Hey guys (gender inclusive)? Guys. I hate to interrupt. But listen for a second.
Look, I know that Miguel Cabrera is an awesome baseball player, and I know he had a huge game last night, and that he's currently two homers back of Josh Hamilton from being in a position to win the first Triple Crown since 1967 (as I write this, he appears to be about five points up over Trout in the batting race, and six RsBI over Hamilton). That'd be a pretty amazing accomplishment, and I know that it's gotta be tempting to go around touting his MVP credentials.
Here's the thing, though, guys: there's no AL MVP discussion this year. That was canceled. I'm sorry, I thought you'd heard. It's Mike Trout. They made the announcement, like, two months ago. I don't know what rumors you've heard or whatever, but nope, that hasn't changed. It's still off.
The various Wins Above Replacement measures break it down like this (through Tuesday; they'll tick a bit in Miggy's favor when the numbers are updated tomorrow):
WAR(P) shouldn't be taken as gospel -- any one of them, or even, I guess, in cases like this in which they all strongly point in one direction -- but that's a pretty huge difference. When three very different groups of smart people take smart approaches to trying to assign value to ballplayers and they all come out the same way by a huge margin, if you want to argue that the one they all agree is less valuable is actually more valuable, I think you've got a pretty steep hill to climb.
And (as usual), if you take the numbers apart into their components, I think they make a ton of sense. You've got hitting, which is obviously Cabrera's strongest (some might say only) quality. After the huge night last night, Cabrera's hitting .333/.396/.612; Trout is hitting .328/.395/.559. Slight edge to Cabrera based on raw numbers (almost all on slugging percentage), but it turns out that the Big A has been playing as a pitcher's park over the past several years, Comerica as a slight hitter's park, while Trout has also had the cavernous Safeco and whatever-the-hell-they-call-the-Athletics-park-now to contend with. Accordingly, Trout came into Tuesday with the lead in OPS+, 169 to 165; I imagine they're roughly tied today, or that Cabrera moved slightly ahead. (wRC+ probably tells this story better, but that includes other things I want to break out separately.)
OPS+ doesn't capture all of offense, though. It doesn't tell you, for instance, that Cabrera has grounded into 28 double plays, leading the Majors by five, while Trout's had just seven of those. Part of that is opportunity, of course, Trout being a leadoff hitter and all, but it has certainly impacted their values, and a lot of it is the vast difference in speed. That speed has value in other ways, of course. FanGraphs calculates that Trout's baserunning in ways independent of all the stolen bases -- advancing from first to third, avoiding running into outs, etc. -- has been worth 6.1 runs to his team to date, while Cabrera's has cost his team 2.9. Knowing what you know about their abilities -- just about what they look like running the bases -- doesn't that sound right to you, or maybe even a little Miggy-friendly? It sure does to me.
Then there are the steals. SB value is bound up in FanGraphs' wOBA calculation, but extrapolating from the wOBA formula, it looks like Trout's incredible 46steals at a 92% success rate has added about 9.5 runs, while Cabrera's 4-for-5 has added a half run. Doesn't that seem about right, too, intuitively? SBs get derided a lot (and rightly so, I think), but if you do it a lot and are almost never caught, you're certainly creating more runs. This suggests that for about every five steals, factoring in those four CSes, Trout's gotten the Angels an extra run they wouldn't have had if he stayed put. I'd buy that.
Then of course there's defense. Cabrera has been less of a disaster than I thought he'd be at third base, but he's been bad. The metrics disagree on how bad -- BP says -3.4 runs, Baseball Reference -4, FanGraphs -9.4. They similarly agree that Trout has been phenomenal in the outfield, and disagree wildly on how much: BP +4.6, B-Ref +25(!), FG +13. Cabrera then gets a run and a half back, somewhat surprisingly, because CF and 3B are viewed as of equal importance, and Trout has spent more time at the less important LF than Cabrera has at the much less important DH and 1B. Taking the averages of the fielding metrics, because why not, Cabrera has cost his team 5.6 runs, Trout has given his 14.2. Even with the positional adjustment, that's an 18.3-run advantage for Trout.
