FanGraphs Baseball


RSS feed for comments on this post.

  1. I agree with neutralizing for park effects but does ‘run environment’ also account for the quality of pitching/hitting of the teams that play in that park? It seems to me that you could drive the value of a run down by having a better lineup, no?

    Comment by leclair — September 8, 2010 @ 5:06 pm

  2. So then where does Votto’s .436 wOBA place him? Great American is also a good hitters park, albeit not Coors Field.

    Comment by Chris — September 8, 2010 @ 5:12 pm

  3. What does that imply about the value of saving a run in Coors (as a CF, say)? My gut instinct is that saving a run in Coors is less valuable than it is elsewhere, which I’m not sure makes sense…

    Comment by Zach — September 8, 2010 @ 5:18 pm

  4. I strongly disagree with this article. Your implication seems to be that you think the MVP should go to the player who contributed the most wins to his team, meaning that context needs to be taken into account, but in fact I think the converse is true: the MVP should go to the best player, and context should be neutralized for.

    Saying Carlos Gonzalez’s offense is worth less on the Rockies than on the Padres doesn’t change anything about how impressive his campaign is. If he were on the Padres, and had the same season (accounting for park factors), he would be more valuable by your reasoning. The MVP is an individual award and there has been a constant and justified trend in the sabr community towards neutralizing for factors outside the individual’s control. I think the point you reference is clever, but you could write the exact same article except, instead of two teams with different ballpark factors, using two teams on opposite ends of the spectrum in terms of the quality of their pitching. Should a hitter get less MVP credit for his campaign because his pitchers are all terrible, meaning his team needs to score more runs to win, making every run of offense he contributes less valuable? I say no, and I think you would agree. The parallel to your point is direct and unavoidable.

    Your metric is actually kind of similar to using pitcher wins to decide the Cy Young. You’re basically grading a hitter’s production based on something out of his control. With pitcher wins, the pitcher is dependent on run support – the more runs he gets, the better his performance looks in terms of W-L record. With what we’re talking about here, the hitter is dependent on ballpark effects (or the very close comparison, pitcher run prevention “support”), and the better his ballpark or his pitchers are at run prevention, the better his performance looks in your terms.

    Comment by Oscar — September 8, 2010 @ 5:27 pm

  5. I don’t think you could have possibly missed the point any more than you did.

    If Carlos Gonzalez had the exact same statline in Petco Park, he’d be the MVP because that would be freaking amazing. The argument has nothing to do with pitcher wins or team context. The point is that the same event does not have the same value in all parks, because of the way offense is differently promoted in different environments.

    Just think of it as inflation. $1 in 1950 is worth more than $1 today. Both are the same unit, but the value is different because it gets you less stuff now than it used to. Carlos Gonzalez’s HRs get the Rockies “less stuff” than Adrian Gonzalez’s HRs get the Padres.

    Comment by Dave Cameron — September 8, 2010 @ 5:28 pm

  6. We control for home team roster construction when evaluating the run environment of parks.

    Comment by Dave Cameron — September 8, 2010 @ 5:29 pm

  7. MVP Candidate yes, however, when I do my list of MVP watch I have him outside the top ten. So is he good? No doubt, 5 WAR is 5 WAR and from a 24 yr old!? Praise the lord! But MVP!? No.

    Comment by My echo and bunnymen — September 8, 2010 @ 5:37 pm

  8. I understood the point. Respectfully, you missed mine, but I’m trying to put it a less crooked way. When I dump a complicated thought into a few paragraphs without editing, it can get jumbled.

    I will have a response for you in a minute, and I have a real concern with the article. But first, a question:

    Is the wOBA you use in the article park-adjusted or not?

    Comment by Oscar — September 8, 2010 @ 5:53 pm

  9. Oscar, I read it the same way you did. The part I tripped up on this article was when dave started talking about winning games. There’s clearly a park adjustment to be made to cargos line, no one would argue otherwise, but it’s nothing to do with winning.

    Im on the subway so I’ll let you tackle it, but I’ll chime in later

    Comment by The Duder — September 8, 2010 @ 6:04 pm

  10. Hm, I’m kind of thinking out loud here, but since both the home and away teams are affected equally by the Coors effect, shouldn’t the run differential needed to win a game be unaffected by park? If that’s true, why do more than the standard park adjustment to put the batting line in context?

