VORP Shmorp

David Sheinin wrote the following in yesterday’s Washington Post.

VORP can open your mind. It can bring your world into crystal-clear sharpness. Go ahead — try some. Did you know, for example, that if you exclude the resurgent Cristian Guzmán (VORP: 21.5), the Washington Nationals’ offense has a negative VORP — which, in essence, means if you released every last one of their position players (except Guzmán) and replaced them with cheap, waiver-wire scrubs, the team would be better off?

Or, to be accurate, the Nationals would be expected to be better off — because there is a theoretical aspect to VORP, which stands for value over replacement player.

What is it, you ask? It measures the number of runs a player contributes (or, in the case of pitchers, prevents) beyond what would be expected from a “replacement-level” player — which is to say, one that could be had as a cheap fill-in and who would be expected to produce at around 80 percent of the league average at his particular position during a particular year.

As for a team — that is, a theoretical team — made up entirely of replacement-level players? According to Keith Woolner, the sabermetrics pioneer who invented VORP in the late 1990s, “They would be expected to win between 45 and 50 games, which is comparable to the worst teams we see.” Well, not all of them: The 43-win Detroit Tigers of 2003 own the worst offensive VORP (-50.8) of the past 50 years.

I don’t use VORP, and it’s not that I don’t understand the concept: I just cannot figure out what I gain by using it over other metrics. Sheinin gives the following reasons that I should change my mind.

Why should you care about VORP? Because it presents the most complete picture of a hitter’s or pitcher’s true value. Unlike most other statistics, for example, VORP accounts for why a catcher — for whom it is difficult to find a replacement, because not as many players are capable of playing there — is more valuable than a left fielder with similar offensive numbers.

I don’t understand how VORP helps me here, and I don’t believe I’ve ever heard this explanation. Are catchers really more valuable than equally-talented batters who play left field because of scarcity? There are plenty of catchers in the minor leagues and major-league teams often carry three catchers. Teams don’t normally carry nine outfielders, do they? Imagine how bad a ninth outfielder must be–see the 2008 Braves if you are having trouble imagining this. Teams can draft and develop more catchers if they believe there is a shortage. I suspect that catchers are more expensive because they offer a greater defensive contribution. After all, they are the only other player besides the pitcher involved in every pitch. I admit that valuing catchers is difficult, but if VORP has a breakthrough at the catcher position by better-valuing the on-field contribution, then I’d like to see this spelled out.

It accounts for the fact that a run prevented is more valuable in 2008 than during a low-scoring year such as 1968.

You can make the same corrections to any baseball metric. The simplest is OPS+, but you can make more precise adjustments for any offensive statistic as Michael Schell has done. To me, that is superior to VORP, because I know what each of those statistics is telling me.

And it also accounts for the fact that, say, a .600 slugging percentage for someone who plays home games at Coors Field isn’t as impressive as the same percentage for someone playing at Petco Park.

Again, you can adjust any statistic for home-park bias. Park-effect corrections are hardly a novel contribution.

There is also the potential gains of valuing a player relative to “replacement level” as opposed to the average. I still don’t get the advantage of this. First, you have the task of defining replacement level. What is the point of this exercise? It is just an alternative benchmark to the average. I can explain to any baseball fan: “this player is above/below average.” To explain a player relative to “replacement level” requires a long, boring, and unnecessary conversation. Below-average players are valuable, and this isn’t difficult to understand.

What about using VORP to judge salaries? For example, if the league-minimum salary is $390,000, and a team signs a replacement level outfielder for $1 million, hasn’t the team overpaid? Not at all. Player value is determined by opportunity cost as determined by marginal revenue product (MRP). If a player generates many millions of dollars, his value is determined by this, not by how much he makes. Teams pay players with less than four years of service (approximately) less than their MRPs because the collective bargaining agreement allows them to do so. A team that plays a young and reserved player forgoes the potential return from trading the player to another team or from keeping down his service time. Signing a veteran for $1 million can be cheaper than promoting a young player who would provide equal value now by holding down service time.

