Subscribe via RSS Feed

Rivera’s Value

[ 57 ] July 15, 2011 |


There has been an interesting discussion in the comments about the value of Mariano Rivera to the Yankees.   On one level, I don’t disagree with the arguments of the skeptics.   While has value is almost certainly understated by WAR — which, if I understand correctly, doesn’t take leverage into account, hence understating the value of a close, I certainly agree that in the regular season Rivera has obviously not been nearly as valuable as Jeter, Posada, Williams, A-Rod, et al.    Another way of looking at the question is Tom Tango‘s study in this year’s Hardball Times book, which assesses Rivera’s value as only about 2 wins a year more than the rest of the Yankee bullpen, which can’t be considered to be significantly more valuable than an ordinary closer and is probably less. Having said that, I think this underestimates Rivera’s value to the Yankees for two reasons:

  • As I assume is widely understood, what makes Rivera by far the best closer ever is not exceptional single-season performances but his remarkable consistency.   His peak value is no higher than the man he replaced, John Wetteland — an excellent but obviously not Hall of Fame caliber closer.    What makes Rivera extremely valuable to the Yankees is that he’s healthy and pitches about as well as any closer in baseball every year (2002 aside.)  Brad Lidge, in his best years, is about as good as Rivera — except that his ERA+ in a given year ranges from 225 to 60.    (Yeah, they made it to the World Series in his worst year anyway, but won only 93 games with an outstanding offense and good rotation; in a good division it would have cost them.)   Rivera may only be a game or two better than an ordinary closer, but he gives the Yankees that edge every year.   They haven’t had to worry about a closer blowing up on them (a la Keith Foulke in 2005), costing them several games.    That’s real value, more than his value in any given season reflects.
  • And, obviously, when assessing his value one has to take into account his postseason record — 139 2/3 mostly high-leverage innings with a 0.71 ERA and 109/21 K/W.   I don’t think there is any serious question that he’s by far the most valuable postseason performer ever, surely relevant when assessing his overall value.   The Yankees don’t win five World Championships with a Joe Nathan or Trevor Hoffman doing an Incredible Shrinking Closer routine in the playoffs.

Comments (57)

Trackback URL | Comments RSS Feed

  1. wsn says:

    This has probably been covered by some other comment, but the Yankees being the Yankees it doesn’t really matter if they are overpaying him relative to his value, only whether keeping him precludes them from getting someone better.

    This may be ignorance on my part, but aren’t there players who can’t handle the New York media scrutiny? If so, it’s probably worth it (again, especially for the Yankees) to pay a premium for someone who won’t get rattled when he occasionally blows a big game and the NY Post starts running mean headlines.

  2. howard says:

    Since I was part of the earlier discussion and since i’m about to get on a plane, let me just say that I think you summarized this very nicely, Scott.

  3. Erik Loomis says:

    “They haven’t had to worry about a closer blowing up on them (a la Keith Foulke in 2005), costing them several games. That’s real value, more than his value in any given season reflects.”

    This is the key to me–that unlike many teams, the Yankees have not had to worry about their closer situation in 15 years. There’s no Keith Foulke. Imagine if Rivera retired after last year. The Yankees offer big money to Rafael Soriano to take over. How would that have turned out? Who would even be closing for them now?

    In the single most inconsistent and unpredictable position in the game, the Yankees have the most consistent individual in history.

    • ralphdibny says:

      Mariano Rivera is the difference between the New York Yankees and the Atlanta Braves. That is his value.

  4. Linkmeister says:

    I don’t know this, but it wouldn’t surprise me if the opposing batter, seeing Rivera come into the game, mentally concedes his chances of success just dropped by some not-small percentage.

    Reputation counts.

    • Joshua says:

      Just like the old “Tiger in the last round” effect. Of course, that’s there until it’s not. Thing is, I thought Mo was showing signs of breaking down four years ago.

      Can someone introduce Rachel Uchitel to Rivera please?

  5. Rich C says:

    Baseball Reference’s version of WAR does use a leverage index in calculating WAR for pitchers (it only really matters for relievers). Fangraph’s doesn’t seem to, but their calculations a harder for me to follow. I’m surprised by this, since Tom Tango developed the leverage index that BR uses, and FG generally utilizes measures developed by Tango (FIP, Tango’s position adjustments, Tango’s weights for wOBA, etc.)

  6. the man he replaced, John Wetteland

    Bar bet!

    Comparing WAR across positions misses the issue of supply.