    Comment by JH — September 8, 2010 @ 6:15 pm

  11. I mean, Dave’s main point is very simple – one run is worth less when it takes more runs on average to win a game. To take an extreme example, a run in baseball is worth more than a point in basketball.

    But I don’t understand the point of the article. The higher run environment of Coors is inseparable from the higher batting stats of the hitters at Coors. So if it takes 5 runs to win a game in Coors and 4 in a neutral park, hitters’ batting stats will be proportionally higher at Coors than at Petco, and those runs those hitters generate then get proportionally devalued through Dave’s method.

    The whole thing seems redundant; at best, an alternate method of adjusting for park factors. Either you take a hitter’s unadjusted wOBA and devalue the runs it produces because of the higher run-scoring environment, or you simply park-adjust the wOBA first.

    Comment by Oscar — September 8, 2010 @ 6:21 pm

  12. I think Oscar is on point here. If placing numbers within context of a park is important when thinking about MVP votes because one hitter’s numbers in a hitter friendly park do less to contribute to a win, then placing them within context of the hitter’s pitching staff makes just as much sense. The only difference between the two factors (park vs. pitching staff) is that pitching staff will fluctuate wildly over the course of a few decades and the park will probably play relatively the same.

    But if a HR at Coors isn’t as valuable as a HR at Petco, then a HR hit by a player whose pitching staff is horrendous is worth less than a HR hit by a player backed by the best pitching staff of all time. And I don’t think it makes sense to punish the former hitter while rewarding the latter. I guess Dave’s counterargument is that we can’t equate the very static and fixed nature of the park to the less-easily defined nature of a pitching staff, but I don’t see a problem with it in the context of this article. WAR is WAR and I get that, but talking in depth about how C. Gonzalez’s HRs are “less valuable” (at least at home) then A. Gonzalez’s allows for an extension into Oscar’s point as well.

    Comment by LD303 — September 8, 2010 @ 6:24 pm

  13. I’d be interested in hearing about these 10 guys ahead of him.

    Comment by James — September 8, 2010 @ 6:34 pm

  14. My echo and bunnymen…id like to see the 10 players you have above Cargo, cause I cant think of 5. Keep in mind, MVP doesnt always mean the best player in the league, but the most valuable. Without Gonzalez, the Rockies are arguably a below .500 team. Hitting .387 in the second half and leading your team into a division race would argue that he deserves credit for MVP.

    Also….Dave, how would you go about Ubaldo’s stats (even though hes been struggling lately) Is there a ‘curve’ for him at all?

    Comment by Cargo for MVP — September 8, 2010 @ 6:39 pm

  15. All he did was park-adjust the runs, guys. That’s why a ..432 wOBA in Colorado is worth 39.2 runs and a .388 wOBA in San Diego is worth 37.4 runs. That’s the park adjustment. It’s much harder to hit in San Diego than in Colorado, so if you put two equal guys in those different parks, the guy in Colorado is going to have better stats. Forget about the pitching staffs, that doesn’t matter. When you park-adjust their numbers, the two guys come out virtually even.

    Comment by Nathaniel Dawson — September 8, 2010 @ 6:48 pm

  16. Ironically, I think that’s what Dave thought I meant when he said I completely missed the point, and in fact you have missed it, but in fact what this whole article comes down to is a second way to park-adjust hitters.

    Comment by Oscar — September 8, 2010 @ 6:55 pm

  17. I figured as much, thanks!

    Comment by leclair — September 8, 2010 @ 7:37 pm

  18. Dave, I don’t think people agree with the concept of standardizing at all. Rather, I think the confusion comes because you’re applying the standardization at the run level, which is not intuitive and not really in line with the way you explained it.

    CarGo has produced significantly more runs than Adrian Gonzalez. That cannot be disputed. When you display 39.2 and 37.4, this reality is obfuscated.

    But, as you point out, when looking at MVP or when trying to assess true talent, we need to neutralize the context. That’s when we need to do the conversion and apply the park factors. The valuation calculus happens when you convert runs produced in to wins, not when you convert offensive production to runs. I think a lot of casual fans are not only confused but get angry when Sabermatricians say clumsy things like “Carlos Gonzalez and Adrian Gonzalez have been equal offensively”.