My point isn’t that VORP is an awful or useless stat. To the contrary, there is clearly useful information contained in it. And those who prefer to hold discussions based on this metric should continue to do so. But there is no need for someone who does not speak the language to learn the ins an outs of a new metric, as Sheinin suggests. I can talk about all its components without dropping the V-bomb. If you want to talk hitting, we can use OBP and SLG. Then you can bring in stolen bases and defense to capture other effects. For pitching, we can use strikeouts, walks, and homers. The big advantage of these is that I can have these conversations with people other than die-hard stat-heads. I can also explain the advantages of these metrics over traditional triple-crown stats, and that is a huge benefit.

I view VORP as an insider language, and by using it you can signal that you are insider. It’s like speaking Klingon at a Star Trek convention. I can signal to others who speak the language that I am one of you. But, the danger of VORP is that once you bring it up the discussion goes down the wrong path as the uninitiated have reason to feel they are being told they are not as smart as the person making the argument. It’s like constantly bringing up the fact that you only listen to NPR or watch the BBC news at dinner parties. The response is likely going to be the same, “well fuck you too, you pretentious asshole!”

Last year, Murray Chass wrote the following.

I receive a daily e-mail message from Baseball Prospectus, an electronic publication filled with articles and information about statistics, mostly statistics that only stats mongers can love.

To me, VORP epitomized the new-age nonsense. For the longest time, I had no idea what VORP meant and didn’t care enough to go to any great lengths to find out. I asked some colleagues whose work I respect, and they didn’t know what it meant either.

Finally, not long ago, I came across VORP spelled out. It stands for value over replacement player. How thrilling. How absurd. Value over replacement player. Don’t ask what it means. I don’t know.

The thing is: I can actually sympathize with Chass here, though for different reasons. I too get the occasional VORP e-mail, and my normal first reaction is to roll my eyes. I don’t speak VORP, and I shouldn’t be expected to do so. If you want to talk about why a player may or may not be valuable, we can have that discussion in a language that I speak.

42 Responses “VORP Shmorp”

  1. Ken Houghton says:

    Isn’t VORP just the equivalent of economists’s discussion of “equilibrium”?

  2. JC says:

    I don’t see the connection. I find the concept of equilibrium to be quite useful.

  3. Zach says:

    It accounts for the fact that a run prevented is more valuable in 2008 than during a low-scoring year such as 1968.

    I’m not so sure I follow. For an offense, isn’t there diminishing marginal utility of runs? For example, the 10th run of the game isn’t nearly as valuable as the 3rd. If that’s true, then isn’t preventing the marginal run in a low-scoring year more valuable than in a high-scoring year? It seems to me that preventing a run in a year in which a team scores an average of 3 runs per game is more valuable than in a year when a team scores an average of 4 or 5 runs a game.

    Another way of thinking about this is the variance of runs scored per game. I would argue that during low-scoring eras (i.e. deadball era), the variance of runs scored is much smaller than during the 1990′s (i.e. a high-scoring era). Any given run in the deadball era is much more likely to be the decisive run in a game. Therefore, peventing a run in those years is much more likely to change the outcome of the game, making it more valuable.

    Am I missing something that the author is not?

  4. Ev says:

    I was a little surprised to read this here. While I understand the points about stat-head lingo and that VORP and such things are not aesthetically pleasing to the casual fan, this is a site I come to to read about the correlation of weather to home runs, temperature to attendance, attendance and GAS PRICES. That you hadn’t heard of VORP until a year ago doesn’t really fall in line with these things.

    While it is certainly a fringe statistic, it is interesting to note, especially when you look at your team and what positions could require improvement. Like you said throughout your post, that’s something you could use other statistics to determine, but I enjoy seeing EXACTLY what a perfectly average SS would do instead of Julio Lugo in that line up.

    Also..and I read this on NPR’s website via the BBC..a study shows that diatribes about taxes and name-calling of executives at groundbreaking ceremonies that is irrelevant to most of us is also pretty annoying at parties, and everywhere else.

  5. JC says:

    I have known about VORP for many years.

  6. Doug says:

    Do you honestly not understand the point about catchers versus leftfielders?

    While he perhaps didn’t phrase it perfectly, I took it to mean: for any given level of offensive production, there are fewer people on planet earth who are capable of providing that level of offensive production and playing catcher than there are people who are capable of providing that level of offensive production and playing left field.