If you think Cabrera deserves the MVP, and think you can argue that the systems and I are way off on any one of those factors (not just that we can't be 100% certain about the exact numbers, but that it's seriously off for some reason), please do. Otherwise, that's what we've got: they've been essentially equal with bats in their hands, and Trout has been much, much better in every other aspect. The baserunning, basestealing and defense alone adds up to about a 36-run advantage for Trout; I think the GIDPs could add another ten or more, but I'm frankly undecided about how to handle that, so I won't (I don't think it's necessary). 36 runs is a huge number; WAR says, for example, that Prince Fielder, who has been really, really good, has been worth about that much over the course of the entire season vis-a-vis a random replacement-level player. So, basically: Mike Trout has been one whole Prince Fielder better than Miguel Cabrera.
Well, almost. Of course, Cabrera has played all but one of his team's games, while Trout (thanks mostly to his late call-up) has missed 22 of the Angels'. That certainly has value...but not even close to 36 runs of value, which would be 1.7 runs for each extra game played. Nobody's ever done that, ever (the all-time single season runs created leader is Barry Bonds, whose 230 in 2001 came out to 1.5 per game played). Cabrera clearly gets a slight advantage with the bat, based almost entirely on playing time (as evidenced by the fact that B-Ref's and FanGraphs' batting runs for the two players are roughly even, despite including stolen bases). But they certainly don't think Cabrera's playing time has pushed him that far ahead. You can disagree, but you've got a lot of arguing and proving stuff to do. It's very hard for me to see an argument that puts Cabrera ahead of Trout, or even particularly close.
So then what else is there? Winning the Triple Crown would be awesome, but it can't be used to somehow increase Cabrera's value. I assume for these purposes that most people reading this understand why the batting average and RBI themselves don't actually add value, but worth noting: (a) per Baseball Reference (see here and here) the park-adjusted league batting average is .247 for Trout and .259 for Cabrera; and (b) the RBI, as usual, are mostly a function of opportunity -- per this, while Cabrera's having driven in 20.3% of the runners on base for him is good for 3rd in the AL, Trout's 18.3% is 14th (and very close to 10th). Most of the difference is that Cabrera has batted with 136 more runners on base than Trout has, including 71 more in scoring position. Also worth noting, if you're caring about those stats, is that Trout has scored 18 more runs in those 21 fewer games.
And that's about it. If you care about whether or not the MVP comes from a contending team (which you shouldn't at all, but now is really not the time), that's a push (Coolstandings gives both teams between a 21% and 23% chance of making the postseason; BPro gives the Tigers 21% and the Angels 29%). I can't think of anything else (if you want to argue leadership or something like that, knock yourself out, you're on your own). It all comes down to Cabrera having really pretty, and potentially historically interesting, raw batting numbers, and Tigers fans being very vocal and protective of their guys. That's all there is. Trout's still running away with this thing.
So, again, sorry guys. I know we're used to having these discussions at this time of year. And we've still got a lot of great talk out there for you to enjoy -- the NL MVP is pretty wide open, for instance, and both Cy Youngs, and we're still technically open for the NL ROY (it's totally Harper's, but we'll hear you out). The AL MVP and ROY are closed up for the season, though. Nothing to see here. Hey, try again next year!
(Note: sorry for the lack of Cram Sessions, and for posting in general, here lately. This was intended to be a Cram Session, and it turned out I had too much to say about it. We'll do better, I'm sure.)
(Note also: I don't mean to say Trout actually has it locked up. I'm actually pretty afraid that Cabrera will steal it. But barring a catastrophic collapse by Trout, it'd be one of the worst snubs in history.)
I will say that Trout was definitely the MVP..........for the month of July. Truly phenomenal month. Good for him. Maybe next year he can put it together for a full season the way Cabrera did this year.
Under my inherent powers as author of this article, I need to insist that we all immediately recognize that @Sabre Player has absolutely no clue what he's saying, no willingness to learn, and no interest in reading or responding to any but the least relevant and damaging points of your responses, and simply isn't worth another second of your time.
My inbox thanks you in advance.
And another thing! Trout's stats are completely inflated by the "no book developed yet" effect. He was torrid when he first got called up. But then, as is inexorably the case with hot rookies, the pitching book gets developed and he reverts to his much more reliable mean. Case and point: Trout batted 284 in August and 279 in Sept/Oct. And he struck out at a 30% clip. Accordingly his OBP and OPS plummetted. These batting numbers are totally respectable. They produce a fine OBP. But the strikeout rate is devastating. Pitchers have figured him out and turned him inot a very good, but not very great hitter. Compare that to Cabrera. Cabrera's book is well developed. The book: he has no weaknesses, so say a prayer and hope he hits it at someone. That's why he is so feared by pitchers and managers alike. Hence, why Cabrera produced through the whole year (357 avg in Aug and 339 avg in Sept/Oct with a remarkably steady 390+ OBP and a slightly increasing OPS on a month by month basis throughout the season). These are very significantly better than Trout. The tigers had the benefit of having the best hitter in the league in their lineup over all 162 games. The Angels had the benefit of having a great, great hitter on their team for only 75 games. There simply aren't enough defensive and baserunning opportunities available for a centerfielder to make up for that relative lack of offensive production over such a large number of game.
