The Blue Jays, Phillies, Mariners, and Athletics put together a blockbuster trade that has rarely been seen in baseball history: nine players will belong to new organizations next year, including two former Cy Young winners very much in their prime.

The aces need little introduction. Roy Halladay will leave the organization that drafted him over fourteen years ago, after throwing over 2000 innings in their uniform-all in the regular season. Cliff Lee will join his third team in a calendar year, just five months after being traded from the Indians to the Phillies, and less than two months after winning four playoff games.

Elite prospect Kyle Drabek, one of the top 25 prospects in the game according to Baseball America, and outfielder Michael Taylor, also listed by Baseball America in the top 25 and in the top 50 by any standard, will both be playing for different teams next year. The Phillies traded both of them to the Blue Jays, but Taylor was shipped to the Athletics for Brett Wallace, currently a third baseman, but one who will likely move to first base or designated hitter at some point. He is also close to the top 25 and certainly in the top 50 depending on whom you ask. Catcher Travis D’Arnaud, another solid prospect, will also be headed to the Blue Jays. The Phillies sent Cliff Lee to the Mariners for three of the Mariners top prospects. The Phillies will receive Phillippe Aumont and J.C. Ramirez, perhaps the Mariners’ top two pitching prospects, and Tyson Gillies, a toolsy outfielder, in return. Aumont should be in the mix for making the top 100 prospects in the game, with Gillies a bit behind him, and J.C. Ramirez a step behind that in value due to how raw he is and how much risk involved in a pitcher that far from the majors. That is not to say Ramirez’s upside is lower than the other two prospects, but his expected value is probably lower.

Additionally, and very importantly, the Phillies received $6 million from the Blue Jays to pay for some of Halladay’s 2010 salary ($15.75 million) and Halladay received a three-year, $60 million contract from the Phillies with an option at $20 million for 2014 that will vest if he pitches either 225 innings in 2013 or 415 innings in 2012 and 2013 combined. Supposedly, there is a similar vesting option for 2015 as well. That deal, absurdly below Halladay’s market value, is what may make this deal a good one for the Phillies, although it is hard to see the point of trading Cliff Lee when the Mariners gave the Phillies far less value in return than Lee should have been worth. The Phillies are the only team that does not clearly win in this deal, but I’ll return to them in detail after discussing the other three teams.


The 2010 Mariners will almost definitely win more games as a result of this trade, and the prospects they surrendered were far from sure things. Although it may seem odd to think of the Mariners as a win-now team, especially after winning just 61 games in 2008, they truly are a team poised to strike at the postseason. No team in the majors improved by more in 2009 than the Mariners did-they won 24 more games than they did in 2008, and finished 85-77.

This was not a fluke. In his short tenure, GM Jack Zduriencik has turned the Mariners around at the major-league level. His initial primary focus was improving the team’s defense, and he certainly did. The Mariners were first in the majors in Baseball Info Solutions’ Total Runs Saved in 2009, first in Mitchel Lichtman’s UZR, first in park-adjusted slugging on balls in play, and first in the American League in PADE.

This defense will surely help Cliff Lee, who walks few batters and strikes out more hitters than the average pitcher, but still allows some contact. Cliff Lee was not an ace in 2007, putting up an ERA of 6.29 in 97 1/3 innings, but he adjusted his approach. He became more of a ground-ball pitcher after being an extreme flyballer earlier in his career, and cut down on his walks while striking hitters out even more hitters than before. Especially in Safeco Field, Lee will only give up significant runs if a lot of hits drop in and the Mariners’ excellent defense is primed to preclude this from happening. Lee is also left-handed, and as Safeco is particularly punishing to right-handed hitters who Lee will see more often, he’s supremely well set up for Seattle.

Of course, Seattle’s defense is good enough to help a lot of different pitchers; the end result is that Lee will probably put up an ERA slight north of 3.00 and throw a good 200 innings, which is worth about 5.25 wins above replacement given the Mariners’ stadium and defense. This is worth a lot of money; for the life of free-agent deals, recent research by Sky Andrecheck suggests that teams will probably pay around $27.6 million or so for this kind of production (though they tend to get better deals in the first year of contracts than that). As a team currently in striking distance and in a pretty large market, the Mariners are the exact kind of team that signs free agents and should be able to get that kind of financial value from him in expectation.