  7. JC says:

    Does the world produce fewer catchers than left fielders?

    I tend to think of catchers as players who have strong arms and have major-league bats but not great bats. Players who can catch but are excellent hitters are moved off the position, like Dale Murphy and Craig Biggio.

  8. Rick says:

    EV-Since when does replacement level = league average? League average equates to a .260 or so BA. Replacement level is around a .230 BA. Huge difference. I do think that many players could replace Lugo though. He is the bane of the existence of Red Sox fans.

  9. Max says:

    but catchers are less offensively inclined than LF’s are. Thats what this guy was saying. VORP, gives us the players value relative to the position. So therefore a LF and a C with the same hitting skills, the C will have a higher VORP, because a replacement level catcher is worse than a replacement level LF. Therefore, the equally talented LF is relatively more valuable. I understand and use the metric. But also, I understand that if the players contributions are the same, than the VORP wont necessarily trump anyone in an arguement as to who is better. But to me, and those who use “Sabr”-metric stats like VORP, it does give a value to players in an accurate way, its like one stop shopping. I dont HAVE to look at OPS+, AVG+, ERA+ etc etc etc if i dont want to cause VORP gives me the value.

  10. JW says:

    Amen, JC. Amen.

  11. Joe says:

    I think reason that value over replacement can be more useful is that it’s theoretically easy to find a “replacement level” player, while a team can’t simply go out and find an average player whenever they need to replace someone.

  12. JC says:

    A replacement player is just a below-average player in a certain range. The difference between denominating use replacement versus average is like the difference between dollars and pesos.

  13. Colin Wyers says:

    JC – The reason you use replacement level instead of average is the talent distribution. Baseball talent is not normally distributed, at least not when talking about MLB’s talent pool.

    Realistically, when we’re talking about major leaguers we’re talking about a portion – a very small portion – of the bell curve, say maybe the top 1% or so. Imagine the rightmost part of a bell curve, and now put a vertical line through it seperating a half of a percent from the rest of the curve. That’s the line that we refer to as “average” in Major League Baseball.

    The practical consequence of this is that an average baseball player is a scarce resource – there is a much greater number of below-average players than average – or even average AND above average – baseball players.

    So when you’re looking at scarcity, you have to realize that at or below a certain bar of performance, there is no scarcity. If you have a left fielder with an OPS+ of, say, 70 – ignoring sampling issues for a second, so I should say a left fielder with a true talent OPS of 70 – then that player is absolutely worthless, at any salary amount; the population of players who can play left field, hit better than an OPS+ of 70 and are available for the league minimum is so great that, if they wanted to, every team in the majors could find a guy like that, and many clubs would still have a guy like that in AAA as a backup plan.

    Now, I’m not the biggest fan of VORP – it completely ignores defense, which is an important part of a player’s value; it uses Runs Created, which isn’t my favorite run estimator by a longshot, and it uses a replacement level well below historic norms. But the concept of replacement level is extremely important, because it captures the very real value there is to being an average ballplayer – average ballplayers are relatively scarce – which using average as your baseline fails to do.

  14. JC says:

    The distribution of talent doesn’t matter. (Though this is a debatable point. After all, many statistics textbooks use baseball batting averages to explain the normal distribution.) If the talent is far below average then the scarcity of players is higher above the average than below. We can account for this without bringing replacement players into the picture. The choice of denominator doesn’t affect our assumptions about the talent distribution.

  15. Colin Wyers says:

    The normal distribution of batting average is a function of the distribution of playing time – it works when you’re simply look at at-bats. At-bats aren’t evenly distributed, however – the better players get more at-bats.

    The reason that replacement level is important is because not all positions are equally valuable. A league average shortstop and a league average second baseman are not the same in value. According to Baseball Reference, this season in the NL the average shorstop hits .271/.331/.396, or an OPS+ of 96. while the average second baseman hits .265/.333/.412, or an OPS+ of 101. They’re nearly identical as hitters – a little bit more pop for our second basemen.