This is an epic discussion. I thoroughly enjoyed reading all the new comments this morning over coffee.
@Kaufmak Whatever it is, it's awesome!
Good comeback, Potsy. I suppose that's more satisfying than having to address some extremely easy to understand laws of probability theory. Do you know what a probability tree is? If so, go ahead and draw one. Start with 20 games of minor league production stats. Then draw another one with about 75 games of super star stats. Then draw another one with 67 games of good, but not great, stats. Add up the outcome. Now compare that to 162 games of super star production stats.
Better yet, use actual OPS numbers in your probability tree. Quintin Berry is a servicable outfielder who got called up from the minors with an OPS of under 700. Plug those numbers in to your April tree for Trout. Do the math. It shows you a player that is a full 100 OPS points below Cabrera for a full season. You can't ignore the fact that Trout didn't contribute in 20 games for the Angels when deciding who was most valuable for the whole season. This is not the MVP award for players who played in games from May to September (in case you are wondering, Trout's per game OPS is 80 points lower than Cabrera's for this period). It is a full season award. Looking at OPS by game as opposed to OPS by at bat is a far more important measure of one's value to a team over the course of a season. This is so basic.
@Sabre Player Are September games more important than May ones? If not, then this doesn't matter.
@Sabre Player You've seen his batting line, right? With the OBP basically at .400? That means that his typical batting night has him reaching base twice in five trips. Which is really good. You really do have no idea what you're talking about.
And where were you yesterday when he had two singles, a double, and a triple?
@Sabre Player "Add up the outcome. Now compare that to 162 games of super star production stats."
Maybe you should do that since you're the one making the argument. It's not on us to provide your evidence for you. Draw it up and show us the actual math that gives Cabrera a higher value than Trout. Not in generalities and theories, which is what you've shown so far. In actual concrete performance numbers.
No they are not. But a batter's ability to effect the outcome of a game is equally applied game by game. Give me a consistently great player over 162 games vs. a great player for 75 games and a good player for the rest. There is only so much marginal benefit that can be achieved for a team over a full season if the great stats are applied to less than half the game.
To analogize. Let's say you and I weighed the same at the beginning of April and we had a stated goal of gaining the most amount of weight in 6 months. I choose to eat a signficantly higher number of calories each and every day for 6 months. You eat at the same or higher rate than I do for 2.5 months and then at only a higher than average rate for the remaining 3.5 months. Who will weigh more at the end of 6 months? I will--and not by just a small amount. Why? Because our bodies can only process so many calories on any given day. The rest disappears as waste.
Trout's stats certainly benefitted the Angels during his torrid period. But, just like calories that your body can't process into weight gain and so disposes of instead, a lot of those stats came in games that the Angels would have won anyway or would have lost anyway. That leaves more than half the season for the truly historically great stats that Miggy produced to have an impact on his team. Whereas Trout applied only above average, but not Cabrerian, production over more than half his teams games.
@jlwoj I must have been watching Cabrera go 4 for 5 with a HR at a critical point in a division clinching game. But tonight, I was able to see Trout K 3 times-- each time with men in base.
It probably has to do with the diminishing marginal returns point. If you have lots of players on the same team producing high WAR numbers, there has to be some waste in there. I haven't checked the game logs, but I bet the Angels and Bosox have a much higher percentage of their wins coming from blowout games (4 runs or more differential). I'm too tired to check if That is true.
@jlwoj Whatever.....the Angels pitchers have high WAR too. The Angels team WAR is far and away the highest in the league. The Red Sox have a much higher WAR than the Orioles and only one win less than the Tigers. You would like to chalk it up to luck and schedules. It is probably more likely that the WAR stat has some deep flaws.
@Sabre Player Well, that was ... unexpected.