    There are a lot of bats out there. There aren’t so many closers.

    It’s sort of like the NBA, where a good 7-footer is worth more than a very-good shooting guard.

  7. DivGuy says:

    Fangraphs’ pitcher WAR is based on component stats (K, BB, HR) instead of actual runs prevented. It significantly underrates pitchers who have skills in either preventing hits on balls in play, or preventing baserunners from scoring. Rivera’s cutter is incredibly hard to hit – obviously – and one of the ways this comes up is in his prevention of hits on balls in play. Fangraphs misses that, and as such their WAR is a bad tool for evaluating Rivera.

    Rivera’s career ERA is 2.22, his FIP is 2.77, his xFIP is 3.01. If you base your estimate of his value on either of the latter numbers, you underrate him significantly. You say Fangraphs WAR surely underrates him, this is one way it does.

    • John says:

      It significantly underrates pitchers who have skills in either preventing hits on balls in play, or preventing baserunners from scoring.

      My sense is that advocates of WAR and most other advanced pitching stats tend to believe that there is no such thing as having a skill in preventing hits on balls in play or preventing baseunners from scoring.

      • Bill Murray says:

        I believe the consensus is there is a small effect of preventing hits on balls in play. I have no idea what preventing runners from scoring means outside of getting batters out

        • John says:

          Ability to induce double plays?

          • djw says:

            GB% and FB% are, if I am not mistaken, tracked by and accounted for by FIP. I don’t believe anyone thinks inducing double play grounders is a skill independent of GB%.

            My understanding is that having a influence on BABIP (beyond the effect that GB/FB rates have is generally considered to possibly be an actual skill for some pitchers. I don’t think there’s really any evidence that strand rate in itself is considered a skill, or that there is any good reason to do so.

            • Bill Murray says:

              plus in Rivera’s case he rarely enters the game with runners on — about 1/3 of the time in the playoffs, probably less in the regular season.

            • There is no skill for controlling BABIP per se, but it is accepted that someone like Rivera is hard to hit and therefore hitters don’t square them up very often (and this shows up in batted ball data, as Rivera is consistently at the top of the league in IFFB%). Of course, saying Mariano Rivera is the exception to the rule is kind of superfluous.

        • DivGuy says:

          This is not correct. While BABIP skill – or skill at preventing good contact – is just about impossible to isolate in season-long samples, a survey of career-long data shows a significant “skill” effect.

          See “Solving DIPS” by Tango, Erik Allen, and Arvin Heu.

          Also, there is strong evidence of pitchers with the ability to strand runners above and beyond normal good pitching. Tom Glavine is the poster boy for this effect.

          What is useful about stats like FIP is that BABIP rate and strand rate have huge variance over season-long samples, and so pitchers who do very well or very poorly are often simply suffering the effects of that variance (or “luck”).

          What is problematic about stats like FIP is that some people have drawn the statistically incorrect conclusion that because the skill is hard to isolate in a sample of a certain size, that the skill either does not exist or is not of meaningful size.

          I quote Tom Tango:

          While there is so much noise in the BABIP metric that it almost makes the BABIP metric itself useless, the underlying true talent in BABIP is real and significant. In practical purposes, this means: Get your nose out of your calculator, and watch some baseball!

          • DivGuy says:

            The reason that BABIP can both be just as large a part of a pitcher’s skill as K rate or BB rate, but also be hidden behind variance, is that BABIP effects are much larger than K or BB effects.

            Each 1% change in BABIP has more than 5X the run impact as the same change in K/9 or BB/9, and 4x the impact of HR/9.

            So, the small differences in BABIP skill which are hidden by sampling issues nonetheless matter just as much as strikeout and walk prevention abilities. We shouldn’t use one-year samples to find these skills, but we should recognize that over a career, these skills will show up.

            • But the only way a pitcher can consistently control his BABIP is to consisently have a high infield fly ball rate. Anything past that is too dependent upon defense, which a pitcher can’t control in any meaningful sense. And I can count the number of pitchers who have done that in the last 10 years on one hand.

              Alternatively I suppose a pitcher could have a lot of their fly balls end up as home runs, which would take those hits out of the equation. I doubt that’s what you’re suggesting, however.

              • DivGuy says:

                It’s certainly interesting that particular component parts of ball in play stats can be shown to correlate with BABIP prevention, but I don’t understand what it’s supposed to show in relation to my larger point.