The prospects they gave up are worth far less than that. Research by Victor Wang a couple of years ago approximated the win values of John Sickels’ Top 10, 11-25, 26-50, 51-75, and 76-100 prospects, as well as some lower-level prospects, and adjusting some of the dollar values of what he did we can get a pretty decent estimate of the prospects. Values like these can be pretty noisy and given that there may be more precise estimates than Sickels’ rankings, we can adjust them accordingly, too, but the result is clear: the Mariners made out like bandits in this deal.

Phillipe Aumont is a borderline top 100 prospect, probably just outside this range. Pitching prospects ranked between 76 and 100 on Sickels’ list were worth about 2.9 wins according to Wang’s research, and those listed as “Grade B” prospects (the next category) were worth about three-quarters of this value. Wang’s method entailed discounting future wins and then inflating the dollar figures by historical inflation rates, along with a few other adjustments. Rather than follow that convention, I will simply keep the dollar value of wins constant without discounting the wins. What I will do is take the cumulative production in wins that the players produce in those six years, adjust for their salary costs. As I stated in “How to Make up a Good Trade Rumor”:

That player is going to get the league minimum for three years and three years under salary arbitration where they will get salaries of approximately 44 percent, 61 percent, and 64 percent of their market value. … Therefore, getting six years of a player only requires “paying” for 1.7 years of their actual value. So, it’s like getting 4.3 years of their value for free.

Thus, approximately 72 percent of a player’s win values are like free talent. Thus, if Aumont is worth between 2.2-2.9 wins in expectation if he were free over those six years, he is worth about 1.8 wins for free.

Similarly, Tyson Gillies seems to fit more snugly into Sickels’ “Grade B” prospect range and thus is equivalent to getting 2.2 wins, costing 28 percent of this, and being worth a net value of 1.6 wins. J.C. Ramirez, a high upside talent even further from the majors may fall into the “Grade C” range Sickels’ used which would make him worth about 0.6 wins in total, and given costs, worth about 0.4 wins net.

Thus, these three prospects amount to about 3.8 wins for free over the coming years. Naturally, it is unlikely that much of this value will come in 2009. Instead, the value will trickle in maybe across 2010-2018 or so.

Lee, as I mentioned, is worth about 5.25 wins to his team this year. As he only costs his employer $9 million, which would likely purchase about 1.7 wins on the free agent market, he is worth about 3.55 wins if the Mariners got no compensation for him. However, Nate Silver approximated the value of draft-pick compensation for Type-A free agents at about $9 million a couple of years ago, when the value of a win was slightly smaller in dollar terms, which could presumably guys about 2.0 wins. Thus, Lee is worth about 5.55 wins.

Here we can see that Lee basically added about 3.55 wins this year and 2.0 wins over the coming years, and the prospects subtracted cost the Mariners about 3.8 wins. The net total of wins is therefore +1.75 wins, a definite gain for the Mariners.

The only hold up would be if the Mariners were not a win-now team, but the fact is that they are. Even though they were a disaster as recently as 2008, their current minor league system is in the bottom half of clubs, and their current major league club is in the top half. The team is not all that young either. Felix Hernandez is turning 24 this April, but he will be far more expensive in a couple years if the Mariners retain him, and thus now is the time to capitalize on him. Their best hitter is Ichiro, 36, and their best other player is arguably the league’s best defender, Franklin Gutierrez, who will have turned 27-right in the thick of his prime-by the time the season starts. Chone Figgins will be a key contributor; he’ll be 32 this season. Their great glove man at shortstop, Jack Wilson, will be 31 this season, and their power-hitting second baseman, Jose Lopez, is 26 and entering his prime. Ian Snell and Ryan Rowland-Smith are 28 and 27, respectively. This is when those added wins from Lee may be most valuable to the Mariners, so this was an excellent move as a result. Zduriencik has been excellent in his first fifteen months as GM of a decimated club, and the Mariners may be the division favorite in 2010, especially with the Angels subtracting John Lackey and losing Chone Figgins to the M’s. The Mariners obviously improved here.