FWIW, it wouldn't be so much WAR that overvalues defense as the defensive systems that the different WAR measurements use (and FanGraphs, Baseball Reference, and Baseball Prospectus also use different defensive models -- and I'll just note that BP values Bourn MUCH lower than the other two). That said, the defensive stats are pretty straightforward in terms of their models -- they give credit for the extra outs created by defenders, and look at how many runs those outs are worth. There's difficulty in measuring the outs created, but in terms of taking those outs, once you have them by whatever method you've used, the runs conversion should be fairly uncontroversial.
This isn't proof or anything, just my own sense of what feels right, but: the top defenders being +20 or +25 or so over the course of a season doesn't seem crazy to me -- that's only a run every five or six games.
OK...I've been playing around with numbers. And I have to admit that while I don't have a formula I care to share, this distribution factor is less than I would have surmised. I guess I have to admit that Trout was probably more valuable in his 139 games than Miggy was for 139 games. I'm still troubled by 23 games not played....but Trout was probably the better player for a substantial majority of games played this year. I still think that WAR massively overvalues defense. Michael Bourn is the poster child for this. But again, I guess I have to capitulate. Trout is the 2012 MVP.
@Sabre Player I do not, though I'm curious to see how you apply this to Trout's 139 games played.
@jlwoj Do you reject the idea that if high WAR numbers are accumulated in a very small number of games, then it is very likely that each marginal increase in WAR is marginally less impactful to contributingto a win?
@Sabre Player while there is room for doubt that the ten theoretical wins represented by WAR translate well to ten actual wins in thirty games, without a method for actually discounting how many real wins we'd expect such a torrid streak to be worth, I don't see another way to vote.
@Sabre Player Please do cite WAR and WHIP in the same paragraph in a way that implies that they are equally useful.
If you have an individual-level stat that accounts for luck and schedule and all the other things that go into team wins, I'm sure we'll all be very open to it. Hopefully it also accounts for division alignment, because the Angels had more wins than the Cardinals and Tigers.
OK. Thanks for answering. One more question: if a player achieves a WAR of 10 in 30 games and then sits out the remaining 132 games.....should he win the MVP over a player who plays all 162 games and achieves a WAR of 9?
Well, by now you know that I completely reject the whole concept of WAR. Trout and the Angels are microcosm of its flaws. Trout isn't the only Angel with great WAR. In fact, there are 8 Angels in the top 70 players in the American league in this specious stat. And they have 4 pitchers...all of whom are in the top 25 in the league for WHIP. It seems that the Angels should have never lost a game. Yet, they'll be watching TV tomorrow night.
@Sabre Player theoretically, sure.
Also, a ten-game swing is theoretically possible in a 30-game stretch without considering WAR. A player could be the difference between a win and a loss (though how we determine THAT is questionable) ten times in a 30-game period.
let me ask you a question: Can a player achieve a WAR of 10 in 30 games?
@Sabre Player I like that you accuse me of using a spreadsheet when what you did in the above post boils down to this:
Compare Trout's OPS in a given game to Cabrera's. Whoever's is higher wins that game. Whoever wins more games wins the MVP.
Given that this is not in the least how baseball works, forgive me if I don't buy this argument at all, especially since you're still insisting on monthly breakdowns, which are highly arbitrary.
at least 82 wins in a theoretical season compared to Trout's no more than 78. In reality, the gap is probably wider.
@Sabre Player how. much. is. this. worth.
If each player played in the exact same game where the two oppossing teams were perfectly controlled to play exactly evenly to a tie for 162 games, with the same exact pitcher and the same exact fielders. and the only variable was either Trout or Cabrera in the batting order (with the same other 8 guys each producing at the same production level for both teams). Leaving running and fielding aside for a moment.......then the Cabreras would win every single game in which they had an OPS advantage. So the numerical value would be something like 108 wins for the Cabreras to 54 wins for the Trouts. All of the Trouts wins would come in July and May. The Cabreras would win every other game. Half of those games (April, August and Sept/Oct) would have been won by substantially larger margins than all the other games (Cabrera is 25% more productive in those games). The May and July games would clearly be won by the Trouts by also a significant margin....and that margin would be more signficant once defense and baserunning is introduced into the equation. So that leaves June games. Again in theory, the Cabreras should win all of them based purely on OPS. But I'll grant that, although rare, there are enough base running opportunities and defensive opportunities in a 27 game stretch to turn many of the Cabreras wins in those tight games into ties or losses. Heck, maybe every single game. Assuming it was every single game, that would at the very best case for the Trouts create a tied season. But there is simply not enough game altering baserunning opportunities or centerfield/3B defensive opporunities available in April August and Sept/Oct to affect the outcome of those games given the giant shortfall in Trout' batting production compared to Cabrera.