                If it is the case that most BABIP prevention skill is found in the tendency to produce infield popups – which is not my understanding of the current state of research, and I would like the see a link for that – then that’s interesting, but it shows us nothing at all new about to what degree BABIP prevention is a real skill.

                The PBP stuff has advanced our understanding of how pitchers differ in the sort of contact they allow, and by extension how many hits they’ll tend to allow, but the Solving DIPS discussion and the other thread I linked remain the definitive locations for the state of the field on whether these skills exist and whether they are of significant size. The answers in both cases are “yes”.

                • “BABIP prevention” is NOT a skill. The ability to maintain a relatively consistent batted ball profile does seem to be a repeatable skill, and that’s where you’re getting “BABIP prevention.” What you mean, basically, is xBABIP, which correlates to batted balls, because different types of batted balls go for hits more often. A line drive, for example, has an xBABIP of .730, meaning it goes for a hit about 73% of the time. An infield fly ball has an xBABIP of .000, since they basically always result in an out (99.5% of the time) or an error.

                  And this basically makes sense. Batters won’t square up a good pitcher very often, so they won’t hit many line drives off of them. A pitcher who allows a lot of line drives, therefore, obviously isn’t very good. But you have the cart ahead of the horse.

                • DivGuy says:

                  I agree with all of that.

                  I did not think we were having a terminological discussion, but you are correct that I was being terminologically imprecise.

                • JRoth says:

                  But you, Brien, are being too reductive in claiming that the “only” way to control BABIP is IFF% (which, bear in mind, DIPS people used to claim wasn’t a skill – the state of the art is changing, and making definitive statements is a recipe for looking foolish in a few years).

                  Most obviously, we have only very approximate ball in play data – GB/LD/FB/IFF are broad categories. They don’t distinguish between a one-hop smash and a swinging bunt, calling both a GB with a 28% chance of becoming a hit. LD% is supposed to be a proxy for solid contact, but the flaws in that premise are glaring.

                  It all works as a generalization, but the whole point is that, when you see a pitcher with a career-long BABIP that’s out of line with expectations, you’re not seeing luck, and you’re not necessarily seeing IFFs. But you’re seeing something real, and it’s OK if you can’t point to a stat that explains it. When Field f/X comes online in a few years, all this may change.

          • That’s petty outdated too, in that we now have xBABIP, which uses the average BABIP of the different kinds of batted balls to calculate what BABIP “should” be, and then compare it to what it is. Long story short; players don’t actually outperform their xBABIP by all that much, what you see instead are hitters with high BABIP hit more line drives than the average hitter, pitchers give up fewer line drives, etc. So:

            1) The same basic framework holds; players aren’t good because they defy sabermetrics, they’re good because they’re good.

            2) The players who do prove to be the exception to the rule are almost always *great* players, and they’re kind of by definition exceptions to the rule. Saying something is invalid because a Hall of Fame player didn’t fit into the box makes no sense, when it’s the Hall of Fame caliber player who’s obviously outside the norm for a baseball player.

            • DivGuy says:

              I don’t really understand your point here.

              I certainly never said anything about “defying sabermetrics” – I linked to some of the most dense sabermetric discussions that like anyone’s ever had.

              The issue is that “BABIP is not a skill” is false. “BABIP is not a skill for most pitchers” is false. This has been shown to be the case through sabermetric research.

              It is a sad fact that the popularization of sabermetrics has led people to believe that a pitcher who does a good job of preventing hard contact, and in so doing prevents hits on balls in play, “defies sabermetrics”. He doesn’t. He’s a good pitcher in a way that is perfectly in line with sabermetric research.

              2) The players who do prove to be the exception to the rule are almost always *great* players, and they’re kind of by definition exceptions to the rule. Saying something is invalid because a Hall of Fame player didn’t fit into the box makes no sense, when it’s the Hall of Fame caliber player who’s obviously outside the norm for a baseball player.

              If your implication here is that only HoF pitchers have high-level BABIP prevention skills, that is absolutely not the case. I would point to Sid Fernandez, for one. Catfish Hunter, Jesse Orosco, Barry Zito. Certainly it’s more Hall of Famers than not, since they’re better pitchers, but it’s not exclusively Hall of Famers.

              And, of course, your Glendon Rusches and Ricky Nolascos and Javier Vazquezes, who consistently fail to match their K/BB/HR numbers with similarly strong run prevention numbers, have demonstrated a similar lack of skill in the other direction.

