered at first whether you had inputted your own $/WAR value to calculate the value from the WAR totals).
I’m not sure why they have the discrepancy either, but it’s large. I’m not trying to reinvent the wheel on my end. I hate pulling data.
There is such a thing as thinking too much and forgetting obvious points. Different innings have different value. The Yankees are the most profitable baseball team, what constitutes value for them is different from it would be for the Padres. This is why “moneyball” will never win a championship, it isn’t designed to, only to make it cheaper to be competitive. As for which players are valuable, we hardly need all the stat-head numbers for that. Also, stats will never adequately tell you when a player is a bad attitude case like Soriano, or predict when a player will be unable to perform under more pressure like Crawford moving to Boston. This is why closers are always under-appreciated statistically, because getting the last 3 outs of a close game in October is hardly the same as 3 outs with a 3-1 lead in April, or in September when both teams have fallen out of contention and are playing their minor leaguers. If all innings were the same then yeah $15 mil for a closer is too much. But when he pitches only in the most important situations, and re-signing him keeps him away from Boston, then it’s a different ballgame.