    In a sense, it’s true. They’ve helped their teams win games due to their offensive production in equal amounts. But while sabermatrician understands the steps it takes to get from the basic stats to that conclusion, the casual fan doesn’t. The casual fan is thinking, heck no, CarGo clearly produced more runs. The sabermatrician knows this, but jumps straight in to the value adjustment. This happens all of the time and is almost always where the conversation breaks down. The casual fan often agrees 100% with the conclusion when they are walked through it, but when they aren’t taken through it, they give up.

    WAR is great, but more of an effort needs to be made to explain the conceptual framework. When we start producing context adjusted stats without clearly explaining why and how, the non-initiated among us simply tunes out and thinks were monkeying with the data to match our pre-existing conclusions. If we want to be convincing in these discussions, we need to do a better job at explaining ourselves, clearly describing what we did to the numbers and why.

    Comment by RMR — September 8, 2010 @ 7:40 pm

  19. If he hit .387 but his teammates performed even worse and the Rockies still weren’t in contention, would that make him less valuable? If the Rockies were already in contention and he hit .387, would that make him less valuable?

    If so, aren’t you basically saying that a player’s claim on the MVP is dependent on how well his teammates play?

    From this perspective, hitting .387 is hitting .387, but it’s still only a fractional contribution to the team outcomes. Whether or not his teammates played well enough to make that .387 matter is beyond his control – why should a player specific award be based on things outside of their control?

    I’ll never understand why people think they need to go beyond the player’s performance and discuss it’s contribution to team success to that performance. Do people really think the award is meant to recognize the guy who played really well while his teammates were good enough to put the team in the playoffs but not good enough to get there without him? It’s such a convoluted concept.

    For me, it’s simple: which guy did the most to help his team win games – the end. And hitting really well in a park that makes it easy to hit well means that doing so probably isn’t as valuable from a win contribution perspective.

    Comment by RMR — September 8, 2010 @ 7:50 pm

  20. Dave has explained it here:

    Scroll down to the bottom, he posted a series of articles detailing the steps moving from a players’ basic stats to wins and valuation. For anyone who hasn’t read this series, I highly recommend it.

    He can’t keep re-iterating these things in every article he writes. At some point, readers and commentors have to take it upon themselves to research and understand the tools and concepts that are used on this site.

    Comment by Nathaniel Dawson — September 8, 2010 @ 7:54 pm

  21. “I don’t think you could have possibly missed the point any more than you did.”

    Maintaining the title for Most Smug Writer on the Web, I see.

    Comment by Wells — September 8, 2010 @ 7:55 pm

  22. Mmm…I’m not sure I even agree with the assumption that a hitter should be evaluated in this way, for the purposes of choosing an MVP. There is a reasonable, and I think convincing, argument that a hitter should be judged based on their production, and not on their value added. For example, if I had to choose blind between voting for the WAR leader and the WPA leader as my MVP choice, I’d choose WAR. I’d rather pick the best hitter than the one who contributed to more wins – and that’s not always the same thing.

    Also, I don’t think the objections have reflected a lack of understanding of the process. There are some really interesting questions about what Dave’s talking about here, and I found your post to be dismissive of that. If you carefully read my original comment, the one that Dave said showed I had completely missed the point, I’m sure the actual point will come through.

    Comment by Oscar — September 8, 2010 @ 7:58 pm

  23. Exactly! Haha, this makes my response below obsolete.

    Can you explain the advantages of Dave’s approach over simply park-adjusting Gonzalez’s stats like we usually do? I don’t see any.

    Comment by Oscar — September 8, 2010 @ 8:00 pm

  24. By your logic Oscar, Adrian Gonzalez is still more valuable, but Carlos Gonzalez gets a leg up on Votto based on the fact that the Rockies have the 3rd best FIP and 4th best xFIP. Not to mention that if you took off almost half a run for everyone of Ubaldo’s starts (about the difference between Citizen’s Bank Park and Coors Field in runs based on ESPN’s park factors) and I’m sure that Ubaldo is a minimum top 3 Cy Young if not edging out Halladay.

    Comment by Chris — September 8, 2010 @ 8:06 pm

  25. and uh, who exactly are these 10 people? (or more, since you didnt say he was #11)

    Comment by fredsbank — September 8, 2010 @ 8:06 pm

  26. And at some point, the longtime readers of FG need to accept the fact that the pool of commenters no longer consists of nothing but a big group of Dave fanboys. New people are passing through at an escalating rate- there are going to be people asking questions that have been answered before. It’s a good thing, not a bad thing.

