Batter's Box Interactive Magazine - Dissecting The Deal

Dissecting The Deal - Freddy Garcia to the White Sox

I'm trying a new kind of trade analysis out today. Yesterday, the White Sox sent Miguel Olivo, Jeremy Reed and Mike Morse to the Mariners for Freddy Garcia and Ben Davis, plus an undisclosed amount of cash.

What I decided to do, is to look at how the various players' PECOTA projections compare, for the length of time that they are contractually bound to their new teams. The hidden assumption here is that once a player is a free agent, they will effectively be of zero current value because their future value will be determined by the market. There is a small amount of resdual value to be gained by having sole access to bargaining right vis-a-vis the player while he is under contract; but agents are increasingly wise to the game and tend not to sign many below-market deals before a player hits the open market.

Freddy Garcia will be a free agent at the end of the season. So the White Sox get 55% of Garcia's season. Ben Davis will be a free agent (I believe) at the end of the 2005 season. So the White Sox get 55% of Davis's season and his 2005 season.

Miguel Olivo will be a free agent after the 2008 season. So the Mariners get 55% of Olivo's 2004, plus his 2005-2008 seasons. Jeremy Reed is not currently in the major leagues; presumably the Mariners will wait until Reed has no chance of being a "super-two" to call him up, and will get six full seasons of Jeremy Reed, 2005-2010. Plus, if we assume he's a September callup, 15% of Reed's 2004.

Prospectus make only the five-year forecasts available, requiring me to interpolate my own numbers out beyond that mark. Obviously, five-year forecasts are pretty vague; but in my opinion BP do a good job of tempering expectations for distant years where that's appropriate. I am using the Wins Above Replacement forecast, reduicing every player to a single number.

PECOTA forecasts:

Freddy Garcia

2004 2.8 wins * 55% = 1.5 wins
Total 1.5 wins

Ben Davis

2004 0.9 wins * 55% = 0.5 wins
2005 0.7 wins
Total 1.6 wins

Total value to Chicago - 3.1 wins
Chicago will also receive free agent compensation for Garcia if offered arbitration

Miguel Olivo

2004 1.0 wins * 55% = 0.6 wins
2005 0.7 wins
2006 1.0 wins
2007 1.0 wins
2008 0.7 wins
Total 4.0 wins

As we can see, Olivo alone is probably worth Garcia plus Davis. He's a young catcher with a good arm and a power bat, who runs well enough and strikes out little enough to keep his average north of his weight.

Jeremy Reed

2004 1.4 wins * 15% = 0.2 wins
2005 1.6 wins
2006 1.9 wins
2007 2.1 wins
2008 2.1 wins
2009 1.9 wins
2010 1.7 wins
Total 11.5 wins

Total value to Seattle - 15.5 wins plus Michael Morse (no PECOTA card; ranked 29th prospect in ?Southern League by OnDeck; 7th by pure statistical measures in 2004).

That's what we mean when we say it's a steal for the Mariners. Kenny Williams just set a match to $15 million of his bosses' money. All in a day's work for your typical major league GM, but not exactly the kind of thing that would endear you to me if I was writing the cheques.

What do people think of this sort of analysis? Is it helpful? I'd be more than happy to run this for other recent trades if there is any interest in this.

Posted by Craig B on Monday, June 28 2004 @ 12:39 PM EDT.

Dissecting The Deal - Freddy Garcia to the White Sox | 38 comments | Create New Account

The following comments are owned by whomever posted them. This site is not responsible for what they say.

_Nigel - Monday, June 28 2004 @ 12:49 PM EDT (#46493) #

I think this analysis is interesting. It's difficulty lies in the questionable value of PECOTA (or any other projection system) for minor leaguers (as intuitively it should be). It backs up your essential analysis which is that Kenny traded a whole lot of future cheap (which shouldn't be overlooked) value for some (and its really in doubt how much) current improvement in his rotation.

One minor note. Chicago gets the compensatory picks if Garcia gets offered arbitration rights by the Chisox and signs elsewhere. Its hard to see Kenny not offering arbitration after having given up the guts of his farm system. The picks do have some value in this whole mess for Chicago and should be taken into account somehow (although I'm not sure how).