Now instead of going month by month or game by game, what you want to do is take the year end averages and apply them to all 162 games played this year. There is a fatal flaw to that. For if you do that, the run differntial of my theoretical season is exactly the same for all 162 games. Every game is much closer because Trout and Cabrera are much closer in year end OPS averages.Throw in defense and baserunning and yes, maybe the Trouts win all 162 games in my theoretical season. But baseball is played on the field and not on your averaging spreadsheet. You are ignoring what actually happened this year on a game by game basis. You are literally taking Trouts stats and evenly apportioning them to all 162 games. In fact, Trout had some phenomenal games--especially in July. But he had a huge gap of time when he produced far lesser numbers....including 23 games that he didn't even play. Cabrera was a significantly better player than Trout for at least half the games this year....and roughly equivalent for several more after that. Trout was definitely better for many games, but it was a clear minority of games. He is not MVP worthy relative to Miggy.
@Sabre Player suppose it is. What's the numerical value of it?
It is a mathmatical fact.
@Sabre Player But you still have not calculated the effect of those disparities. The RE24 analysis by Cameron and the WAR systems do. They look at Trout's good parts and his bad parts and they come out to a number. Then they do the same for Cabrera. And every single time, the number is higher for Trout.
I reassert that your only argument is the unproven assertion that consistency is more valuable. Is that not true?
I'm not basing my overall assertion on the merits of Cabrera over Trout with only production with ROB. I happen to think it is important, but not dispositive. I find it interesting, but again not dispositive, that Trout performs at his very worst at the most important times. Miggy is the opposite of that even when accounting for GIDP's.
My main assertion is based on the full body of work over all 162 games. It is just a simple fact that Cabrera was way, way way more productive and valuable than Trout for half the games this year. This is completely indisputable. We can discuss the other half of the games. I'm prepared to grant you that Trout was more productive and valuable for the other half.....but not nearly to the extent that Miggy was for the first half.
@Sabre Player I've responded to every single point you've made that deserves response.
Of course the analysis there is all at-bats. The point is to give credit (and debit) for everything the player does based on the context in which he does it. Sometimes a player bats with nobody on, sometimes he has an RBI chance. You can't just look at the runners-on situations and completely ignore the rest if you want to ask about the player's value to his team. The runners-on situations are more heavily weighted than the others by the natural fact that the run expectancy swings in those situations are greater. Hitting into a double play with one out is far more damaging than flying out with nobody out. The RE24 method accounts for that.
"points to a stat that speaks to performance above an expected outcome for all at bats" -- I don't think you understand the stat. Every game situation has a run expectancy attached to it. It's empirical, not theoretical (in the most common usage -- there are also theoretical models possible, but most people don't use those) -- it asks, for instance "how many runs did all teams score after having a runner on second and two outs?" It assigns that value to that situation.
So what this method does is examine the difference between that value before and after the individual player's at-bat. If there was a runner on second and one out and now there's a runner on second and two outs, the batter's RE24 is credited with the (negative) difference between the run expectancies in the two situations. If instead now there's a runner on first and one out and one run scored, the batter is credited with the difference between the two base-out states plus the one run that scored.
So what this does, as Dave stated, and as I stated, is it properly weighs the values of the various hits, walks, outs, etc. that the two players made over the course of all the at-bats they faced and the particular contexts in which they faced them. I don't actually believe in valuing players this way because it depends on teammates creating those situations, but since you do value context (citing hitting with runners on, etc.), it's not clear that you should have any objection remaining to the idea that, over the course of the year, Trout was more valuable than Cabrera with the bat.
There's still this questions about consistency, but we're not getting anywhere with that one. You appear to believe that the burden of proof lies elsewhere than it really ought to.
Very original and creative response. Feel free to actually address the point if you like.
I just went to the link you sent. Thank you. I have to say, there is some compelling stuff in there. However, and correct me if I'm wrong, this analyis pertains to all at bats--not just at bats with ROB. Right? The guy opens up with the bold statement about proving that Trout was more productive with ROB...and then points to a stat that speaks to performance above an expected outcome for all at bats--including at bats with no runners on base (a very frequent occurrence for a lead off batter). It is already well documented that Trout is generally very productive....and it is also well documented that he is even more productive when the bases are empty.