Blue Jays

The Blue Jays also improved in this trade as well. As I documented in detail last July here and here, the Jays absolutely had to trade Halladay last year to give a team his full value for two pennant runs, and Ricciardi blew it. Alex Anthopoulos improved matters by taking care of that old business.

Jays fans are understandably attached to Doc. He joined the organization in 1995 and has been their ace for many years. They have not seen a playoff team during his tenure in Toronto, but the one thing that fans knew was that when Roy Halladay pitched, they were likely to see a win. Losing Halladay takes that away from them, and you have to feel bad for them.

But the fact of the matter is that any Jays fan who remembers 1992 and 1993 is simply not going to compare the feeling of a championship with a nice stroll out of a park having watched a sub-.500 Blue Jays team win a Halladay-pitched game by a score of 3-1 on a September afternoon while fifteen games out of first place. Anthopoulos kept his eye on the ultimate prize, and even got extra value thrown into that future by putting money into the deal now. Drabek is a top pitching prospect, and pitchers ranked between 11 and 25 added 4.4 wins on average over their six years of service time. Fortunately for the way Sickels’ ranks prospects and the way Wang computes their value, pitchers ranked between 26 and 50 also added 4.4 wins. Obviously this may be a rankings issue, but with Drabek somewhere near the middle of that pack, assuming he is worth 4.4 wins sounds about right. Given the 28 percent cost I explained earlier, this means Drabek is worth about 3.2 wins to be added in the future.

Michael Taylor is more likely in the 26-50 range of hitters, who averaged 6.4 wins over their first six seasons. Some have Taylor higher than that, but hitters ranked in 11-25 range averaged 7.9 wins. Discounting these values by the 28 percent cost of those arbitration payments, and we get about 4.6-5.7 wins. Let’s call it 4.8 to pick a number.

Travis D’Arnaud is likely outside of the Top 100. Hitting prospects given a “B” grade according to Sickels were worth about 2.3 wins seemingly, and discounted 28 percent, this gives us about 1.7 wins.

Adding this up, we see that the prospects that the Blue Jays got are worth 3.2 + 4.8 + 1.7 = 9.7 wins. These are wins that come in the future when Anthopoulos may turn them into a competitor and those wins may push them over the top. Now, Halladay’s wins simply show up on the flatter portion of Nate Silver’s curve of a the marginal value of a win.

Halladay is a 6.25 win pitcher who costs $15.75 million this year-good for three wins on the free agent market. Thus, Halladay adds about 3.25 wins more than his salary would buy the Jays on the free agent market. He would also have netted about two wins as a result of Type-A free agent compensation picks that would come down the line.

The Blue Jays did agree to send $6 million to Philadelphia to cover some of his 2010 salary, which may have added about 1.15 wins this year had they spent that money on a free agent, so they really cost themselves 4.4 wins this year to net those future wins.

Thus, the Jays traded 4.4 wins in a useless year for about 7.7 wins when they may be competitive. How did they do it? Well, as I will show, the Phillies got a pretty good deal on Roy Halladay because of the extension he signed. The contract he signed was an incredible bargain and the money saved by paying him less than his value on the free agent market for future years is worth several wins if they spend it on free agents, and this was because Halladay wanted to play for the Phillies. He knew he was leaving money on the table, but he wanted to come to the Phillies and the Blue Jays would not have gotten much more than the 4.4 wins of value if they traded him to a team where he did not want to sign a below market extension. The fact of the matter is the Blue Jays maximized his value by trading him to a team where his value exceeded 4.4 wins.