Craig B - Monday, June 28 2004 @ 12:53 PM EDT (#46494) #

Nigel, I alerady noted that... except I made the mistake of saying Seattle will get the picks. I'll fix that.

_Nigel - Monday, June 28 2004 @ 12:55 PM EDT (#46495) #

Sorry, I was reading too quickly.

_Max Parkinson - Monday, June 28 2004 @ 01:00 PM EDT (#46496) #

Craig,

There has to be some amount of future discount, no? Is a win in sept. 04 exactly equal to one in 2009? Probably not to Kenny Williams, as there's no guarantee that he'll be there, especially if he continues to trade top prospects for decent starting pitchers...

Craig B - Monday, June 28 2004 @ 01:13 PM EDT (#46497) #

Max,

No, that's not right. The win in 2004 is worth the same as the win in 2009 - provided the marginal dollars per win are the same for both. Of course, that's a business analysis. From a purely sporting analysis, the wins are exactly the same (provided, I guess, that the marginal pennants per win are the same. I tend to think that for a team like the White Sox, already likely to win this year, that they will be).

Kenny Williams isn't making trades for himself. He's making them on behalf of the organization. Yes, he'll be fired if he doesn't produce a pennant this year; but no, that doesn't give him the right (from the White Sox perspective) to give away all their best assets to marginally increase his chance of hanging onto his job.

robertdudek - Monday, June 28 2004 @ 01:14 PM EDT (#46498) #

Further to that...increasing wins now leads to more revenue for this season and coming seasons (and increases the chance of a playoff appearance, generating still more revenue). That has to be accounted for on the plus side of the ledger for the ChiSox.

Craig B - Monday, June 28 2004 @ 01:30 PM EDT (#46499) #

increasing wins now leads to more revenue for this season

Actually, I think it only increases revenue for coming seasons, at least in any significant way. I'm pretty sure that the wins/attendance correlation is pretty weak once you tak out prior seasons' performance. Still, it's a factor.

_tangotiger - Monday, June 28 2004 @ 01:33 PM EDT (#46500) #

Craig, I do similar analysis, but I also bring in the cost factor.

For example, if I had a home worth 200,000$ and you had one worth 300,000$, would you trade them straight up? What if I paid off my mortgage, and you just started one?

When I looked at the ARod/Manny/Soriano deals, I took the expected value of the players (as you have) and converted them into dollar values (using 1.85 million$ per win).

Then, I compared that to how much present value dollars their contracts are worth. It is that difference that matters. It is for this reason that Manny Ramirez is worse than useless from a team perspective, because 100 million$ is better spent elsewhere, or just left in the bank. You may win less that way in the immediate future, but that money can be used when a bargain is available.

That said, Jeremy Reed!! My gosh...

Craig B - Monday, June 28 2004 @ 01:37 PM EDT (#46501) #

Right, the next step is analysis of what a player is actually making/likely to make. Finding out what the typical cost savings is on particular player (and types of players) in arbitration years is going to be a key to this.

_tangotiger - Monday, June 28 2004 @ 01:57 PM EDT (#46502) #

Completely agree Craig (in most cases). In this case, it's such a huge advantage to Seattle, it won't make a difference.

For years 1 through 6, just making an educated guess, but the top-line salaries would be 0.30, 0.80, 2, 5, 8, 11. Since we don't know that Reed will be that guy, maybe it's more like 0.3, 0.7, 1.5, 3, 5, 7.

So, he's going to cost them about about 20 million$ over those 6 years, with an "out" after every year. (It's an incredible system for a smart team to take advantage of. Taking the human factor out of the equation, it's the one place where smart management can make the most impact.)

Now, BP's got Reed at +11.5 wins over that time period. That's 21 million$ of contract value he'll generate. You know, that sounds pretty low. For a guy to generate an average of +2 wins above replacement, that means he's producing at the league average (the replacement level is about 2 wins below average). Reed's MGL's MLE's are off the charts. I mean if Reed is only +2 wins per year, is there any prospect that's even higher?

Having never looked at this issue, I'll still say that BP must be wrong. Though, it has piqued my interest in trying to get this right.