@Sabre Player "I'm glad to see you that you do actually get the point. I was starting to worry about you." Oh, fuck off.
"I'm glad you are acknowledging that there is value to consistency at least at extreme levels." I did no such thing.
"I don't have the math formula to prove it for baseball." Welp.
First, I'm glad to see you that you do actually get the point. I was starting to worry about you.
Second: In my example, I did indeed use absurd extremes. And I'm glad you are acknowledging that there is value to consistency at least at extreme levels. My theory is that it is applicable at all levels. I don't have the math formula to prove it for baseball. Alternatively, you definitely can't disprove it. Glad to hear that smart people are working on it. As you know, usually people engage in such activity in order to prove a theory or to explain why something that is seen with the eye actually occurs.
But, I digress. Fine, I'll work within your "extreme" stipulation. in reality, there are huge extremes in the data for Trout and Cabrera. Cabrera's OPS was 25% higher (yes, 25%) than Trout's for half of the games played this year. Yet for the 35% of the games when trout was better his OPS was only about 13% better. These are huge extremes in terms of deviations from the mean. Very extreme......and applicable to half of the games. This is a big deal. If you were to draw Cabrera's production bell curve over 162 games (with Y axis being number of games....and x axis being OPS with low numbers left and high numbers right), Cabrera's curve would be very, very high and very narrow. It would also be very symetrical. Trout's curve would be much, much lower and it would look skewed. It would go out further to the right than Cabrera (to account for July)....but it would not be very fat (not enough games) on the right side. It would, however, be very fat on the left side (half the games with weak OPS).
@Sabre Player Just saw this http://www.fangraphs.com/blogs/index.php/trout-versus-cabrera-offense-only-context-included/ in which Dave Cameron does the work for you on measuring, in runs, how the two hitters (not fielders, not baserunners) have performed taking into account the context of runners on base, outs, and so forth. Guess who wins.
@Sabre Player "It is pure common sense" -- no, it isn't. We're not dealing with the extreme situation and we're not dealing with analogies. We're dealing with two slightly different positions on the spectrum from extreme consistency to extreme volatility. You've shown that the far extreme of volatility is not as valuable as more consistent positions, but you haven't shown that it's as simple as "the more consistent, the more valuable."
Nor, I imagine, will you, because a lot of people who spend a lot of time thinking about baseball have worked on this question without really coming to a satisfactory answer. In the absence of a satisfactory answer, you're falling back on what you deem to be "common sense." I'm falling back on the simpler system on the principle that it's better to use that if there's no evidence that the more complicated system gives a better result.
well. you are right that I was taking as a given that even distribution of production is more valuable. It is pure common sense....and I think you are smart enough to know that. We can argue about how much more valuable, but not the basic premise. If you want to argue the basic premise, then you are simply acknowledging that you have never taken a basic statistics class in your life. If it will be easier for you, let's be even more absurd in the stats. Let's say player 1 jams all his production into one epic lopsided game. His team bats around forever and forever. He hits 15 home runs in the game. His team wins the game by some silly margin--lets' say 63 to 5. But then he is completely unproductive for the other 29 games. Mathmatically, he is unproductive for 29 games......and 90% of his production in one game is completely wasted. As I said, you can't win 1 game twice. In this case, you can't win one game 10 times. Are you just being argumentative or do you really not get this? In all seriousness. This is so basic. I'm happy to engage in a conversation around how to measure the clear value of consistency and the clear detraction of value that comes with volatility.
I'll try one last analogy to try to help you. Let's say there are two bags with 30 coins in it. Bag 1 has 30 coins worth 1 dollar each. Bag 2 has 29 worthless slugs and 1 coin worth $30. As you probably know--even though you appear to know nothing of statistics--both bags are exactly equal to each other in terms of the expected dollar outcome from reaching into the bag and pulling out 1 single coin. Fine. But you have to go one step further if you want to apply this analogy to baseball. As I said, you cannot win the same game more than once. So, to apply that to my coin analogy, there must be an addional rule that any coin pulled out of the bag that is worth more than 6 dollars must have a 100% tax applied on the amount above $6. So @4 of the 3 dollars is of absolutely no value (akin to the fact that excessive production in a single game is completely, totally wasted and useless). Now the expected outcome for bag 1 remains $1 dollar.....and the expected outcome for bag 2 becomes 20 cents.