    They’re not equally valuable, though! Our average shorstop can move over to second base and, when you factor in his above-average defense for the position, be an above-average second baseman. Our average second baseman, moved over to shortstop, is an inferior defender at the position – that’s, after all, why he’s not a shortstop; almost all second basemen in the majors were shortstops at some point – and thus a below average shortstop. The average shortstop is worth more than the average second baseman.

    This is the underlying reality that replacement level is trying to model. I guess it’s possible that you can construct a model that doesn’t explicitly use replacement level and still manages to get the positional comparison correct. If you have such a model, I’d love to hear how it works.

  16. JC says:

    The reason that replacement level is important is because not all positions are equally valuable.

    Does anyone think that positions are equally valuable? I can make comparisons across positions by comparing players to the mean at different positions.

  17. Colin Wyers says:

    Does anyone think that positions are equally valuable? I can make comparisons across positions by comparing players to the mean at different positions.

    Right, and those comparisons end up doing things like underrating shortstops in comparison to second basemen – an average shortstop is a better baseball player than an average second baseman. An average shortstop is a better second baseman than the average second baseman.

    Or look at the American League this season – the average 1B hits .256/.341/.412, OPS+ 104. The average DH hits .248/.337/.420, OPS+ 105. The average LF hits .264/.337/.422, OPS+ 106. The means are nearly identical – any difference is pretty much splitting hairs. But the left fielder is more valuable than the first baseman, and the first baseman is more valuable than the DH – the average left-fielder is also an average-or-better first baseman and an average DH, and the average first baseman is also an average DH. The average DH is a below-average 1B or LF, though – that’s why he’s a DH.

  18. JC says:

    Right, and those comparisons end up doing things like underrating shortstops in comparison to second basemen – an average shortstop is a better baseball player than an average second baseman. An average shortstop is a better second baseman than the average second baseman.

    If we know that the average shortstop is more valuable than the average second baseman, then I don’t see where the problem is. We know that shortstop gets more defensive opportunities on balls in play than any other position. Naturally, your best infielder (with some caveats) will play this position. What does this have to do with using replacement level as a benchmark?

    My own system for valuing players gives value to different positions based on defensive position. You don’t need the replacement level concept to make this correction. That’s my point.

  19. Matthew Knight says:

    JC,

    You’re kind of missing the point of the column. It’s a weekly column which introduces more advanced thinking in baseball to Washington Post readers, not a one-off column convincing you to include VORP in your sabermetric toolbox. Last week the article was about why it makes financial sense for both the team and player to lock up Hanley Ramirez or Evan Longoria for a long term deal when they have minimal major league experience. If my memory serves, other topics have included various ways of measuring defense and the break even point required for stolen bases to be valuable.

    While you might disagree with a particular metric it espouses, I think it is a very valuable column for those of use who like thinking about more than just the standard HR, RBI, AVG statistics and wish that more people would look beyond those numbers when considering baseball players. My uncle, for example, has been a huge baseball fan his whole life. To him, the only statistics that matter are the ones that were on the back of a Topps card in the 50′s. But he reads that column every week and learns that there are other ways of valuing players than just those stats, and that there is more to executive decisions than what is generally reported. Whether or not you like the methods in the article, I think you should be praising David Sheinin and the Washington Post for publishing these “new” ideas on a weekly basis. Think of all the new sabernomicists that may result!

  20. Ken Houghton says:

    A catcher needs to be able to throw 121 feet accurately. A LF needs to be able to throw 200′+.

    The reason the Biggios and Murphys tend to get moved is that, if you have that much offensive production, you want to have it for more years–catchers, as you note, are involved in almost every play, and crouch for every non-IBB pitch. (Piazza stayed at catcher too long–he was lousy at holding on and/or throwing out baserunners, for instance–and it probably cost him some offensive productivity.)

    So, yes, I would assume a scarcity of catchers over LFers, because the difference in arm strength needed is negligible, and more than outweighed by the additional effort needed to perform at the position.

    I may be beginning to understnad part of your objection to VORP, though. If I’ve got a player who is hitting .245 where the average is .260 and the Replacement Player would be .230, they’ll look as if they have a positive value, even though they are below average. (That is, they are not just “taking up a spot on the bench.”)