    Comment by LD303 — September 8, 2010 @ 8:27 pm

  27. LD, that’s why I linked to that series of articles. As new people ask questions, it’s a good thing to point them to something that can answer them.

    Comment by Nathaniel Dawson — September 8, 2010 @ 8:45 pm

  28. No one is arguing that park adjustments are unnecessary, right? We all understand that, in order to “equate” a hitter’s performance, we need to park adjust, correct? For the record, FanGraphs WAR takes into account park-adjusted batting runs.

    Oscar mentioned that, at best, what Dave mentions in his article is the same as park-adjusting your run total. Indeed, they’re both the same. The point of park-adjusting is that you account for the fact that events are worth more or less than the average in a given run environment. Here, Dave describes it in terms of wins, which is correct. The way it is taken care of in WAR is by park-adjusting the runs, which is also correct. The concept is the same, even if the explanation is different.

    Remember, runs and wins are the same (in that they have a conversion rate). Devaluing the value of an event in terms of runs is no different than doing so in terms of wins. The question now is whether or not those who disagree with Dave still feel park-adjustment is important.

    For those mentioning the quality of team pitching counter-example: I think the logic behind a park adjustment is that parks are mostly independent of individual performance. That is, a park adjustment takes into account something that is out of control of the players (a park naturally depresses or increases run environments). On the other hand, quality of team pitching is like RBI/pitcher wins in that it is dependent on teammate performance.

    Comment by Michael — September 8, 2010 @ 8:48 pm

  29. As per my response: there is no difference. Perhaps Dave worded it in a confusing fashion, but essentially the method done on FanGraphs is the same as the one Dave said (hence why Dave quoted FanGraphs park-adjusted batting numbers in support of his point).

    If you do it either way, what essentially happens is you change the value of an event based on the run environment. Dave essentially describes changing the win value of events (as a whole) in the article, whereas FanGraphs changes the run value of events via run park factors. It’s all the same, since runs can be directly converted to wins.

    Comment by Michael — September 8, 2010 @ 8:51 pm

  30. I see what you’re saying about “to find the league’s most valuable player, find the player that provided the most wins to his team”

    I guess though, it’s not too much of a stretch think that not all wins are created equal and that some are more valuable than others, which then brings us back to the original point of helping to win more important games being a factor in assessing the MVP…

    Comment by Resolution — September 8, 2010 @ 9:17 pm

  31. I understand the process, Nathaniel. I don’t need to reread the series.

    I completely understand the need to park adjust the run values. My point is that by park adjusting the run value before reporting it out, you make it more difficult for a person initially trying to understand the full formula what’s going on.

    My contention is that, in an article describing player value, it is more clear to talk straight run value, then an adjusted one, then convert to wins. Or alternately, use a process which better matches the lay description and use a different runs/win conversation factor based on the run environment in each park.

    All the math is, in the end, the same. But by grouping the calculations differently it is easier to explain what is being done and why.

    Comment by Rick — September 8, 2010 @ 9:40 pm

  32. But none the less gives you a basic value of what a run will be worth. Using team FIP would be a way of ‘neutralizing’ individual performances. After all, we’re talking about the value of a run to a team to determine MVP status. You have to account for your pitching if that is the case. A pitching staff with a 4.34 FIP is obviously going to inflate the value of whatever runs it’s offense scores. Also, if you go that route you must look at team runs. The Reds have the most runs in the NL, the Pirates have the fewest. Therefor, any runs produced by an individual on the Reds would have less value because some one else would likely step up and score it for you, however, any time that Neil Walker can drive in a run he’s producing a much more significant portion of his team’s total offense.

    Scoring runs is a team effort and you can’t ‘factor out’ the team itself. If no one gets on base ahead of you, you wont drive in runs. If no one gets hits after you, you wont score runs. You could spend a whole season batting 1.000 but if you can’t score or drive in runs then you wont win the MVP because your run values are based only on your park factors and not team pitching or team offense?

    I’m either incredibly off track with this or people aren’t seeing the can of worms opened up by this post and I’m pretty sure I’m on to something here.

    Comment by Chris — September 8, 2010 @ 9:42 pm

  33. Is the inference then that if Carlos and Adrian switched teams their numbers would essentially flip-flop? I don’t know if I buy that.