The Blue Jays also spent $6 million to contribute to his salary, and this shows their future commitment to winning. This money, spent on free agents, would add barely a win to this year’s sub-.500 team and would be a waste. By paying the Phillies this money, they effectively redeemed it for abut 1.14 wins worth of talent from the Phillies as measured in prospects. Had they not done so, they would have gotten lesser prospects back and had to spend inflation-adjusted $6 million more on free agents later to get the same value. Anthopoulos is already spending money on a future winner, and this demonstrates the Blue Jays commitment to winning.

Swapping Taylor for Brett Wallace is tough to calculate-when two prospects of generally equal value are swapped, it is pretty much a wash. They are betting on each other’s guys, and perhaps improving at positions of need.


Oakland is both easy and impossible to analyze. Prospect for prospect is not my area of my expertise, and I will leave that for others to decide. They certainly took a bet on a guy they liked a lot in their outfield, which certainly is an area of need. No loss and likely a gain for Athletics here.


The only team that does not necessarily come out ahead in this is the Phillies. The team really made two separate trades-three prospects for Roy Halladay conditional on an awesome extension, and Cliff Lee for three inferior prospects. The former will turn out to be a success and the latter will not.

It is pretty easy to see that the Phillies lost the wins that the Mariners gained as a result of the Cliff Lee trade, and Lee is not going to give a discount to the Mariners just as he was not going to give a discount to the Phillies. The results above show that the Phillies will lose 3.55 games more this year as a result of this move and gain 1.8 wins in subsequent years.

The Halladay deal was a success but it came at a huge cost. Those three prospects above were estimated at 9.7 wins of free talent, as they would generate 13.5 wins at the cost of about 3.8 wins. This year, the Phillies will pay just $9.75 million for Halladay to generate them about 6.25 wins. That $9.75 million would only net them about 1.85 wins on the free agent market, meaning that they added 4.4 wins to 2010 as a result of this deal.

Halladay’s extension provides a huge bargain as well. If Halladay were to be able to auction off his 2011-2013 services (we will ignore the vesting options momentarily) in December 2009 and require no commitment beyond 2013, he would get probably $95 million for those services.

More specifically, if Halladay is a 32-year old pitcher now who is worth 6.25 wins, then it is reasonable to assume that he will be worth about 5.75 wins in expectation in 2011. Purchasing this on the free agent market, assuming a modest 8% salary inflation rate in salaries as the economy improves, would cost about $32.6 million. Halladay saves the Phillies $12.6 million, which nets them 2.2 wins in 2011 if they spent it on free agents (again assuming eight percent inflation in salaries).

If he drops down to 5.25 wins in 2012 and salaries inflate eight percent again, then those wins would be worth $32.1 million on the free agent market. Again, he will cost only $20 million, thus freeing up $12.1 million to be spent on about 2.0 wins. If he drops down to 4.75 wins in 2013 and salaries inflate eight percent again, then those wins would be worth $31.4 million. The same math says he will save the Phillies enough to net them about 1.7 wins.

Thus, even without considering the vesting option, we have the Halladay trade (before prospect cost) giving the Phillies about 4.4 more wins in 2010, 2.2 more wins in 2011, 2.0 more wins in 2012, and 1.7 more wins in 2013. They also cost themselves about 9.7 wins for Taylor, Drabek, and D’Arnaud, much of which will be spread across 2011-2013 as those guys are nearly ready for the majors now and would generate most value in the first three years. Thus, the net gain for 2010 is 4.4 wins, the net loss combined for 2011-2013 is 3.8 wins. They also got themselves some interesting vesting options. Cot’s Contracts lists the 2014 vesting option as being $20 million only if he either pitches 225 innings in 2013 or a combined 415 innings in 2012-2013. Although I projected him to be worth abut 5.25 and 4.75 wins in those years respectively, and using the same half-win a year back of the envelope approximation, about 4.25 wins in 2013, this is not the expectation if we already know he has thrown so many innings in 2012 and 2013. If so, the odds are that he has aged very gracefully and may be worth even more than 5.0 wins in 2014. With all that inflation, those five wins would cost $35.7 million and at $20 million, this would be $15.7 million in expectation. Will that happen? Estimating maybe 1-in-4 odds, just to make up a number, would make it worth about 0.5 wins doing the same salary inflation adjustments. The 2015 option would probably be close to 0.3.