I have to say, I am thoroughly enjoying this epic conversation. Keep it up fellas! @Bill_TPA @Sabre Player @jlwoj
@Sabre Player @jlwoj I so want to be done with this, but the baseless assertion that "No amount of amazing defense and baserunning can overcome" an 86-point edge in OPS -- which isn't actually all that much, even putting aside the fact that such an edge doesn't exist and you totally made it up -- is pretty hilarious. Ozzie Smith must've been a really terrible baseball player, then.
@Sabre Player Here's the core: "In reality, Player 2 is substantially more valuable and productive." You still have not actually proved that, or even marshaled evidence suggesting it, since proof is hard. You've just stated it. Repeatedly. It might be true. It might not be true. But saying it because you believe it doesn't make it so.
In terms of backward-looking value, performance with runners on and so forth is important, but I'd want a more systematic way of knowing how much it's worth before I start going down that particular rabbit hole. You've just cited a bunch of OPSes without providing any run values.
The math is pretty straightforward. Trout played in 139 games. Miggy played in 161 games. So the Tigers were the beneficiary of Miggy's production potential in 99.4% of their games. The Angels (or any team...it doesnt matter) were the beneficiaries of Trout's production potential in only 85.8% of their games. Before I get into the actual stats of what they actually produced, will you agree with me that this simple fact is critically important in deciding MVP? The rhetorically obvious answer is that you can't be valuable to your team if you are not playing.
Now let's get to the math. For the 23 games that Trout didn't play, I will assign a 690 OPS for those games. For the 1 game that Miggy didn't, I'll assign the same 690 to keep things apples to apples. Let's go month by month and look at OPS:
April. Trout and his replacement proxy: 650 (trout played 3 games in April with a 348 OPS). Miggy: 940. .
May: Trout: 941. Miggy: 839.
June: Trout 950. Miggy: 990
July: Trout 1259. Miggy: 1086
Aug: Trout: 866. Miggy: 1092
Sep/Oct: Trout:894. Miggy 1077
When looked at this way, Miggy's OPS was 86 points higher than Trouts for 2012. That is a massive margin. No amount of amazing defense and baserunning can overcome that kind of production for the reasons I've already articulated.
But this doesn't even get to my main point. In an earliery post, you used the word 'volatility". Yes!!!d Thank you. That is the perfect word to describe a forumula that the Sabr Wonks need to create in order to penalize someone who produces great stats, but with high volatility. Trout is the poster child for high Vol. And this gets to the problem of looking at just averages and/or WAR. An average takes a player's production and divides it by plate appearances. WAR is even worse than that because it is more or less an absolute number without placing any value on distribution. To illustrate my point (and I'll use extremes here to make my point)....take 2 theoretical players and consider the following scenario. Player one plays 30 games. He produces an OPS of 1000 for all of the plate appearances in that game. Player 2 plays the same number of games and produces the exact OPS. They both hit the exact same number of home runs. They score the exact same number of runs. Batter 1 goes hitless and walkless in 25 games and strikes out in half of his plate appearances in those games. Player 2 gets a hit in all 30 games and never strikes out. In your world (the world of existing statistical measures), Player 1 and Player 2 are equally valuable and productive. In reality, Player 2 is substantially more valuable and productive. Yes, Player 1 plays phenomenally great in 5 games....and his stats are so gaudy in those 5 games that he is more likely to affect the outcome of those 5 games. But even with those gaudy stats, the outcome of those 5 games will still be decided mainly by whoever is pitching for both teams. And presumably, some of the wins produced in those 5 games will be produced by massive run margins, in large part because of Player 1's obscenely high production. But a win is a win...regardless of margin. So, a signficant percentage of the gaudy stats from those 5 games are of absolutely no value to the outcome of those winning games. And they are certainly of zero value to the outcome of the other 25 games. You can't win one game twice! For the other 25 games, Player 1 is a sea anchor to his team.
