Jayhawk Bill

Thanks for the response. 1) Rather than Win Shares, I'd use WARP; better than either, I'd use VORP along with a second-generation or third-generation fielding metric, but those fielding metrics range into the proprietary method world of knowledge. I guarantee you that Boston doesn't use Win Shares to evaluate trades; there was a great piece in BP in early August 2005 that suggested that they possibly used a sophisticated accounting model coupled with a model not dissimilar to WARP and PECOTA. I don't expect that they use either WARP or PECOTA in their internal modeling, but I bet that they use a system that's very similar in its detail. Win Shares is grossly inaccurate regarding defense, and I doubt that they use it. 2) Yup, you're overestimating Johan Santana. Glad that I made you happy.

TheKilo, my recollection is that Kelly Shoppach wasn't even an essential part of the trade--the Guardians were rumored to have wanted Mota and Marte for Crisp, and Boston then expanded the trade to include a swap of catchers. Rician Blast, how does having Kelly Shoppach face tougher pitchers than any other batter on the 2005 Red Sox (minimum 10 PA) do anything either to improve his trade value or, as I originally contended, do anything to support young players? My point is that Terry Francona has had a tradition of screwing over young players, not supporting them. 2007 did perhaps mark a break with Terry Francona's pattern with respect to screwing over some of his younger players: he treated Pedroia better after a tough start where he was giving Alex Cora some better opportunities. He also treated Kevin Cash much better than he'd treated many other backup catchers.

From the linked article: Wily Mo Pena and Kelly Shoppach apparently weren't interviewed in the preparation of this article.

First, if we're looking at talent value less contract value as investment value, then we're in concurrence, more or less. I've already explained why, IMO, (Crisp + Lester) ~ Santana in investment value. The price for Santana seems to be higher, so unless the up-front years are valued much more highly than the out years, I don't see the rumored trades for Santana as good investments. But you're looking at Crisp vs. Ellsbury. Who's going to be better over the rest of their obligated service? The funny thing is, each of the players in question has some odd spikes in performance. If I may: 1) I look at Coco Crisp's MiLB record--a .299/.372/.411 batting line, put up at about the right age for each level--and I expect an MLB hitter posting a .270/.335/.370 batting line, give or take. In 2006/2007 Crisp batted roughly .266/.326/.383, pretty close to my expectation. In 2004/2005 Crisp batted .299/.347/.456, similar with respect to IsoD but way higher with respect to BA and, especially, SLG. It would seem that there might have been some factor that made Crisp hit relatively better in MLB than MiLB in 2004, but that the factor in question had gone away by 2006. 2) Whatever was making Crisp a better hitter in 2004-2005 might've been making him a worse fielder--there's nothing in Crisp's past that suggests his superb work in CF in 2007, except that he was better in 2006 than he'd been in 2005. MLB players peak defensively very early--this looks to be a performance spike, unless there's an external factor of which we're unaware. 3) I look at Ellsbury's MiLB batting line of .314/.390/.426 and I expect a .285/.350/.385 MLB hitter. Ellsbury's .353/.394/.509 MLB batting line from 2007 appears unsustainable. Overall, the biggest advantage Ellsbury has over Crisp is possible upside; the second is remaining obligated years; the third is salary. I'd rank likely 2008 hitting as the fourth-biggest advantage, trailing all three of these other factors. Ellsbury should be better than Crisp as a hitter, but maybe not by as much as we're hoping. Crisp has two advantages: demonstrated MLB performance over more than 44 games and superior defense, at least in 2007. In the short term, disregarding salary and counting defense, Crisp is roughly equivalent to Ellsbury. If that's your point, you've got my support--as long as it's understood that the other "investment factors" make Ellsbury a much more valuable commodity than Crisp.