All in all, this means that we should add almost another win of value to this. They effectively got a net value of 1.4 wins over the next decade as a result of this trade, assuming that the Phillies do not offer arbitration in 2016 (or the player’s union gets rid of it in the next CBA).

Halladay plus two wins worth of free agent compensation (or no bargain on the extension) would have been worth about 6.4 wins at a cost of 9.7 wins, and lost the Phillies 3.3 wins. Instead, it netted the Phillies 1.4 wins and the Blue Jays 3.0 wins. Thus, both teams benefited and the Jays bargained for more of the surplus beyond trading Halladay to a team where he would have played out his contract. However, this is a gain for the Phillies on the Halladay end.

However, it is a loss of 1.75 wins on the Lee end of things, and thus, the whole trade is worth about negative 0.35 wins overall. Is this good? No, it is not. You are supposed to gain from trades. However, it is hardly an apocalypse. Some time, maybe in 2014 or so, the Phillies will surrender a grand slam in the first inning of a game that looked like a coin toss. They also shuffled a few wins around over the next decade, too. But the fact is that the net effect on the Phillies is pretty neutral. If you value current success more than future success, it may even be a win-or more likely, if any of my obvious ballpark estimations are wrong, this deal could be a win. The problem is that they just did not gain as much of the surplus as the other teams.

The reason was the Cliff Lee trade. The reason the Phillies did not go with a three-headed monster of Halladay/Lee/Hamels that the city clamored about was either about budgetary reasons or about re-stocking the farm system, depending on who you believe. They lost wins on the deal, so re-stocking the farm system at seventy cents on the dollar seems like a bad reason.

The justification of going down in budget itself, I actually understand. The Phillies are currently huge favorites to win the NL East in 2010, though at least the Braves should put up a fight with their pitching staff. The marginal value of a win, as seen in the Nate Silver article linked above, is high right around the point where the win is most likely to push you over the edge of making the playoffs. It flattens out after that and has less use. If I were managing the Phillies books, I probably would have put a $140 million cap in the current situation as well. That is probably how they maximize profit, and as an economist myself, I see the validity of that. Roy Halladay clearly does not his maximize monetary profits at the expense of his family’s preferences, and I commend him thoroughly for that and hope that I would do the same, but it is not something I expect from people, certainly not owners of a business worth hundreds of millions of dollars.

I don’t even take issue with trading away Lee due to the three-headed monster rotation of Halladay/Lee/Hamels. The Astros in 2005 lost the Series with Clemens/Pettitte/Oswalt, and even a far superior set of hitting teams like the Braves in the Maddux/Glavine/Smoltz era only put a ring on one finger of each of their hands despite being in place for a decade. Baseball Prospectus research in Baseball Between the Numbers shows that the three things that are even weakly linked to post-season success include strong defense, strong bullpen, and a strikeout-happy rotation-but not top of the rotation starters per se. This is not a reason not to do the deal.

The issue I take is that they could have non-tendered Joe Blanton (or, better yet, traded him for some inferior prospects) and Chad Durbin and won two or three more games in 2010 at the same $9 million salary. I say that even as someone who believes Blanton is vastly underrated, but they could have gotten better value by doing that. If Blanton was worth nothing on the trade market (which I am skeptical about), then the Phillies were foolish to offer him arbitration-clearly no one else would have even offered him his arbitration estimate on the free-agent market, so why pay him that much? Figure out what other teams would pay him and offer him that and a cookie.

The fact of the matter is that the Athletics made a decent gamble, the Blue Jays made a deal that absolutely improved their chances of winning a World Series later this decade, and the Mariners did too. The Phillies are the only team that looks like they may be more likely to have lost than won in this deal. By much? No. But why not hold on to Lee and save budget another way later? Then a win-neutral move means that they can enjoy the fruits they got from one of the top pitchers in the game handing them about $ 36 million in discount.