_tangotiger - Monday, June 28 2004 @ 01:59 PM EDT (#46503) #

For those interested:

http://www.tangotiger.net/MLE.html

Pistol - Monday, June 28 2004 @ 02:33 PM EDT (#46504) #

I liked the deal for Seattle, but I don't think the Sox are getting killed that badly for what they're trying to accomplish.

Simplistically, Davis and Olivo are a push. But Garcia will replace whoever the worst Sox starter is right now (Rauch?) and that's a pretty significant upgrade.

If you did a similar anaylsis last year for the Urbina trade to Florida for Adrian Gonzalez I'd suspect that the Rangers came out way ahead in the deal. But without Urbina (who was around for only a couple months with no compensation when he left) the Marlins likely don't win the World Series last year. As BP likes to say, 'Flags Fly Forever'.

Funny thing about the trade, I found out about it driving home last night on ESPN radio (and as an aside I recently got satellite radio and love it). It was pretty frustrating to hear them continually say 'Olivo and minor leaguers'. It wasn't until they actually interviewed Williams that I found out that Reed was included in the trade.

_Ryan Lind - Monday, June 28 2004 @ 03:06 PM EDT (#46505) #

This is rather insignificant, but doesn't Ben Davis' totals add up to 1.2 wins, not 1.6?

_Loveshack - Monday, June 28 2004 @ 03:23 PM EDT (#46506) #

No, that's not right. The win in 2004 is worth the same as the win in 2009

Im not so sure. From a certain perspective, wins are worth alot more when you're contending then when you're not because the playoff races can be so tight. For instance, I would gladly trade 5 wins this year for 5 wins (or even 4 or 3) next year if Im the Jays. Im not sure how this would figure into any stat calculations though.

Craig B - Monday, June 28 2004 @ 03:28 PM EDT (#46507) #

Ryan, you're right.

Loveshack, that's why I said provided, I guess, that the marginal pennants per win are the same. I tend to think that for a team like the White Sox, already likely to win this year, that they will be.

_Loveshack - Monday, June 28 2004 @ 03:31 PM EDT (#46508) #

Sorry 'bout that Craig. I'll be getting some new glasses on my way home from work tonight.

_Jim - Monday, June 28 2004 @ 03:44 PM EDT (#46509) #

Garcia already is at 33.2 VORP, I'm not sure how to convert exactly to WARP but one would think that his 2.8 WARP projection is out the window because he has probably already exceeded that - unless you expect him to pitch at replacement level for the remainder of the season.

_tangotiger - Monday, June 28 2004 @ 03:51 PM EDT (#46510) #

No, his expectation, today, is that he will produce at a right of 2.8 wins above replacement over a 162 game season.

That he has pitched better than expected probably pushes the 2.8 estimate to 3.1 or so, per 162 games.

Derek Jeter, regardless of how bad an April he had, had the same expectation of performance from May through Sept. It might have been a little less because of his bad April (the off-chance that it was something real, and not bad luck).

So, Craig's numbers hold.

_Jim - Monday, June 28 2004 @ 04:00 PM EDT (#46511) #

Your expectation might be 3.1 games. A realistic expectation might be more like 4. If he's already at 2.8 the only way to finish at 2.8 would be to pitch at replacement level for the next 100+ innings.

If this was a July 31st deal and he was at 4.2 WARP would you still tell me that it's realistic for him to finish at 2.8?

robertdudek - Monday, June 28 2004 @ 04:12 PM EDT (#46512) #

It will increase revenue this year greatly if it's the difference between the Sox making and not making the playoffs.

Craig B - Monday, June 28 2004 @ 04:38 PM EDT (#46513) #

Jim, it's not a question of where he'll finish, it's what he'll do from today.

Coach - Monday, June 28 2004 @ 06:52 PM EDT (#46514) #

Kenny Williams just set a match to $15 million of his bosses' money.

Craig, your analysis reinforces my gut feeling. It might be worth it if Garcia wins them the World Series, but that ain't gonna happen. Williams overpaid -- I presume Ozzie Guillen also pushed hard for this deal -- and Sox fans may regret it for years to come.