    Comment by David G — September 8, 2010 @ 9:52 pm

  34. You are exactly on point, and you are raising the same points that bothered me. There’s no easy answer, but I’m interested to see what Dave thinks.

    Comment by Oscar — September 8, 2010 @ 9:53 pm

  35. Youre argument makes no sense….Noone has control over what their teammates do. All you can do is perform…and thats exactly what hes done….You say its simple….which guy did the most to help his team win games….well…3 HR’s off of the triple crown? Id say hes done more than anyone else in the league.

    Comment by Cargo for MVP — September 8, 2010 @ 10:02 pm

  36. I think you’re being a little bit misleading with this article. I’m not sure anyone believes Gonzalez is truly the most valuable offensive player in the National League, but maybe that he’s top three.

    Even with Fangraphs’ park-adjusted batting RAR, he is 3rd in the NL, which is REALLY FREAKING GOOD. That deserves to be mentioned.

    Comment by Andrew T. Fisher — September 8, 2010 @ 10:17 pm

  37. That’s why he said he’s an MVP candidate

    Comment by Nathaniel Dawson — September 8, 2010 @ 10:24 pm

  38. That’s why I said a little bit. But when the final sentence is about how his raw performance has to be so much better than everyone to have the same value, excluding the fact that he actually does would suggest to most that he does not. I frankly was surprised to find out that he was 3rd in the NL on my own. It seems like something I would not leave out if I were writing about CarGo.

    Comment by Andrew T. Fisher — September 8, 2010 @ 10:28 pm

  39. Yeah I think he isn’t a top 50 most trade able commodity either, Dave.

    Comment by Sophisticate — September 8, 2010 @ 10:49 pm

  40. So Ubaldo has a 15-1 start with a sub 2 ERA, and the first thing this site says about it is he has been unimpressive. Cargo has two straight 480+ wOBA months and the first thing the site says about it is that it is unimpressive.

    Comment by Sophisticate — September 8, 2010 @ 10:54 pm

  41. As a long time Rox fan all I have to say is if every great offensive season by a Rockies player will be discounted because of Coors Field then Ubaldo’s pitching performance in 2010 should be put on a pedestal.

    Comment by purplerox — September 8, 2010 @ 11:09 pm

  42. “Most valuable” is open to interpretation. I think that is the point of the award, and that’s why they have a vote. Some of us like to believe there is a difference between putting up good stats in meaningless games and putting up good numbers in meaningful games during August and September when there is external pressure from your teammates and fans to succeed.

    Comment by James — September 8, 2010 @ 11:10 pm

  43. Yes, calling him a legitimate MVP candidate is calling him unimpressive.

    Well done, comprehension failure complete.

    Comment by Not David — September 9, 2010 @ 12:10 am

  44. I agree wholeheartedly. Colorado’s hitters are punished by voters and baseball fans in general, while their pitchers (which lead/led MLB in pitching WAR for 2010 AND 2009) are not given enough credit.

    CarGo’s WAR values may not reflect his true value this year, as I do not trust UZR as part of year-to-year WAR, especially for Coors Field. Based on the (somewhat biased) eyeball-test, he is most definitely a plus fielder at all three positions. Flip the sign on his -4.9 UZR, and his overall WAR jumps to ~6.1, up with the top of the league.
    (My apologies if I miscalculated, my in-depth knowledge of WAR is very limited, and I seem to remember that 10 runs =1 win, thus +9.8 runs over his -4.9 UZR (to +4.9) should add a whole win, correct?)

    Comment by hjrrockies — September 9, 2010 @ 12:11 am

  45. Yep again I fail at comprehension, see here I thought I was reading a condescending article that was aimed at best diminishing, and at worst belittling the accomplishments of a this player.

    Of course he is an MVP candidate He is top five in every counting stat short of BB and SB he leads the league in slugging!!! So I should be happy if Dave says he could be in the running for a popularity contest, but really his value is less than the other candidates.

    Comment by Yo Comprendre — September 9, 2010 @ 12:34 am

  46. I chose to go by WAR and also put in pitchers whose WAR was higher

    The 10? Ryan Zimmerman, Joey Votto, Albert Pujols, Andres Torres, Matt Holliday, Adrian Gonzalez, Josh Johnson, Roy Halladay, Ubaldo Jimenez, Adam Wainwright, Brian McCann (because of position).

    I had him 14th at my last MVP Watch which was just a couple days ago, he’s been on quite the streak lately.