              • But that’s what I said. “Weak contact” is best defined by infield fly balls. Without even checking Fangraphs, for example, I can tell you offhand that if you look up some of Zito’s more insane years from Oakland w/ regards to BABIP, you’ll see IFFB% numbers in the 15% range, which is A LOT. Of course, he was also benefitting from the spacious amount of foul territory in the Colisseum, and hasn’t been able to maintain that batted ball profile, so ultimately his poor peripherals caught up to him and his BABIP went way up on average.

                And on the other end, again without bothering to check, I can say you’re absolutely wrong about Vazquez, who always gave up too many home runs, despite his solid K/BB numbers.

                • DivGuy says:

                  Even if it is the case the IFFB% explains most of BABIP difference between pitchers, this affects my point not in the slightest. What point are you trying to make with that?

                  Zito’s become a worse pitcher since leaving Oakland – his fastball is slower, his walks are up, his ks are down. The BABIP problems are certainly partly a function of park, and surely in part variation, but it’s also him pitching worse than he did in Oakland.

                • That’s really not accurate in the slightest. There’s really no meaningful differentiation between his K/9, BB/9, etc. between Oakland and SF, once you account for outliers.

                  In fact, by your standard, Zito is currently set to have the best season of his career, with a .220 BABIP, 76.2% LOB%, 7.5% HR/FB, and 3.18 ERA through his first 34 innings.

                  I’m guessing those numbers are going to normalize if he doesn’t improve that 1.42 K/BB ratio, however.

    • “It significantly underrates pitchers who have skills in either preventing hits on balls in play, or preventing baserunners from scoring.”

      That’s because, in 99.9% of cases, those things are a matter of luck, not skill.

      • DivGuy says:

        You can see the stuff above, particularly the Solving DIPS document. There’s also an excellent discussion in this 2005 Baseballthinkfactory thread. (I stole a bunch of stuff from that thread, I miss when Tango posted regularly at BTF.)

        In small samples, differences in these numbers are usually variance. That does not mean that pitchers have no control over these component numbers, or that we should dismiss career-long tendencies as “luck”.

  8. wengler says:

    The position of closer is extremely over-valued regardless of Rivera’s particular ability. Perhaps his most exceptional quality is having only one pitch and still getting people out.

    My favorite Rivera memory is Game 7 2001, so I guess he is also to be commended for his psychological toughness in coming back from blowing a save that lost the World Series.

    He might be my favorite Yankee, but that’s not saying much.

  9. mpowell says:

    Surely there’s a reasonably systematic way of approaching this. Looking at ERA or something and then extrapolating that to WAR and then trying to convert that into WS titles is just the wrong way to go about this. We can look at every postseason game Rivera appeared in for the Yankees. For each of those we can look at the probability of the Yankees winning the game with an average post-season closer and compare it to the actual outcome with Rivera. If the difference is 96% versus 97% then our job is done, but I’d be shocked if that’s what you’d find. Then you either look at the total number of expected wins Rivera generated or look at each series by itself and see what kind of impact he had based on the expected increase for that series. It would be a lot of work, but not too much that you couldn’t do it if you had the data and really wanted to.

    • howard says:

      mpowell, as i pointed out in the previous thread, and as i’ll note again here, the simple fact is that no other reliever with substantive post-season innings has come close to performing as well as rivera (the closest is papelbon, btw). given that mariano does not suffer from small sample size, i think this is pretty revealing.

      i’m willing to bet, as you are, that there’s a significant difference between the outcome among all teams when you have a lead in the 9th in the regular season and the outcome among the very limited number of very good teams playing the post-season (that is, i’m betting that an empirical analysis would show that in the post-season, there are more late-inning comebacks). however, even if this isn’t true, if post-season closing were just like regular season closing, then we should see a lot of relievers with comparable post-season and regular season records, and we don’t.

      and indeed, the really special thing about mariano, as scott implied up above, is that his performance has been better in the postseason than the regular season, even though he pitches a higher percentage of yankeee relief innings in the post-season than in the regular season.

      • Papelbon has 27 career postseason innings, and his ERA and FIP are both below his career regular season numbers. His ERA of 1.00 is over a run better than his regular season total.

        • howard says:

          brien, exactly: now the question is whether papelbon can do that for another 100 postseason innings.

          • Perhaps. It seems the more relevant question is whether he’ll even throw another 100 postseason innings though, isn’t it?