Back to reality now. There were 22 games played in April. Cabrera's OPS was 289 higher than Trout and his proxy for the first 22 games of the season. In May, Trout was 102 points higher than Miggy. In June, Miggy was 40 points higher than Trout. Then July came. Trout got incredibly hot. His OPS was a red hot 1259. Amazing...truly. His WAR skyrocketed in July. Miggy's OPS was only a "pedestrian" 1086. Herein lies the rub. From a stats perspecitve, Trout's MVP month of July gets spread out over the whole season. But in reality, his production only impacted 25 games. Let's move on. In August, the american league pitchers had developed their pitching book on Trout and his production fell off the table. He was well below 900 for the rest of the season--400 full OPS points lower than his July production numbers. Miggy's OPS was consistently very high all year. His OPS was 226 (sic) higher than Trout's in Aug. His OPS was 183 points higher in Sept. There can be no other conclusion but that Miggy was not a little better, but a whole boatload better, than Trout for roughly half the games of the season. There were about 30 games that were a push (Miggy had a 40 pts higher OPS in June than Trout , but I'll grant that is not enough of a differential to declare a winner--especially given Trout's production in other parts of the game). That leaves only 55 games where Trout outproduced Miggy--most of that coming from 25 games played in July.
In short, volatility of production distribution is Trout's achilles heal in 2012. Miggy, on the other hand, had phenomenal consistency of production distribution. Simple probability laws are very applicable. As I said, I hope someone figures out a formula for it to apply to baseball stats.
One last thing. Why will you not engage me on the concept of production with runners on base? Those are the most important of all plate appearances because batters have a chance to produce both parts of run ingredients (getting on base and getting runners home). Miggy's OPS was 77 points higher with ROB. Which brings me to my last favorite stat: production with men in scoring position and 2 outs. Miggy's OPS was 1211 in such situations. Trout's was 782--a 429 differential. Miggy was at his very best at the most crucial times. Trout was at his very worst.
@Sabre Player your argument relies on the scores of the games. These scores are created far more by teammates and opponents than by the player in question himself. Hence my statement.
@Sabre Player I dispute that it matters at all given that Trout's total production at the end of the year was greater.
Do you dispute the notion that Trout and his triple A proxy for all of April (i'll play your WAR games to humor you) had significantly inferior offensive production than Cabrera for well more than half of the games this year?
It says no such thing. It simply says Cabrera had many, many more opportunities for his consistenly better offensive production to impact the outcome of games this year.
@Sabre Player this entire argument depends on a player's value being largely determined by the performance of his teammates. If that's a road you're OK traveling, then I don't know what to say.
Well, that's hard to do, but no harder than it is do disprove it. Try this:
For starters, Let's just agree that Trout didn't help the Angels win games in September. Let's also agree that neither trout nor cabrera helped their teams in games their teams lost---even if they prodigiously raked in those games. I hope that's not too controversial for you.
Now let's look at May, June and July--the general period that I'm calling the trout torrid period (the TTP). During the TTP, the angels went 49-32 for a 60.5% winning percentage. I'll grant that this is impressive especially given that Pujols didn't start batting until after the all star break.I haven't gone game by game (something you would literally need to do to analyze the actual--as opposed to theoretical--effect that trout had on the Angels during the TTP). But I will grant you that Trout had a posive impact on the Angels during the TTP. But let's look closer. Of the 49 wins during the TTP, roughly half of them came in games where the Angels outscored their opponent by more than 4 runs. I'm sure that trout played well in those games and contributed to the wins in a big way. Those stats are clearly very valuable. But he also produced stats in those games that only contributed to a higher run differential. Those stats are of no more value than stats contributed in a loss. Full stop. So that leaves only 25 wins during the TTP in which the Angels won be a differential of 3 runs or less. These games would be worth analyzing on a game by game basis. Perhaps trout single handedly positively decided the outcome in many of these games. But chances are that the pitcher was a bigger decider.....and there were 8 other batters who may have contributed more. The point is that this is a very small number of opportunities during which Trout's monster stats were applied to marginally close winning games. After the TTP (August and Sept/Oct), the Angels record fell to 32-25 for a 56% winning percentage despite the fact that Pujols hit like he is supposed to. Trout had good, but not great, production during these times. Again, you would have to go game by game to really understand Trout's role in this relative decline. But I'm certain it is there--especially when compared to Cabrera--and I'm sure it is significant.
Cabrera played all 162 games. His production was phenomenally consistent throughout. The Tigers played in 49 winning games that were decided by 3 runs or less. So his gaudy production was a variable applied to double the number of marginally close winning games. Again, you would have to go game by game. But mathematically speaking, it would be impossible for the math to show anything other than that Cabrera contributed far more significantly to the Tigers than Trout did to the Angels in this crucially important subset of games. He had a far greater number of opportunities to do so.
@Sabre Player Can you prove that Trout's more volatile performance added fewer wins to his team than Cabrera's steadier work?