    Now, you may opt at that point to replace the subpar player with someone at another position–pro baseball is not rotisserie. And your 3rd catcher may contribute less than your 8th OFer, if the first is never injured and there are few day-after-night games. But that’s a substitution effect, not necessarily a supply limitation, no?

  21. Sal Paradise says:

    ‘Average’ doesn’t take into account the real impact that a player has on a team.

    For instance, take the Phillies.

    Chase Utley is a career .301/.377/.506 hitter.

    His replacement is Eric Brunlett, at .249/.322/.358.

    As stated above, the average 2B is .265/.333/.412 — significantly better than Brunlett. When we talk about value over replacement player, we acknowledge that most teams don’t have an average player sitting on the bench. We look at each player’s contribution over what the person who replaces him is actually more likely to provide.

    And that level is the replacement player. Sure, we have exceptions (like Eric Hinske this year), but for the large part — especially for the more defensively demanding positions like catcher/SS — the replacement player is utterly miserable.

    If you peruse the depth charts, you’ll notice that there are really very few ‘average’ players sitting around on benches for the higher skilled positions, and that ‘replacement player’ is a far better comparison point for lots of analysis regarding injuries and team depth than ‘average’ is.

    I wouldn’t bring it up at a cocktail party, but if you’re talking about what the Braves would do if Chipper Jones (.310/.406/.550 career) went down. Comparing him to an average .270/.340/.450 isn’t really the best comparison when his replacement is Omar Infante (.253/.300/.386).

    That’s where VORP comes in handy.

    The people at the very top pull those averages way up. The actual replacements are far far worse.

  22. “Does the world produce fewer catchers than left fielders?”
    Yes. Bill James had this to say about the defensive spectrum:

    “Players can generally move from right
    to left along the spectrum successfully during their careers.” -Bill James

    The Spectrum looks like:
    1B – LF – RF – 3B – CF – 2B – SS – C

    http://www.baseball1.com/bb-data/bbd-bj1.html

    It’s not that the world produces fewer catchers than left fielders, it’s that most everybody on the diamond already is a left fielder if they need to be. Only a small fraction are catchers.

    Also, if this were an article in Baseball Prospectus or The Hardball Times we might have something to write home about.

    Sure, VORP is a bit outdated these days. In a perfect world we’d see wOBA, EqA, or linear weights used in the newspaper every day. Unfortunately this is a world where the majority of fans still use archaic counting stats like RBIs and hits, and Joe Morgan, John Kruk, and Steve Phillips still have jobs with the leading broadcasting network.

    I can live with VORP.

  23. JC says:

    Catchers are an anomaly on the spectrum. Catchers can’t just move over the SS. Again, there is nothing special about VORP in integrating the importance of position concept into evaluating players.

    This has nothing to do with where the article was published. I think VORP is useless unnecessary. If you are just getting in to analyzing baseball, I see no gains to becoming familiar with the concept. Don’t bother.

    I don’t use wOBP or EqA either. Linear weights predates VORP. In a world where fans use triple-crown stats to analyze baseball you are more likely to explain to friends why AVG isn’t as informative as OBP and SLG (hey, you can add them together to get OPS!) than by indroducing VORP.

  24. jfalk says:

    Put me down as another JS fan (and economist) who seems confused by your stand here. If the choice between average and marginal as a base just a numeraire problem, then I would agree with you… but you’re assuming a fact not in evidence. I grant that VORP requires you to answer a question which doesn’t have to be answered otherwise — the definition of a marginal player. But it is certainly possible at any given time for, at any position, the average player to be worth much more, or only somewhat more, valuable than the marginal player, particularly since we are dealing with the tail of a distribution. The advantage of the marginal player is that we know what they cost — the league minimum. Further, they are, by definition, freely available at that cost. You then get the problem of a player who is the best say, catcher in baseball who is relatively not as much better than a replacement catcher as the best shortstop is than a replacement shortstop.
    I grant your Klingon point, but it seems to me there’s potentially extra information there to the extent that the distribution of talents has higher variance in one position than another. Or not.

  25. JC says:

    The advantage of the marginal player is that we know what they cost—- the league minimum. Further, they are, by definition, freely available at that cost. You then get the problem of a player who is the best say, catcher in baseball who is relatively not as much better than a replacement catcher as the best shortstop is than a replacement shortstop.