No. Example1, thank you for taking the time to carefully construct complete positions regarding this issue. Others are taking pot shots at small chunks of your work to call you wrong, but they're not laying out equivalent effort to support their own thoughts in almost every case. That said, I still disagree with your assessment. I have two differences: 1) I strongly dislike Win Shares as a metric. Its weakness is in the area of players' defense, so it's not necessarily as bad when used strictly to evaluate pitchers. When one starts evaluating hitters, though, one gets some terrible results. Here's a quick example: I called up the best seasons from 2004-2007 from THT for a quick counterexample. A great one was right at the top of the list: the best two seasons were both achieved in 2004, by Barry Bonds and Albert Pujols. Both, of course, were mostly great hitters. But Win Shares has Barry Bonds at +2.6 as a fielder and Albert Pujols at +2.1. That's absurd: Bonds is a defensive liability, and Pujols is a Gold Glove-caliber fielder, regarded by some as one of the top defensive players in MLB at ANY position. Checking FRAA for a quick contrast, FRAA had Pujols at +9 in 2004, while it had Bonds at -8; other systems might be less kind to the slow-moving outfielder. Want another example? Check 2007 outfielder stats. The system has Manny Ramirez pegged at +2.7 WS for his defense this year. It awards Carl Crawford, the fleet-footed Rays' LF, just 1.6 WS for his defense. Checking RZR, Crawford was almost 20% more likely to field a ball hit into his zone than Manny. How could Manny be considered 50% better than Crawford if he's reaching over 100 fewer fly balls over the course of a season? OK, enough on why I don't like Win Shares. On to my second, and bigger concern. You project that Johan Santana will earn 150% of the value earned thus far in his career over the rest of his career, and you write that he's earned 123 WS so far. Let's examine that, using these fairly generous assumptions: 1) Although he might be slowed by injury, unless he is injured he will never be worse than his current talent level; 2) His current talent level isn't his 2007 level, but the median value of his three most recent seasons, his 2005 level (23 Win Shares); and 3) We'll use Nate Silver's estimate of a cumulative 15% chance each year for career-ending/talent-level-altering injury. Here's what the next 12 years look like using those assumptions: Year WS Cumulative Win Shares 2008 19.6 19.6 2009 16.6 36.2 2010 14.1 50.3 2011 12.0 62.3 2012 10.2 72.5 2013 8.7 81.2 2014 7.4 88.6 2015 6.3 94.8 2016 5.3 100.1 2017 4.5 104.7 2018 3.8 108.5 2019 3.3 111.8 Let's look at this differently: using this method and 2007 WS rankings* as a benchmark, between what two current 2007 AL pitchers would we find Santana's likely performance, INCLUDING the risk of zero value from injury, for each year? 2008 Lackey/Haren 2009 Kazmir/Wang 2010 Blanton/Pettitte 2011 Matsuzaka/J Weaver 2012 Gaudin/Marcum 2013 Westbrook/Robertson These are all pretty good pitchers. For four years he projects to be as good or better than the 2007 Matsuzaka. For two more years he projects to be a 3/4 starter. Now, I understand that Santana isn't likely to be Nate Robertson in six years--he'll either be better or he'll be injured, I expect, but the net value is roughly the same as a guaranteed-healthy Westbrook or Robertson. But even giving Santana very considerable value for the next twelve years--more than I'd expect that PECOTA would offer over the next five, more than BR would consider likely from using its top comparable pitchers--he doesn't end up earning 150% of his value thus far in his career over the rest of his career. Almost no starting pitcher Santana's age would earn that much value. The way that seems plausible is by looking at great pitchers who did earn that much value. Yes, they were similar to Santana at his age. But so were a bunch of other pitchers, and a whole bunch of them flamed out young. You know, it's easy to forget that Matt Clement still ranks 33rd in the all-time MLB list of qualifying pitchers for strikeouts per nine innings pitched. Certainly nobody thought that he'd miss half of his three-year contract to injury when he was signed by Boston in 2004, and that he'd pitch only three months for Boston without displaying the effects of what we now know to have been serious injury. Boston paid Matt Clement over $19 million these past two years to pitch 65.1 IP with a 6.61 ERA. Other teams have done similar things: Carl Pavano earned $18 million to pitch 11.1 IP at a 4.76 ERA these past two years; Pedro Martinez earned just over $14 million to pitch 28 IP for the Mets (but with a 2.57 ERA!) Mike Hampton earned $29 million the past two seasons not to pitch at all. Pitchers get injured. Speculating that Johan Santana might earn 150% of his career value to date through the rest of his career is a reasonable upper bound. As an assumption or a forecast, it's very optimistic. * THT lists AL starting pitchers by whole Win Shares. In each case where Santana projected to other than a whole number, I gave the names of the lowest higher-ranking pitcher and the highest lower-ranking pitcher. For 2011, where Santana projected at 12.0, I gave the two middle names on their list of AL Starting Pitchers ranked at exactly 12 WS.