    Comment by My echo and bunnymen — September 9, 2010 @ 12:52 am

  47. There is so much confusion in this thread between people it’s hilarious. All dave is doing is adjusting for coors at the run level…for no good reason.

    Comment by The Duder — September 9, 2010 @ 12:55 am

  48. I listed my ten above (it’s actually 11 and it was more but he’s been on fire as of late). Value and best player go hand in hand since being the best player makes you the most VALUABLE player.

    Comment by My echo and bunnymen — September 9, 2010 @ 12:56 am

  49. Also if you wish to make an argument for CarGo over Adrian then I’ll hear that (Wainwright and Jimenez as well) but I’m on the side that values Adrian (and others) more.

    Comment by My echo and bunnymen — September 9, 2010 @ 1:02 am

  50. Right and if you adjust Cargo’s numbers at home using a number that his own excellence helps to create, and then weigh his road numbers against people who don’t have to continually adjust the way altitude affects pitches and fly balls your system is flawed.

    Amazing how guys like Posnanski managed to say almost the exact same thing as Dave without provoking an outburst, while Cameron seems to have a gift for stirring up the masses.

    Comment by Yo Comprendre — September 9, 2010 @ 1:06 am

  51. Problem is that when you adjust the value of a run you have to adjust for all factors, not just park factors.

    Comment by Chris — September 9, 2010 @ 2:20 am

  52. Any writer here would tell you that while WAR is an excellent stat, it is not perfect. At this point in time we can’t take every single factor into consideration, as of now we are doing the best with what we have. It’s not 100% correct but it’s pretty damn good.

    It would be incorrect to say that a player is having a better season because he has .3 more WAR than another. It’s very close and quite honestly, in most situations, you can make an argument for both players. I think Adrian Gonzalez and Carlos Gonzalez are prime examples of this.

    Comment by SF 55 for life — September 9, 2010 @ 3:54 am

  53. at what point can we base conclusions on actual results rather than expected value?

    Comment by cs3 — September 9, 2010 @ 4:33 am

  54. Dave is simply double counting park factor. Am I wrong?

    Why would you adjust raw stats for coors THEN ALSO adjust on the run level? Makes no sense. At least the way dave explained it.

    Comment by Locke — September 9, 2010 @ 8:51 am

  55. “It would be incorrect to say that a player is having a better season because he has .3 more WAR than another. It’s very close and quite honestly, in most situations, you can make an argument for both players.”

    A-freaking-MEN! This needs to be stressed over…and over…and over…and over. One can’t just say player x is definitively better than player y because he has 5.5 WAR to the other’s 5.2, or whatever. WAR has its own set of issues and doesn’t need to be treated as gospel, particularly in the context of a single season. It’s a single piece of a 30-piece puzzle, if you will.

    Comment by Jason B — September 9, 2010 @ 9:54 am

  56. @Re: Cargo For MVP “Youre argument makes no sense….Noone has control over what their teammates do.”

    Oh yah? Well, if Carlos Gonzalez were a REAL MVP he’d make his team mates play better. He’d work them up with an Ichiro-All-Star-Game-style filthy speech and he’d run and make plays they were missing in the field. And if they hit poorly or made bonehead plays, he’d be waiting by their locker after the game to tell them how they could improve and hit them in the arm REAL HARD for hurting their playoff chances.

    And hence, by a combination of love and fear- he could control what his team mates did. (Source: The Prince, by Machiavelli)

    Comment by B N — September 9, 2010 @ 11:00 am

  57. Maybe I am missing something. I don’t see where Dave double counted anything. It makes sense to me that a given HR or anything else in a high run environment is worth less, in terms of wins, than in a low run environment. If two players end up with the exact same stat line for a whole season and one plays in Petco and the other plays in Coors, the one in Petco will have generated more wins. I think that is what Dave is saying. What is wrong with this?

    Comment by Cyril Morong — September 9, 2010 @ 11:06 am

  58. Another problem with Dave’s explanation is that he says a HR is worth 1.4 runs at both Coors and Petco. That can’t be true because there are likely to be more men on base at Coors, so a HR would be worth maybe 1.7 runs at Coors and 1.2 runs at Petco, each providing about 30% of the runs needed for a win.

    Comment by newsense — September 9, 2010 @ 11:17 am

  59. That makes sense. But we have to take both into account. A given HR might be worth more runs at Coors, but a given run is worth less in terms of wins. That might be the way to get a more accurate idea of value.