    This is exactly what is wrong with VORP. Marginal players are not worth league minimum, they should be valued equal to their marginal revenue products.

  26. Watercott says:

    JC, You’re being pretty silly here. You say you can adjust for era and park. You say you can assign more value to players of certain positions. But then, you dismiss VORP, which does this, and instead say you want to rely on OPS and SLG? Why not use the metric that does everything you want it to?

    As for replacement level – First marginal players are worth exactly league minimum – that is the whole point of the exercise of using replacement level. They have NO marginal revenue production, because there is always another player who is available for league minimum with the same production (at least in theory).

  27. JC says:

    From the post.

    You can make the same corrections to any baseball metric. The simplest is OPS+, but you can make more precise adjustments for any offensive statistic as Michael Schell has done. To me, that is superior to VORP, because I know what each of those statistics is telling me.

    What about using VORP to judge salaries? For example, if the league-minimum salary is $390,000, and a team signs a replacement level outfielder for $1 million, hasn’t the team overpaid? Not at all. Player value is determined by opportunity cost as determined by marginal revenue product (MRP). If a player generates many millions of dollars, his value is determined by this, not by how much he makes. Teams pay players with less than four years of service (approximately) less than their MRPs because the collective bargaining agreement allows them to do so. A team that plays a young and reserved player forgoes the potential return from trading the player to another team or from keeping down his service time. Signing a veteran for $1 million can be cheaper than promoting a young player who would provide equal value now by holding down service time.

  28. Watercott says:

    Sure, it’s possible that it doesn’t make sense to call up some of your replacement players because of playing time, etc. However, again, the definition of replacement level here is that there is an almost limitless supply of players at that level, some of whom it would make sense to pay the minimum, e.g. career minor leagers, the players already on your bench, players for whom you don’t care about playing time, etc.

  29. JC says:

    The limitless supply of players with MRPs of 390K assumption is just wrong. The opportunity cost of playing a players is the value you could get from shipping him to another team. Or, if you want to acquire a player from another organization you have to give up more than the salary you pay him.

  30. jfalk says:

    Two things: fisrt, marginal players (at least those without tenure) do not need to be paid their MRP. The theorem that equates wage to MRP assumes that a player paid less than MRP can find a competitor who will pay up to MRP. Obviously inapplicable here.
    Second: try never to quote Murray Chass. he’s an innumerate idiot… when he’s on my side of an argument, I usually suspect I’m wrong.

  31. Watercott says:

    I am convinced we’re not speaking the same language here, but I’m happy to keep having the conversation in the hopes that I can perhaps learn something.

    If we assume that everyone is a perfect evaluator of baseball talent, then the opportunity cost of playing a player is exactly the same as the value that he will provide your own team, so unless your organization is taking advantage of a dimwit, the worth of the player is still only what he is able to produce on the field.

    For a replacement level player, that is exactly 390k, because, even if a particular player in your organization would cost more than this because of playing time concerns, etc., there is some player somewhere, who can be acquired or brought up, that will provide replacement level play for that price.

  32. JC says:

    marginal players (at least those without tenure) do not need to be paid their MRP.

    No, but they do need to be valued as such.

  33. Nice discussion. I’ve never quite gotten VORP either, particularly since they make the point of adjusting it by position, but then don’t account for defense at all in the metric, which is definitely part of the value of that ballplayer.

    Seems like it should be OVORP instead, so that it doesn’t confuse the people trying to understand it for the first time or people trying to understand the value of VORP vs. WARP, which does account for defense.

    Also, I’ve always found where they defined replacement level to be very arbitrary, much like the decision that 100 pitches is bad but 99 pitches is not, for PAP. They make it all seem very scientific, but the explanations I’ve seen makes it seem like it was chosen out of the air.

    I think JC’s point is this: whatever adjustment VORP makes to account for a replacement player can be shown just as plainly with any other offensive statistic you chose. He used the example of currency as an example, but I think a better example would be Farenheit vs. Celsius.

    You can think of Farenheit as the system JC is talking about, where the replacement player is the point where water freezes at 32 degrees-F. Celsius is the VORP system, where they adjust the Farenheit so that the the replacement player is now defined as zero degrees celsius, so then anything positive is clearly better than replacement and anything negative is clearly worse.