You appear not to have read my words. I concluded my post with: "I would propose, instead, that we record our differences and move on, looking back, perhaps, if Melky Cabrera might either collapse or have a breakthrough season, at which point one or the other of us might clearly have been correct in our foresight." You respond to my suggestion that we wait to see who might be right with, "It is no fun having a discussion with someone who considers his opinion to be infallible." Dude, you're projecting. I state my reasons and suggest that we wait; you say that I'm claiming infallibility. :thumbdown Here's the deal: you made an opinionated statement regarding Melky Cabrera that flies in the face of the relative merit of his accomplishments when compared to all others who have played Major League Baseball at his age. Having done that, when confronted with the facts, you try to insult me. I'm willing to wait to see who's right. I'm not willing to give you the last word when your every post is an unresearched insult. Give it up, a700hitter...give up the insults, and acknowledge that you cannot see the future better than I can and that, our opinions both posted, we should wait and see who's right. Miggy Cabrera's top ten BR comparables after age 24 are either likely HOF-bound (Vlad Guerrero, Andruw Jones, Ken Griffey Jr.) or HOF enshrined, excepting Hal Trosky. Trosky retired due to migraines at age 28, although he tried a comeback 3-5 years later unsuccessfully. Miggy's top comparables at age 22--Melky Cabrera's age--were equally excellent, but with Tony C and Cesar Cedeno added to the mix. We know the story regarding Tony C; Cesar Cedeno had troubles of his own: CESAR CEDENO Arrested for: DUI, assault, Murder, beating his pregnant girlfriend, battery on a police officer Cedeno was called “the next Willie Mays,” and he justified the hype by hitting .320 in 1972 and 1973. But on December 11, 1973, he shacked up with Altragracia de la Cruz in a Santo Domingo motel. During the night, the woman was shot in the head and killed. Cedeno claimed it was an accident. (Accidents do happen—especially when a loving couple messes with a loaded gun.) The court reduced the charge to involuntary manslaughter, leveling a fine of $100. Obviously distraught over the whole incident, Cedeno kept playing in the majors until 1986. http://www.maximonline.com/articles/index.aspx?a_id=4459 I guess the moral is that only Miggy Cabrera can keep Miggy Cabrera from the HOF. It could be his appetite, it could be drug use, it could be injury...it couldn't be that he doesn't have the talent to be enshrined. Melky Cabrera isn't as good as Miggy Cabrera, even considering defense (remember, Melky's is excellent and Miggy's is atrocious, making them more comparable than one might expect looking just at hitting stats). Only about 30% of his comparables are Cooperstown-worthy. That is, however, 30% more than most players' comparables. Most of his comparables made the All Star team. That was, and is, my projection.

And that's why I used the word "prejudice" to refer to your perspective. You've made up your mind that Melky is a fourth outfielder, and you don't care about how he stacks up historically. Your opinion is that the comparisons don't "have much validity." You're not going to wait until a few years pass to make up your mind: you're sure. It does not trouble me that my opinion differs from that of a prejudiced individual.

But it's better than the others...it's better than Marcels...it's far better than just accepting the previous year's performance as the likely next year's perfrmance. Most here are assuming that Cabrera (and Santana) will never change. That's very unlikely. Young position players improve with age; pitchers already in their primes decline. :thumbsup: I knew that you knew baseball--I hadn't realized that you were a diplomat!

I had tried to avoid pointing out what you just did: I was several days ahead of the actual Crisp + Lester + Lowrie + Masterson/Bowden offer when I was suggesting Crisp + Lester + two, and I don't recall any of the posters who were ridiculing me at the time stepping up to say they were wrong. Regarding the contract, let's use a parallel from the Red Sox: did they increase Papi's salary to match Manny's? No--even after renegotiation, Papi's salary isn't even similar to Manny's, even though he's now the better hitter. He still makes only roughly two-thirds of what Manny does--he even makes less than JD Drew, who platooned with Bobby Kielty. If Santana makes more money than Beckett, Beckett can renegotiate for a raise by offering more years (as Big Papi did). He won't just receive a pay raise. ***************************************************************** a700hitter, why should my credibility slip before the time passes to prove me right or wrong? If you look at what I'm saying years later and it doesn't make sense, fine. But what you're saying here is that you disagree with me, so my credibility should suffer. That's far different. That would be defined as "prejudice."