            • howard says:

              well, brien, just to step back to the original post (about k-rod); i mentioned in comments that when krod first came up with the angels, i considered him as a possible mariano, but clearly that hasn’t panned out.

              as a comparable, i noted that i regarded papelbon when he first arrived as another possible mariano.

              but really, my point was that mariano has been a pretty unique talent: at this stage, i don’t think papelbon is likely to have the career length and consistency of mariano.

              as i noted in the first thread: mariano has pitched the equivalent of two full seasons in the post-season, he has the lowest career postseason era, and he has (by a fraction, to a pitcher with a much smaller sample size in the late oughts) the lowest whip/9 in postseason history.

              so i agree: it’s unlikely that papelbon will even wrack up that number of innings and it’s also unlikely that if he does, he’ll match mariano’s performance (although not impossible).

    • Bill Murray says:

      well in 15% of his playoff appearances (14 of 94) Rivera came in with the game tied which is where the most leverage is. It seems unlikely Rivera could be worth more than 5 games over an average closer. He has come in 20 playoff games with a 1-run lead, this probably 3-4 games and 13 games where his team was behind, which gives maybe another win. So Rivera probably added ~10% extra wins for the Yanks over 94 games.

      A better stat is probably WPA — win probability added http://www.baseball-reference.com/about/wpa.shtml. From 1995-2009 Rivera added about 48.5 wins, which is about 3 wins a year, and in his career has 11.646 WPA, of which 9.8 occurred when he was the closer (6 wins, 42 saves and 6 blown saves). Over his career, Mariano has 51.8 WPA which is very similar to his WAR of 54.6. His WPA is the 6th best since ~1950 for pitchers and about 50% better than Trevor Hoffman the next best career closer. With the same WPA he would be 27th but as some games are missing from 1950 to 1973 and all games from before 1950, some of the players that span that gap (Musial, Ted Williams, Duke Snider to name 3) 33 or so seems better, making Mariano about the 40th best player since 1950 which sounds about right to me

      • Bill Murray says:

        in his career has 11.646 WPA, of which 9.8 occurred when he was the closer (6 wins, 42 saves and 6 blown saves)

        in the playoffs that is and this about 0.5 wins per series, which seems about right to me

        • Bill Murray says:

          Hoffman and Mitch Williams were both negative WPA for their postseason careers.

          Rollie Fingers pretty much matched Mariano for his 72-74 years and didn’t pitch much in the post season outside these years and wasn’t particularly good when he did

          Eckersley was never close although usually not negative (excepting the big -0.828 WPA for game 1 in 1988).

          I am out of great relievers from 1973 to the present

          • howard says:

            bill murray, this is my point: the postseason isn’t the regular season.

            if mariano’s postseason performance were no big deal, then there should be other great relievers with comparable postseason performance.

            if there aren’t, it’s suggestive – given rivera’s large sample size – that we are talking a pretty unique performance quality.

    • Actually, this isn’t that difficult at all. The best way to do it is probably to look at Mariano’s postseason game logs in the prism of the Shutdown and Meltdown stats. You can either add up his WPA and divide it by games, or you can just count how many shutdowns he’s had in the postseason (a shutdown equal +0.06 WPA). I might do it later.

  10. I’m 95% sure fWAR now includes leverage for relief pitchers based on WPA (which would make sense, given that they’re trying to popularize the Shutdown/Meltdown thing), but I could be wrong. I can tell you that that page is obviously outdated, because fWAR does take baserunning into account now, through Fangraphs new base running runs above average number.

  11. c u n d gulag says:

    I know that intangibles are not things that sabermatricians particularly want to talk about because they cannot be measured. And I understand that completely. However, I suspect most of us believe that they’re there. Or, maybe I’m wrong…

    I remember reading an article in the NY Times magazine about the 2011 Philly rotation, and they were talking to Mike Schmidt who is considered, if not the greatest, than certainly one of the top 2 or 3 best Third Basemen of all time, and he was asked to compare these pitchers to ones he had faced. And he talked about Seaver and Gibson and Ryan, and how you not only faced embarrassment, but getting killed by getting hit in the head. He said that you only had a split second to decide, and that if you ducked at a curveball, you’d look like a coward, but if you guessed wrong, an inside fastball from those guys could kill you. He said he didn’t sleep the night before facing those guys. He said he wouldn’t fear the current Philly starters. Respect them – yes. But lose sleep over facing them? No.

    How does not sleeping the night before a game affect the hitter? Most times, I don’t suspect it helps.
    So, fear – of embarrassment, or of physical damage, is an intangible. No?