You need to be logged in to comment. Login or Subscribe
Good points, Matt. I would have enjoyed the Doc/Lee/Hamels trinity, but guess I can't blame the Phils for thinking longer term and getting some AAA-level talent replenished. I was surprised Toronto didn't grab Aumont, was sure that would happen when I heard Seattle was involved - kudos to the new GM for not making that move just to grab some easy local goodwill
Excellent work, Matt.
Great article Matt, enjoyed the read.
I haven't read up on stats as much as others on here, and this is the first time I've come across UZR. I thought a better fielding metric was interesting, so I found a couple articles on it along with 2008 and 2009 UZR totals on fangraphs, and I'm curious about something. The #2 UZR/150 player in 2008 was Mark Ellis at +23.7. Ellis was at +2.4 in 2009. A few others fluctuations at the top of the list: Rollins: 15.0 to 2.9 Fukudome: 13.1 to -18.1 O. Cabrera: 13.1 to -13.7 In some of these cases, players jump from Top 20 one year to Bottom 20 the next. Are fangraph's numbers reliable, and if so, is this stat reliable?
UZR is one of the best fielding metrics available. Total Runs Saved (based on +/-) is another excellent metric. Fielding metrics remain in their infancy and there remains a lot of noise surrounding them. Looking at UZR over a longer time span gives more insight. It is also possible that defensive skill is very streaky. Regardless, I always prefer to throw a few metrics up there to see what sticks when it comes to defense.
Fukudome's 2008 to 2009 is comparing RF to CF which are very different positions.
And Ellis injured his throwing shoulder at the end of 2008 and his calf midway through 2009. That definitely would impact any defensive metric. So I would say these swings are no comment on the effectiveness of the stat.
Ellis was not the same guy in 2009. That was clear to me even without looking at the numbers. There was no question he was among the best in the game a couple years ago, but he didn't look the same in 2009. That said, yeah, even using a good stat like UZR, defense is difficult to project.
Its actually kind of incredible to me that the Jays weren't able to get more out of the phillies for the deal. If they don't trade him the phillies don't get another chance at such an incredible deal. The union would have rejected halladays deal if it went through on the open market. So that should have given the Jay's a ton of leverage. Also, as to the NL East, this site for one says that on third order wins the Braves were the best team in the NL this season. Yes the phillies are to have a full season of a much better pitching staff, but the NL East could end up being the second best division this season (let's not even talk about the declines that Werth Ibanez and Howard are sure to have); I bet the phils will miss Lee in the end.
If the Jays had traded Halladay to another team, he would not have been worth as much. He claimed even before the trade that he wanted to go to Philadelphia all along, largely due to their Spring Training facilities being nearly in his backyard. He would not have signed this extension with, say, the Angels. Had he, they would have surrendered prospects worth their value to him which would have been the same 4.4 wins as stated above (assuming the $6MM Jays subsidy), plus another 2 more for draft picks to reach 6.4 wins, because that was his value to other teams. The Phillies value to adding Halladay as described above was 11.1 wins. Thus, there was 4.7 wins of surplus value. The Phillies only got 1.3 of these 4.7 wins. It is just as easy to say "The Jays could have gotten more" as it is to say "The Phillies could have surrendered less." The 3rd order wins that we have are a way of approximating the value of performance as measured by plate appearance outcomes. It is not something that can be extended forward without considering other forms of luck. The Braves, quite simply, had an 8.1% HR/FB this past season, which is not something that they can expect going forward. Pitchers fail to show persistence in this statistic and the Braves play in a pretty neutral stadium. 3rd order wins are a way of eliminating luck on run distribution among games, hit distribution among innings, and opponent difficulty. There are other forms of luck, including BABIP and HR/FB luck for example. Had the Braves had neutral home-run luck, their 3rd order record would have looked more like their regular record. Breaking down the player by player projections this season, the Phillies are very likely to come out ahead. The decline of Ibanez may be on its way, but projecting Werth and Howard to decline considerably is just ridiculous. Werth only had a "career year" because he got more playing time. Howard has had the same season plus or minus some BABIP luck for about three years running and this season he got in shape and improved his defense. They will be 31 and 30 this year. That's not a time to fall off a cliff.