    But if you go to Tom Ruane’s article on “The Value Added Approach to Evaluating Performance” at

    You can see that the value of a HR was not much less in the 1968 AL (1.388) than it was in the 1996 AL (1.404) or the 2000 AL (1.407).

    Comment by Cyril Morong — September 9, 2010 @ 11:29 am

  60. The point of the article was to point out the fact that a run generated by Carlos Gonzalez is worth less than a run generated by Adrian Gonzalez. This is done in order to make a side-by-side comparison of the two.

    I do see your point, though, in that it basically seems to be a form of park-adjustment. Playing 81 games at Coors field should adjust a players wOBA down, in comparison to playing 81 games at Safeco bringing a players wOBA up (if the wOBA has not been weighted in the first place).

    Comment by Ty — September 9, 2010 @ 11:32 am

  61. Because I think he’s saying that you have to do this in addition to adjusting raw stats (wrong), or in the place of other adjustments (which is fine, but needs more explaining.)

    Comment by The Duder — September 9, 2010 @ 12:02 pm

  62. The Rockies are scoring 6 runs per game at home this year. In 1930, The Yankees scored 6.9 per game for the whole season. Tom Tango has the run value of a HR for the Yankees for that year at 1.46. Go to

    So my guess is that the value of a HR in Coors this year is pretty close to 1.4. Tango also has the HR values for the Rockies for entire seasons (but not broken down by home/road). But they do tell us something since they are all pretty close to 1.4. Here are some selected years


    If the value in their home games is 1.7 in any of those years, the value in their road games would have to be about 1.1. That just is not possible. Even in the 1968 AL, Tango has all of the teams around 1.4

    Comment by Cyril Morong — September 9, 2010 @ 1:17 pm

  63. what seems to be overlooked is that every time a rockies player puts up amazing offensive numbers, all that happens is it must be detracted from and lowered so they look worse to people who are inherently biased against colorado… like someone else said, the rockies pitching has been, as flawed as FIP-based WAR is, the WAR leaders for this season and 09, and no one gives that any credit whatsoever. when you make it your mission to destroy their offensive accomplishments, how about you properly mention their pitching.

    Comment by fredsbank — September 9, 2010 @ 2:16 pm

  64. Nope.

    In every run environment that you could conceivably encounter in the Major Leagues, the run value for a home run is around 1.4. Yes, a home run is likely to score more runs at Coors, but the baserunners are are also responsible for those added runs. In a contrasted run environment like Petco, where baserunners are at a premium, you can see how valuable a home run would be worth.

    I can’t point you to something that would explain and confirm this, but if you searched Tom Tango’s blog, “The Book”, I’m sure you can find much information about this.

    Comment by Nathaniel Dawson — September 9, 2010 @ 5:52 pm

  65. That’s where my initial confusion came from. I thought he was double counting park factor, when instead, he’s not really doing anything. The article is unnecessarily confusing.

    In a sentence, all that’s happening here is that Dave is park-adjusting Carlos Gonzalez’s value in an unorthodox way. There are subtleties to the method but nothing ground-breaking. It’s almost a semantic point more than anything else.

    Comment by Oscar — September 9, 2010 @ 6:38 pm

  66. How about we stop splurging over CarGo and give some love to a guy on the same team, within .3 WAR, who has 120 fewer plate appearances: Tulo?

    Comment by Dwight K. Shrute — September 9, 2010 @ 8:06 pm

  67. Tulo is amazing no doubt but when he is behind in every major stat save OBP/BB he is likely to be left out of discussions like this. The question I want answered is if Cargo has his home performance weighed by factoring Coors, is his road performance weighed factoring his teams road woes?

    Comment by Yo Comprendre — September 10, 2010 @ 8:52 am

  68. also dont forget cargo spent 44/122 of his starts batting leadoff

    Comment by fredsbank — September 10, 2010 @ 11:08 am

  69. We are a group of volunteers and opening a new scheme in our community. Your web site offered us with helpful info to paintings on. You’ve performed a formidable activity and our whole community will be thankful to you.

    Comment by Cathy Ulisch — November 15, 2011 @ 3:35 am

Leave a comment

Line and paragraph breaks automatic, e-mail address never displayed, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Close this window.

0.159 Powered by WordPress