Minor? Hey, you picked it, ORS, not me... A whole lot of established pundits have pooh-poohed forecasting by studying comparable players for years. What they don't do is to make their own forecasts to prove that they're better. Most years BP does a post-season study of how PECOTA ranked among the major forecasting methods. It's almost always at or near the top. The study was in BP Unfiltered this year if you want to check it out. The moral is that forecasting by comparables does work...at least as long as one looks at most-likely possibilities, not for exact certainties. And lack of progression...progression is rarely linear for any player, and Melky's September strongly looks as if it was affected by an unreported injury. He was hitting .296/.347/.426 before the slump, and .116/.192/.145 after September 8. That's not a slump: there's a one-in-2,800+ chance of a slump that bad (using binomial theorem baed upon BA). Something changed...it could've been illness, injury, or other factors, but Melky isn't a .116 hitter. I'm wondering if it's a cold-weather issue--his September was bad in 2006, too, except for a series with Boston played in comparatively warm weather, IIRC. But if I trust forecasting more than your own two eyes, let's agree to differ.

Jayhawk Bill- I think you are missing two important points here. First, you need to factor in the risk that the player never reaches his potential, second you need to factor in the element of time. Third, certain players are more valuable to different teams. In Beckett's case, his value maybe much higher today than it was in November 2003. Though he's being paid more and is closer to retirement, he's much more likely to have a great season in 08 than he was in 04 and is much riskier. He's also less likely to get injured as well. WRT your first point, once you've got a few years of MiLB/MLB stats you've got a very good idea of the career projection, even better if you add a bit of scouting and phenotype information. Your second and third points I certainly understand. In Beckett's case, his expected performance next year is higher than it was at any previous point in his career. His period of future obligation is less, though, and his salary, while reasonable, is significant. Look back at my quote that you cited. It's not just the expected wins the next year.

Do you know what RZR is and what it measures? FRAA makes assumptions regarding the pattern of balls in play given the pitching staff. RZR counts balls in play exactly. If you want to find a way Coco is better, you can check OOZ plays--he had 25 more than Melky in just 144 more innings. We don't know how close those were to his zone, though: we know that Coco's CF OOZ rate was barely better than Cabrera's in 2006, and that his stats from LF in 2005 look, at face value, significantly worse than Melky's LF stats in 2006 (.697 RZR vs. .915 RZR). Regarding The Fielding Bible, it uses three-year stats, something that I strongly question except when two players both in their primes, with no injuries, playing the same position each year, are being compared. That applies to neither Crisp nor Cabrera. *** The OOZ plays are the reason for the difference several systems cite between Crisp and Cabrera. If there were a multi-year pattern where Coco handled OOZ balls so frequently, I'd see your point better, but as it is we're talking a couple dozen balls caught in one year. Speaking of cherry-picking, though, the bit about Coco vs. Melky was part of a very long post from me. I take it that you concur with the rest? Or are you cherry picking?

What does have predictive value? You're obviously basing your words on something...what is it that's better than comparable players? And, FWIW, from age 22 to 23 Dustin Pedroia raised his BA from .191 to .317, while Gary Sheffield raised his from .194 to .330 at the same ages. I know that you're referring to Sheffield because he's the 20th of 20 comparable players on Pedroia's 2007 PECOTA Card with intent to ridicule, and you, I and the audience know that expecting 100% parallel development to every comparable is impossible--but it's striking how your example, chosen for humor, was actually a case where PECOTA captured a surprising jump by both players.

Yup. As I posted, though, we still won. Quote me or retract your post. You're changing meanings for the purpose of ridicule...you're better than that. Or at least I'd thought so...:dunno:

First came up...do you mean the 49 innings at AGE 20??? Were the nine games in 2005 that memorable??? See, in 2006 Melky was primarily a left fielder. He was the best qualifying left fielder in the AL by RZR, posting a .915 to lead his league. (For contrast, Manny was at .643, missing roughly four times as many fly balls as Melky.) In 2007 he was a center fielder, and by RZR he was second in the AL, at .910 a mere 11 points behind Curtis Granderson. Melky has played outstanding defense for two years. "Looked lost and confused???" Not with those stats. He's hitting .275 over two years and he's as good defensively as Coco Crisp, who's hit about ten points lower than Melky with equivalent power. How has he not earned his position--he's played as well as the CF on the World Champion Boston Red Sox? No. Lineup protection may change the type of offensive production from walks to hits, but there no study that shows that being surrounded by better hitters makes a difference in offensive value. Two points: 1) Ellsbury's MLB sample size is quite a bit smaller than Melky Cabrera's, and he's a year older; and 2) I wasn't discussing Cabrera vs. Ellsbury, I was discussing Melky's likely value based upon career path.