Matt, I don't think it's a great idea to take Victor Wang's argument, which is a useful heuristical way of determining prospect value, and treating it as if it is literal, as you seem to do here. All of the prospects involved in this deal are lottery tickets, even if some look like they're more likely to pay out than others. But there's no way of saying that prospect X will return Y value with any kind of certainty, so it seems unwise to analyze the trade as if you can. I would hedge a bit on the certainty expressed here, and simply say, if you'd like to, that it *seems* like the Phillies undersold on Lee.
This is all expected value as this type of analysis always is. The point is to value it. That's like casinos as a business because theoretically, the house could lose. the point is to play the odds. Wang and I both expressed this as historical tendencies indicating probabilities indicating expected value. The Phillies undersold on Lee in the sense that they could have gotten prospects for him that were more likely to succeed than the ones that they got.
While it is expected value analysis, a more interesting way of doing it might be to take the true distributions (not just the EV) and add them up. Because the value of prospects is very skewed. So most prospects end up worth nothing or close to nothing. A few end up incredibly valuable. In addition to knowing the EV it might be interesting to know the median value expected or the distribution of value expected. Like maybe the Jays got +3.3 wins on average, but is it a case that most of the time the Jays lose 2 wins but 20% of the time they net +15 wins? Their spread is likely quite skewed like this. The Philies are skewed the other way. While they might be neutral or even -EV on the trade, most of the time they probably win because odds are the prospects underproduce their numbers (it is just that when the Philies lose because the prospects do develop into true studs then they lose by a bigger amount). So the Philies probably have less risk now. Assuming everyone hits their EV is a reasonable SWAG, but doesn't give the full picture. And for most teams a player who is worth 3 wins a year is more valuable than 2 players who are 50% likely to be worth 0 wins and 50% likely to be worth 3 wins.
I doubt the EV approach is too far off base here, because obtaining 3 prospects hedges the risk from the Jays point of view. I think this is why you almost always see 2 or 3 prospects as the return for a major leaguer, regardless of the caliber of players involved. Of course this argument is null if you believe your last sentence, but i don't see why that would be true, especially when you're discussing prospects that can be stashed in minor league systems for little cost.
All good points. Victor Wang's article linked above did a terrific job at establishing rough distributions. He got win values per year for the first six years for sets of prospects with certain grades in different years, and reported the percentage that were in certain ranges. Lately, I've been thinking variance is good for baseball teams who aren't the Yankees and maybe the Red Sox. Think about it. Two teams have expected win totals of 85 wins. Do you want to be the one who has an 100% chance of winning 85, or a 50/50 shot at winning either 95 or 75? One way you make the playoffs less than 1% of the time and the other way about 50%. Of course, if your expected win value is 95 wins than maybe you want less variance. Thanks for the post though. This is the type of topic that should get raised to extend this expected value trade analysis and usually people take it as given. I tend to think the overall variance ends up similar since draft pick compensation takes up part of Cliff Lee's value and those picks have more variance than the prospects dealt, but it's all a good point.
So Cliff Lee is worth Blanton, Durbin and 3 prospects? I want to see what Pecota has to say about that. And if you're suggesting that Halladay could've expected a 3-year, $90 million contract at the end of this year ... well, then I don't think much of whatever data you're inputting to wind up there.
Was this reply just the words "I disagree" without any coherent reason. Why do you think Cliff Lee is worth less than those players? And why on earth are "3 prospects" treated like some sort of currency that you can just counted them up like dollars? Different prospects have different values. Blanton is someone that I believe that could have gotten fair value for if they waited-- about 2-2.5 wins by most estimates. Durbin is replacement level, and the "3 prospects" are not Top 100 prospects and are probably worth 3.8 wins combined. Even if they could have gotten any one of those middling prospects for Blanton, they would have gotten better value. If the market valued Blanton so much lower due to some sort of injury likelihood or something I'm missing, then it would be even more likely Lee was better. The reason 3/$90MM seems extreme to you is because you need to think about the details of deals. The reason is that you would think that someone like Halladay would get 5/$125-130 or something on the market. Thus, a lower AAV. The thing is that most free agents have accelerating salaries and decelerating value. If Halladay had only sold his first three years of free agency, they would have been valued similarly to how Sabathia's and Santana's were based on their trajectories at the time which would be about 3/$90MM.
