Saber-Friendly Blogging 101: Trade Value Calculator
New to Beyond the Box Score and like what you see? Subscribe to our RSS feed or follow @BtB_Sky on Twitter.
In case you're not following along, Dave Cameron has a stellar series going on over at Fangraphs. He's running through the most valuable contracts in MLB, which you could also view as a raning of trade value, assuming all teams currently had the same financial and competitive situations. (Check out last year's top five.)
If we had enough data, a project like this could be automated. We'd need complete contract information and multi-year WAR projections just to start. Unfortunately, I don't have any of that lying around (if you do, let me know.) But it's still worth crunching the numbers player by player, especially if you disagree with any of Dave's rankings. It's really not that helpful to get into a war of opinions, and bringing some data to the shouting match is a great strategy.
In that vein, I've put together a little spreadsheet you can download to help estimate a player's excess value to his team, based on expected future production and salary. Pick a player, plug in the data, and post the results. Disagree with someone's assessment? Simply change the numbers and spit out your own analysis. The download link and tutorial after the jump.
Download the Excel file here. Some folks will want to jump right in (there are directions in the file), so go for it, but I've included a thorough tutorial below for everyone else. Here's what the calculator looks like:
The output is the information inside the box: year, actual/estimated salary, projected WAR, projected FA value based on WAR, and net savings. And we really only to care about the totals at the bottom, especially the net value number in the orange box. That's the monetary savings of a player to his team compared to paying for the same production on the free agent market. Aka, it's his context-neutral trade value.
Tutorial
To figure the actual/estimated salary (the column after year), you either need to enter a number in the Sal column (the second to last column, this one's for either free agent years or pre-arb years) or a percentage in the Arb column. We estimate arbitration players will earn 40% of free agent value in their first arb year, 60% in their second, and 80% in their third, so enter one of those three figures.
In the Halladay example above, I entered his actual salaries for 2009 (halved) and 2010 and pretended he'd be in the second year of arbitration in 2011, thus entering 60% in the Arb column and no value in the Sal column. Either way, a number pops up in the first Sal column, pulling the proper data via magic (IF and ISBLANK Excel functions).
The WAR column will need to be estimated by you. You could eye-ball a weighted average or crunch your own multi-year projections. The Val column is automatically calculated based on WAR, at $4.5M per win, plus $.4M base salary. The net value column is also automatically calculated and is the cost savings to a team of each player's contract. For years a player isn't under contract, make sure no cells in that column are filled in. Finally, if you're sure a player will walk after their contract is up, enter the value of the draft picks the team will receive as free agent compensation in the solitary green cell. As per Victor Wang, Type A's return about $5M while Type B's return about $2.5. Feel free to hedge your bets and pick numbers in the middle or lower than $2.5M.
Now take a look at the totals. You'll see the total cost of the contract, total expectel production (WAR) provided by the player, total expected free agent value of the player, and the total net value of the player's contract (that's the number in the little box). Zero means a player has zero trade value (team context aside) because you could supply the same amount of production by signing a free agent at the going free agent rate without giving up anything else. Positive numbers are good. Negative numbers mean a team should have to include something of positive value in order to get another team to take the player.
Prospects
You could run prospects through this chart, but there's an easier approach. Use Victor Wang's research, which Erik Manning summarized nicely in this table. Based on whether the prospect is a pitcher or hitter and where they fall in top 100 prospects lists or John Sickels' ratings, we can estimate the total net value they'll provide to their team. That number can range from about $36M for a stud hitting prospect to $7M for a Grade B pitcher to peanuts for Grade C prospects.
Where to Find Data
- Fangraphs and BaseballProjection player pages have historical WAR you can use to estimate future WAR.
- Cot's contracts has phenomenal salary information.
- Favor: give those sites credit when you post conclusions using their input.
One last tip. If you're going to analyze multiple players and don't want to lose your work for each one (maybe because you want to come back and change some things later), just copy the worksheet each time and re-name the tabs to keep track of the various players.
Some Analysis Using This Tool
- What's a Fair Return For Roy Halladay? (BtB)
- How About Packaging Vernon Wells With Roy Halladay? (BtB)
- Braves Should Trade For Willie Harris (BtB)
- Economics of the Astros (Crawfish Boxes)
- Blue Jays Biggest Trade Values (Bluebird Banter)
As always, please share any links to work you've done related to this article in the comments. We all want to see what others come up with in terms of trade value.
3 recs |
17 comments
|
Comments
Man do I love Excel
Very cool, Sky
by Tommy Bennett on Jul 15, 2009 12:39 PM EDT reply actions 0 recs
This is awesome.
godfather of futureredbirds.net
by erik on Jul 15, 2009 12:40 PM EDT reply actions 0 recs
I've been playing with this thing for way too long.
Thanks. This is awesome.
><
by Blicks on Jul 15, 2009 2:06 PM EDT reply actions 0 recs
thanks sky
For making hacks, I mean “analysts” like myself obsolete
I'm not a sabermetrician, but I do play one at Driveline Mechanics.