    So the point for the VORP supporters is that you can see right away what the +/- for a player is relative to this theoretical replacement player who is available for the grabbing off the waiver wire or via a trade (though in reality, that is not the same cost, a waiver wire would entail just the cost of the player, but a trade would require you to give up another player, theoretically of equal value).

    But JC’s point is that you can still discuss the same thing in another way because whether you use the average player or the replacement player as your zero point, the main thing is that you can still see the difference between a replacement player, an average player, and star player. Why is it so important to measure against the replacement level player when you can still see the relative difference, relative value between these three types of players?

  34. Replacement level is such a fundamental concept in baseball economics.

    Please tell me that I shouldn’t expect this kind of pointless obstinacy in your book, which I still haven’t read.

  35. Sal Paradise says:

    JC, out of curiosity, if I were to ask you, “If Chipper Jones goes down with an injury for the rest of the season, how many games can the Braves expect to lose that they would have won with Chipper?” how exactly would you go about answering that?

    I entirely agree with you that the replacement player is not equal to a league minimum wage, but I disagree that VORP isn’t useful in the context of many (real) baseball situations.

  36. JC’s point, and he’s free to correct me if I’m wrong, is that replacement value is sort of an arbitrary scale where the actual value of the number being measured is no different than if it was presented on a “league average” scale.

    And because “replacement level” as a number is inherently arbitrary (while “league average” is not) it’s less useful than simply putting it in the “league average” scale.

    It’s not a terrible argument, particularly for players with similar playing time. I’m not sure I agree with it either. Where I believe “replacement level” gains value is when you start to compare players with vastly differing amounts of playing time (like Hall of Fame arguments).

    Even then you’re still a bit on “favorite flavor of Ice Cream” territory as there’s plenty of reasonable arguments to using “league average” for Hall of Fame purposes.

    I like replacement value as a theoretical concept. Something that I think you need to understand when discussing player “value” in an abstract sense. Practically, you’ll almost never compare a player you’re thinking of replacing with a theoretical “replacement player,” the potential replacements will all have names.

  37. DJ Tanner says:

    JC,

    I think it is a rather flimsy argument to dismiss a very useful metric b/c you feel that it is esoteric. An explanation of what VORP measures does not have to be complicated. If you feel like quoting that particular stat in a conversation with the “casual fan” all they truly need to know is that: It measures how many additional runs a player contributes to his team (prevents), relative to freely available talent. That’s pretty simple. The person understands the meanings and its usefulness, all in one sentence.

    I too like OPS+/ERA+, but statistics do not have to be mutually exclusive. MLB teams, and fans alike, look at very complicated advanced statistics everyday — statistics that personnel in other front offices probably have trouble understanding (i’m looking at you Brian Sabean) — but that has zero impact on their usefulness.

    My point: Ok, you don’t like VORP…but, jeez, man…not everyone who tries to quote the thing is a Joyce reading, NPR listening, uptight douche with an inferiority complex. Now, FRAA…that’s a differnt story.

  38. JC says:

    Sal,

    I would replace Chipper’s expected performance with Infante’s. You could use VORPTM to make such a measurement. I don’t have a problem with this. But, I could make this same comparison if VORP never even existed.

    My point isn’t that VORP is an awful or useless stat. To the contrary, there is clearly useful information contained in it. And those who prefer to hold discussions based on this metric should continue to do so. But there is no need for someone who does not speak the language to learn the ins an outs of a new metric, as Sheinin suggests.

    What I am bothered by is this clubish (dare I say cultish) use of VORP, which is offputting to those who are capable of understanding similar concepts with stats they can understand. I knew what OBP and SLG were when I played Little League. I understand what league average is. Now imagine you are a columnist who writes about what a good hire Dusty Baker was for the Reds, and you flamed by irate stat-heads claiming, “Dude he played Corey Patterson who has a career VORP of – 27!” I’m just going to delete that. If you say, “Dusty batted a player leadoff with an OBP 100-points below league average,” then I’m more inclined to pay attention.