I also wonder how much of a factor Lee not demanding the ball when his team was down 2-1 in the World Series played in this one.
Fantastic article, Matt. This is one of the better reads I've seen on BP in a while.
I enjoyed the article. But your analysis overvalues Lee. There is too much variability/deviation to his season-to-season performance i.e. there is a much greater chance Lee will have a "clunker" of a season next year than Halladay - the small difference in expected wins doesn't account for this. Also, your inflated valuation of what Halladay would receive on the open market for 2011-2013 ($95 million or $31.67 per year) also skews the analysis.
I've used a few sources on approximating Lee's value, and based on his huge change in peripherals starting in 2008, it is certainly a reasonable approximation to put him a 5.25 wins above replacement. I've seen higher and lower ones. The reason $31.67/year seems high is that you are used to seeing contracts that are over longer periods of time, typically a couple years into a decline phase. The trade Halladay might expect on the open market is probably 5/$130MM or 6/$145MM, but even though the salary escalates, most of his value comes early in the deal when he is younger. A team paying Halladay $140MM over 6 years is probably paying him $95MM for the first 3 years and $45MM for the last three years even if they are paying more of it out later in the deal. Also keep in mind that was assuming no draft pick compensation, which would have made it closer to 3/$86MM, as part of a 6/$136MM deal.
Matt, without devaluing your exhaustive article, which I think is up to BP's highest standards, the really difficult issues here are those that I don't see as amenable to BP type analysis- or if they are, no one's done it yet. I think they are the sources of the uncertainty that prompted Amaro to trade Lee. FIRST, how will Cliff Lee perform in 2010, having pitched 270+ innings in 2009? His past career does not shine any bright lights on that question. SECOND, what would it take to retain Lee at the end of 2010, and could the Phillies have paid it? On this latter question, I think the answer is also unknowable, unless you already know how Lee will perform in 2010? Lastly, a rotation consisting of Halliday, Hamels, Happ, Blanton and Moyer would not cause any NL manager to consider slitting his wrists. So, maybe, Amaro is right? We will find out; that's why you play the games. Regards,
Peter, thanks for your comments. The fact is that this is all expected value analysis. You simply cannot predict anything with certainty. Even if you knew every player's exact talent level and exact number of days that would be spent on the disabled list, you would still be off the average teams win total by about 6 or 7 games. However, doing expected value analysis is the only way I see to establish a guess as a starting point, unless you can somehow unravel the trillion different ways that the next year could play out. As far as what it would take to sign Lee, the answer is about market value which is enough to do the analysis. Check out my article from last month 'How to Make up a Good Trade Rumor.' In there, I explain that you can expect that if a team retains their own player, the surplus value will be approximately the two picks the team would have gotten for him if he left due to the bidding process. The Jays only got so much from the Phillies precisely because Halladay's willingness to effectively pay $36MM to live in Philadelphia for the next four seasons made him more valuable to extend than the picks would be worth.
Matt, I'm slightly confused. For the 2010 season, the Phillies have Halladay and no longer have Lee. I don't see how the prospects affect their win total in 2010. If the Phillies had traded Halladay to the Jays straight up for Lee, would that equate to more wins for the Phillies in 2010?
I'm not sure which part of the article you are referring to. I see the trade as a net positive for the Phillies in 2010 probably. Keep in mind, though, that Drabek and Taylor are both nearly major league ready and both may see time in the big leagues in 2010. I'm not sure that it's as trivially positive as you'd think.
There's a small problem with your wins vs. free-wins method. It assumes that an extra win is revenue-neutral, and we know for a fact that that is not true.