    Back to intangibles and Mariano.
    I have read many accounts of both opposing Managers and players talking about him “shortening the game,” that they don’t think about with other closers. They feel he is as close to ‘automatic’ as you can find.
    It means that they feel that they HAVE to have a lead before he comes in, and so they may change their approach. To have hitters sacrifice or hit and run in earlier innings rather than swinging away. They may bring a certain reliever in earlier than they normally would to prevent the Yankees from scoring and getting any lead.

    I’m not mathematically adept enough to be a sabermatrician, but the reason I love reading them is that they try to take what used to be called ‘intangibles,’ and figure out what there is that is really ‘tangible’ about them. And I think there’s been a lot of great work done there. But, I still think that there are still areas that are undefined and unexamined, and maybe because they are not definable or can be examined at all, and those are what we call intangibles – which are really, and I hate to use the word, nothing but feelings.

    I’d be interested if some of you more sabermetrically inclined commenters would want to look at this ‘shortening the game’ thing, and tell me what you think about this. Is it an intangible, or can you look at a Managers and hitters prior patterns in other games, Yankee and otherwise> Can you determine whether ‘shortening the game’ is something that can be analyzed, and, if so, is it really an ‘intangible,’ or is it a fable or BS. Sabermetrics have already eliminated a lot of what Managers used to call ‘intangibles’ and managed accordingly. Once in a while, decades ago, some Managers put guys with high OBP in the leadoff even if the guy couldn’t run a lick, while most others put on some speedster whose OBP was only slightly higher than his BA. Which, unless he’s hitting .330 plus, is a detriment – particularly when the speedster then proceeds to successfully steal less than 75% of the time. Until fairly recently, when asked why the Manager would lead off with a slow 3rd Baseman rather than the speedy SS, he would say he had a feeling that this would work better. That that 3rd Baseman had intangibles as a leadoff man. Look at how many horrible leadoff men there were in baseball for decades because Managers went by the book. Thing about how many RBI Bobby Richardson and Tony Kubek cost Mantle and Maris by hitting ahead of them. How many HR’s?

    Enough rambling. What do you guys think?

    • My sense is that 99% of what players from other eras say in the “back in my day” vein is complete and total bullshit, if only because EVERY player seems to think he played in the Golden Age of baseball. Obviously it’s a way to puff yourself up even more, and I basically think it’s nonsense. I mean really, to hear them tell it you’d think Gibson gave up 10-20 runs per year just because of all the damn batters he hit. And by the same token, I’d stack Clemens, Pedro, and Randy Johnson up against anyone from any other era in terms of intimidation factor.

      As to the rest of it, it’s basically just a long, subjective, narrative filled way of saying “Mariano Rivera is very good at his job.” But we knew that already.

      • c u n d gulag says:

        Brien,
        I agree.

        I just want to add the following:

        “…you’d think Gibson gave up 10-20 runs per year just because of all the damn batters he hit.”

        I was more like the mound that he pitched from. Have you ever seen how high the mounds were before 1969 when they lowered them after “The Year of the Pitcher?”

        If you look at the mound in LA, it was amazing that Koufax and Drysdale didn’t need Sherpa guides and oxygen tanks to get up there to toe the rubber.

        • Bill Murray says:

          Gibson hit about 6 batters a season with a high of 13, so it really wasn’t that. The Dodgers had about a 20″ mound back in the 60s, compared to 10″ today.

  12. Joe says:

    “Rivera’s value as only about 2 wins a year more than the rest of the Yankee bullpen.”

    Eh. Don’t believe it. How do you even figure that out? It’s a game of fantasy to try.

    I agree his overall value is his consistency year in, postseason, out. The reputation value alone helps you out there.

  13. I think the best way to look at the value of relievers is to first look at the value of wins themselves. A team’s 95th win has much more marginal value than their 75th win or, say, they’re 105th win. So by that token, if you’re a 95 win team, a 2 WAR closer is a valuable commodity to you, but the same pitcher is basically worthless if you’re a 75 win team. And of course their value goes up greatly once you’re in the playoffs, because the marginal value of each win increases by orders of magnitude.

    So basically, the best way to think of Rivera’s value is to say that he’s been very valuable to the Yankees because they’ve otherwise been a very good team during his career. Put him on, say, the Pirates, and though he’s still the same pitcher on the merits, he has basically no marginal value for his team.

Leave a Reply




If you want a picture to show with your comment, go get a Gravatar.

  • Switch to our mobile site