Amazing job by Bavasi to get so much in return; I wonder what his other offers were like. Imagine if J.P., in a misguided attempt to win this year, had traded Rios, Quiroz and Hill for three months of Garcia and too much of Davis -- I'd be furious.

_NDG - Monday, June 28 2004 @ 07:02 PM EDT (#46515) #

No, that's not right. The win in 2004 is worth the same as the win in 2009 - provided the marginal dollars per win are the same for both. Of course, that's a business analysis.

Businesses don't discount cash flows because they worry about the intrinsic value of the dollar (your analogy), they discount because of the uncertainty involved in projecting value 5 years into the future. It's the same case here, Garcia can be expected to contribute now, Reed MAY contribute in the future. As Max and Robert pointed the financial impact of winning the pennant is great, and hence wins above say 95 are worth more than wins 80 and 81.

BTW I also believe that Mariners did well in the trade, I just don't think it's the fleecing that many are making it out to be.

_Daryn - Monday, June 28 2004 @ 07:43 PM EDT (#46516) #

I agree, a win in the hand is worth two in 2009

I would also happily move 5 wins from a year we are losing to a year we are winning, so it seems to me that the value if a win increases depending on where you are in your "cycle"...

_Jim - Monday, June 28 2004 @ 08:29 PM EDT (#46517) #

Well how do you know what he'll do from today if you don't know how he'll finish?

If you had a projection system that said Jim Thome was going to hit 35 home runs and you traded for him today wouldn't you expect him to hit more then 10 home runs in the next 90 games?

If I've got a projection that says Freddie Garcia is worth 3 wins in 2004, and it's July 1 and he's already has more value then that then why would I measure a trade on a inaccurate projection?

I don't disagree that this is a bad trade, I don't think Garcia is much of a upgrade and I think they overpaid greatly based on what Kansas City got for Beltran. I do still believe that Garcia should be worth more then 1.5 wins the rest of the way based on how he's pitched for 3 months.

Craig B - Tuesday, June 29 2004 @ 08:43 AM EDT (#46518) #

NDG, your analogy doesn't hold. The projection system already discounts the performance projection for risk/uncertainty (at least, I believe it does - check out Mark Prior if you don't believe me), so we don't need to discount again.

Incidentally, the primary reason future cash flows are discounted is the opportunity cost of the capital

If you had a projection system that said Jim Thome was going to hit 35 home runs and you traded for him today wouldn't you expect him to hit more then 10 home runs in the next 90 games?

No, I'd expect him to hit 90/162 * 35 home runs in the remaining 90 games. Of course, I'd want to reflect that his recent good performance (which would adjust the projection upwards) but I don't have a projection system that can incorporate that data.

You haven't read a single word of what I've written, Jim, and it's frustrating. I'm not making this point again.

_Jim - Tuesday, June 29 2004 @ 09:11 AM EDT (#46519) #

I've read everything you've written.

Here is something you wrote:

'What do people think of this sort of analysis?'

I think this analysis would be better if it weren't based on old flawed forecasts.

_Geoff - Tuesday, June 29 2004 @ 09:31 AM EDT (#46520) #

Jim, you're not still not getting it - if Thome was projected to hit 35 homeruns this year, then his projection right now would be 25 + approximately 19 home runs (35 * 90 / 162) - so about 44 homeruns - do you understand how this works?

Craig B - Tuesday, June 29 2004 @ 09:35 AM EDT (#46521) #

It's probably helpful here to point out that how a player performed over his prior career is a much stronger predictor than how he's performed over his last 15 starts, which is why we wouldn't want to boost Garcia's projection too much based on this year's performance.

_NDG - Tuesday, June 29 2004 @ 10:13 AM EDT (#46522) #

Craig, while it's true the projection system discounts for player performance, it does not discount based on team performance, which is what the non-Jim counter-argument is. Even if Reed contributes 5 wins in 2009, if those wins are #77-81, they aren't as valuable as Garcia's win #96.

Future cash flows are discounted to determine opportunity cost, not because of it. All projects are discounted, not just some, so the selection between projects is based on what provides the best outcome versus opportunity. The reason for the discounting is uncertainty.

Mike Green - Tuesday, June 29 2004 @ 10:23 AM EDT (#46523) #

Craig,

Were the PECOTA projections for Jeremy Reed generated at the start of the year? I think so, but I wanted to make sure.