    I address every single one of these concerns discussed in the comments in my analysis of baseball. I don’t use VORP. I don’t even know where to find it on the web. And the general response I have received in this thread (see #34, for example) confirms by suspicion that using VORP is more about signaling behavior than anything.

    If you use it, I have no problem. Just don’t expect others to use it as well. Engage, maximum VORP.

  39. Watercott says:

    If VORP didn’t exist, though, how would you translates Jones’ and Infante’s OPS+ into runs/wins? And if you have to perform that translation, why not just use VORP, which already did it for you?

    I don’t actually like VORP all that much either, and replacement level really is only useful in certain circumstances, but if you really want something that looks and smells and sounds like a duck, just get a duck.

  40. jpwf13 says:

    I really don’t understand JC’s position, it’s almost like he;’s been deliberately contrary.

    I think VORP is flawed (and misnamed), but highly useful.
    1: It should be “runs over bench” or some such thing rather than “value over replacement player”
    2: BPro’s replacement level is too low.

    What I find useful about VORP is that it’s a combined rate stat and counting stat.
    A: A SS who generates 100 runs is more valuable than a 1B who generates 100 runs. The average SS may produce 60 (bench 40), with 1Bs, the average may produce 80, and you can find any number of lead footed AAAA slugs who could put up 65-70.

    B: Hold a mock draft, 2 teams, one drafts by raw runs (linear weights/baseruns/erp/rc, it won’t matter- the other drafts by vorp, until each team has 9 players-
    The vorp team will have more total offense by the end.

  41. Jason W says:

    JC,

    I think you’re making the assumption that there are only two types of baseball fans out there: “dumb” fans who think average, RBIs, and pitcher wins are the only stats that matter; and everyone else, who “really” knows what’s going on. In particular:

    If you are just getting in to analyzing baseball, I see no gains to becoming familiar with the concept. Don’t bother.

    I consider myself reasonably smart, stat-wise. I understand what VORP is (even if I don’t know how to computer it). I have only the faintest clue what, say, Eqa is, so I don’t use it in my discussions. Similarly, there are things in your blog that I don’t “get,” because, well, you’re an economist with a far better education than I. That doesn’t mean I think it’s useless or you think you’re “cool” because you can go off about marginal value or PrOPS (which I’d never heard of until your Chipper Jones post). And it seems odd that you’d have a beef with people talking about VORP being too esoteric for “regular folks,” yet you use far, far more complex metrics in your writings.

    Saying that “there’s no reason for someone to learn about VORP” is like saying you should go straight from using ERA to using xFIP to determine pitchers’ values, skipping right over “intermediate” mechanics like WHIP and opponents’ OPS.

    Jason
    (Level 6 Half-Elf Stat-Head)

  42. cc says:

    marginal players (at least those without tenure) do not need to be paid their MRP.

    No, but they do need to be valued as such.

    Value needs to take into account opportunity cost. Because the supply of freely available minor league players (by which I mean minor league FAs, Rule V draftees, AAA lifers granted their release when another team offers them a spot in the majors – not prospects) exceeds demand, their salary is deflated below its “value,” here pushed down to the MLB minimum salary.

    Thus, if you can get a AAA lifer who produces, say, $5M worth of raw value for $390K, there is a significant opportunity cost to signing a FA instead. So:

    Net value
    = raw value – opportunity cost
    = (FA runs * $/run) – ((AAA runs * $/run) – $390K)
    = $390K + ((FA runs – AAA runs) * $/run)
    = $390K + (VORP * $/run)

    Yes, you can write the same equation using a different baseline, but your constant will be derived from the major league minimum salary ($390K) and the production you can acquire at that price point.

    Replacement level is also the only baseline that does not require that playing time be held constant. Using average as a baseline undervalues PT (since average production has value), while using zero as a baseline overvalues PT (a full season of horrific production hurts, not helps, a team). VORP normalizes for playing time so that 20 VORP from a part-time stud is worth as much as 20 VORP from a full-time average player, which is a substantial advantage as a metric.

    That’s not to say that VORP is perfect. Defining replacement level is an art, not a science, and the assumption that there is always a replacement-level alternative freely available may not always hold. But it certainly has an important place in baseball economics and sabermetrics, and I really don’t think arguments to the contrary hold water.