Melky Cabrera is most likely to develop into an All Star. He has a significant chance, albeit certainly not a probability, of being in the Hall of Fame. That doesn't mean that his work at ages 21 or 22 has been All Star caliber. But most All Stars aren't All Stars as 22-year-old rookies, they're in the minor leagues. Melky Cabrera has been in MLB for two years already and he's still a year younger than Ellsbury. That suggests very significant talent.

Where in all of this do you consider these two points: 1) Johan Santana will cost $18 million or more per year, while Ellsbury is almost free; and 2) Santana's salary will be guaranteed for 5-7 years, but he's roughly 15% likely in each year to have a career-altering/ending injury? Considering those factors, I'd request that you'd explain again why Santana is significantly more valuable than Ellsbury. I think that you and example1 both make good points, but I think that you're seriously undervaluing Ellsbury.

Perhaps your belief that IsoD is not typically a skill which one learns is influenced by the normal career progression where players usually develop in college for a few years, MiLB for a few years, and arrive in MLB at around age 25. Melky Cabrera is barely 23 years old: he's still developing his core skills, even if he's already good enough to have reached MLB. But let's look at players who ARE similar to Melky Cabrera, players who reach MLB while young. Going to Baseball Reference and taking the ten players most similar to Melky through age 22, and then taking the average batting line of those ten from age 23 through the end of their careers, one gets this batting line: .295/.365/.435 Which is exactly what I'd posted below as a reasonable expectation for Melky. Let's look at PECOTA comparables for a double-check. Melky's top comparable coming into 2007 was Rick Manning. Rick Manning didn't develop IsoD. The next four comparables were Carlos Beltran, Roberto Alomar, Richie Ashburn and Pete Rose, all of whom certainly DID develop IsoD. Many of the other 15 comparables did so as well, with what seems at a glance to be a correlation between developing their power hitting and developing IsoD. FWIW, 30% of the players on both the BR comparables list and the BP comparables list were HOF-caliber players. There are a few who washed out, but it's more likely that Melky will be voted to the HOF than it's likely that he'll wash out as you describe. A most-likely scenario has him developing into an All Star outfielder...and All Star outfielders are valuable players.

This looks to be a win-win trade to me. The Mets are in "Go for It Now" mode, as they should be. For a guy a few years away from his prime, they acquired a batter lethal against right-handed pitching and the best defensive catcher in the National League. The Mets are more likely to reach the NLDS in 2008 because of this trade. The Nats, who won't contend in 2008, have acquired an outfielder who'll look pretty darn good by 2010. Even next year they're going to feature Wily Mo Pena, Lastings Milledge, Austin Kearns, Ryan Zimmerman and Dmitri Young in the same batting order, and the Nats probably aren't going to be cellar-dwellers any more. Well done by both sides, IMO.

Melky Cabrera just won't be a big offensive player? It's easy to forget that Melky Cabrera is almost a year younger than Jacoby Ellsbury. He's going to develop more than Ellsbury because of that: you're looking at a career average maybe 25 OBP points and 50 SLG points higher than he is right now, with a peak value that much higher again in his good years. A good defensive outfielder who hits .295/.365/.435 most years and .315/.390/.485 his peak years is one of the better hitters in the league. Regarding the pitching talent, I hear echoes of Joe Sheehan's piece, written while it looked as if Santana were swiftly heading for Boston. Sheehan is a Yankees fan, and he'll do anything to belittle the Red Sox or, in this case, the quality of their offer. I disagree with his logic this time. I agree that the Twins have pitching prospects, but they're not MLB-ready for the most part. Sheehan ranks Anthony Swarzak as the Twins' fourth-best pitcher. Swarzak missed 50 games via suspension for drug abuse this year, and he's not projected to reach MLB until 2009 IF he can keep his nose clean. Kevin Slowey and Boof Bonser are both ranked ahead of Swarzak...look, Bonser and Slowey will be capable 3/4 starters, but these aren't the top names of a strong pitching staff. Hughes is such a name, and Hughes would rank with a healthy Liriano as the best pitchers for the Twins once Santana were gone. Phil Hughes is a big asset for Minnesota--that's why they demanded that he be included! Crisp, Lester, Lowrie and Masterson/Bowden are a very valuable package of talent. I just see Hughes and Cabrera plus one or two others as exceeding the value that Boston has offered.

Not really. The Yankees' offer is stronger with the inclusion of Hughes.

Defense matters. Defense matters big-time in the middle infield, and Mazeroski was perhaps the best at turning a double play in MLB history. Bill Mazeroski is admittedly not an inner-circle HOF player, but he's very far from the three least worthy.