_Jim - Tuesday, June 29 2004 @ 11:10 AM EDT (#46524) #

Good Lord. I understand what you are saying.

Hopefully I can explain myself better this time. The case can be made that he is a better player then 2004 projection. 15 starts isn't a huge amount of starts, but the results could be showing a new 'true talent'.

There is some evidence that Garcia will be worth more like 5.0 wins this season - not the 2.8 that PECOTA projected. That would make 55% of his season worth 2.75 wins - not 1.5 wins if you use the 2.8 win projection.

I know that many of you disagree, but there is a CHANCE that Garcia is a better pitcher then that projection and that me might very well be worth more from here out then 1.5 wins. One MIGHT believe that based on the fact that he has so throughly outpitched his projection through today.

Craig B - Tuesday, June 29 2004 @ 01:24 PM EDT (#46525) #

I get it Jim. You're right, but I think the effect is smaller than you might imagine. It probably wouldn't serve to bump Garcia's expectations from 2.8 wins to 5.0 wins unless he'd done something Pedro-like. (Pedro's 5-year projections at the start of the year were 6.8, 4.2, 4.8, 4.8, 3.9; Mark Prior's were 6.1, 4.6, 3.7, 2.9, 3.7)

_Jim - Tuesday, June 29 2004 @ 01:36 PM EDT (#46526) #

There is one thing I don't love about PECOTA. It seems that the long term forcasts work out like this:

For Reed he's a league average player for the better part of a decade according to the system. It seems that 'average' comes from:

a 60% chance he is a good player (4 wins per year) and a 40% chance that he washes out and is worth 0 wins per year.

I understand tempering a prospect's future - but if PECOTA were right about everyone (I know that noone is claiming that it is) - there wouldn't be any MVP candidates in 2007. No player has a very good long term forecast.

_tangotiger - Tuesday, June 29 2004 @ 01:52 PM EDT (#46527) #

I'm not sure that we are all on the same page regarding the future performance. This is how it would work:

At Apr 1, our best guess is +2.8 wins over the next 162 games.
At July 1, our best guess is +2.8 wins over the next 162 games.

As of July 1, there is only 81 games left (or so). So, our best guess is +1.4 wins over the next 81 games.

Now, between Apr 1 and July 1, he already got +3.3 wins (over 81 games). So, it stands to reason that we may have been a little pessimistic about his true talent level.

So, also considering his Apr1-Jul1 performance,
At July 1, our best guess is +4.0 wins over the next 162 games.

As of July 1, our new best guess is +2 wins over the next 81 games.

So, that's what they are paying for... that he will produce at +2 wins over the next 81 games.

(Note, all numbers for illustration only.)

_Jim - Tuesday, June 29 2004 @ 01:59 PM EDT (#46528) #

Ok - that looks different then what you were saying before.

Mike Green - Tuesday, June 29 2004 @ 02:13 PM EDT (#46529) #

Assuming that the PECOTA forecasts have not been updated, I would venture a guess that Reed's PECOTA forecast would be modestly lower now than it would have been at the start of this year.

Craig B - Tuesday, August 03 2004 @ 01:52 PM EDT (#46530) #

Just an update on this deal; I'm busy doing a "Dissecting the Deal" for some of the deadline trades and thought I'd stop off here for a moment.

Garcia is 5-2 with a 3.67 ERA in a White Sox uniform; his strikeout totals have skyrocketed (49 in 49 innings) which is an excellent sign for the White Sox.

Davis has hit .250/.250/.464 in a backup role in Chicago.

Olivo has hit the ground running for the Mariners as their starting catcher, after he returned from a bout with kidney stones that started shortly after the trade. Since the trade, he's hitting .245/.302/.510 for the Mariners.

Reed has rediscovered some of his batting stroke in Tacoma, hitting .302/.354/.509.

Mike Morse is still hitting in San Antonio, albeit somewhat less than he had in Birmingham where he was on fire.

Dissecting The Deal - Freddy Garcia to the White Sox | 38 comments | Create New Account

The following comments are owned by whomever posted them. This site is not responsible for what they say.