Can't get enough of me? Check out my Twitter feed.
by devil_fingers on Jul 15, 2009 3:58 PM EDT via mobile reply actions 0 recs
I am going to be a little greedy, but can you set it up so the $/per WAR can be enter (if it changes) instead of having it at 4.5/per WAR?
Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.
by Jeff Zimmerman (TucsonRoyal) on Jul 15, 2009 4:02 PM EDT reply actions 0 recs
I can do that.
Beyond the Boxscore Not a member? Sign up.
by Sky Kalkman on Jul 15, 2009 4:17 PM EDT up reply actions 0 recs
Very good
Im favoriting it
I Have Spoken.
by The_Fan on Jul 15, 2009 4:03 PM EDT reply actions 0 recs
Sky, you are such a jerk
here i am rubbing two sticks together and youve had a Zippo this whole time!
this is really outstanding. thanks.
by Charlie Scrabbles on Jul 15, 2009 4:20 PM EDT reply actions 0 recs
Agreed
The Crawfishboxes
A good friend of mine used to say, "This is a very simple game. You throw the ball, you catch the ball, you hit the ball. Sometimes you win, sometimes you lose, sometimes it rains." Think about that for a while.
by DyingQuail on Jul 15, 2009 6:40 PM EDT up reply actions 0 recs
Outstanding
Thank you for taking the time to put this together.
by WiHaloFan on Jul 15, 2009 8:21 PM EDT reply actions 0 recs
how well does it hold up for low WAR players?
I’m interested in the draft and prospects so I was curious to see how the spreadsheet handled different quality players over thier first 6 years.
So I set the SAL for the 1st 3 years at 0.5M and then used the standard 40,60,80% progression for years 4-6.
A 6 WAR/year player – an instant star – will return a surplus value of 118.4M including 5M for type A consideration.
I’ve screwed around with this type of analysis before using BP’s MORP and that looks fine to me.
If I drop it to 1 WAR/yr and eliminate draft pick compensation I get 18.9M for a player halfway between replacement level and average. Looking at Victor’s data from the AL West chart that player is more valuable than every prospect group except hitters from the Top 50. That just doesn’t seem right to me.
Drop it even lower to 0.5 WAR/yr and you get a 9.4M value. That’s noticably better than grade B pitchers and hitters. Victor’s research is in WSAB so it’s not quite apples to apples, but I don’t think he was suggesting 0.5 WAR/year was a likely outcome for grade B prospects.
Actually, here’s a good real world comparison. Again based on Victor’s work we have two draft picks for type A free agents worth 5M. In order to get down to a 5M value I have to use 0.26 WAR/yr which is just extremely low.
That is saying that Beane can either hold onto Holliday for 2 picks worth 5M or trade Holliday for a prospect who will produce 0.26 WAR/yr worth 4.9M and those two outcomes are basically interchangeable. That just can’t be true.
I know Tango has made pretty convincing case that WAR/$ is linear and not exponential based on emprical free agent data, but at these really low levels aren’t we seeing that break down?
Since players at this lew level of yearly productivity don’t get offered arbitration or reap big FA dollars could it be that the relationships that hold for FA and arb (mostly the 40/60/80 progression) just don’t apply to very low WAR players?
And if so, where would you draw that line and how else would you come up with a valuation for a low WAR player (say a guy who porduces 1-4 WAR over his 6 pre-FA years)?
Curious to hear your thoughts.
Oh, and thanks a bunch for posting it. I’ve been playing around with it non-stop since. :)
by philly on Jul 16, 2009 11:30 AM EDT reply actions 0 recs
Part of where things break down is scarcity.
There are about 15 full-time players on a team. You can’t have 50 .5 WAR players and be a 25 WAR team, for example. Those guys are still deals, perhaps, but they each take up one roster spot. For prospects, a B level prospect is worth $7M because of how often he tanks. If he actually makes it, he’s worth much much more. So three B-level prospects might rate as useful as a .5 WAR player, but four of them are likely to only take up one roster spot at the MLB level, leaving room open to find deals at other positions.
Also, a 1 WAR player for six seasons seems unrealistic. If a guy is a 1 WAR player for his first couple years, chances are he has a career year that’s better or is out of baseball after a couple years.
And, I believe we probably underrate the value of knowing you’ll have a 1 WAR player for six years. We tend to assume a B prospect will do that, but you get a ton of B prospects that turn out worthless. At least you’re getting 1 WAR per season out of the MLB scrub instead of nothing, every time but the odd time the B prospect works out.
I obviously didn’t solve the dilemma you brought up, but hopefully I provide some other perspective.
Beyond the Boxscore Not a member? Sign up.
by Sky Kalkman on Jul 16, 2009 12:36 PM EDT up reply actions 0 recs
Thinking about it more, I like this idea:
could it be that the relationships that hold for FA and arb (mostly the 40/60/80 progression) just don’t apply to very low WAR players?
4/0/60/80 is clearly a rough model, probably the roughest in the overall model. Good-fielding, high-OBP, low-AVG shortstops will earn much less in arbitration than high-AVG, high-RBI corner outfielders. (Same for the free agent market.) And it’s could certainly be the case that arbitration awards aren’t as linear as free agent signings. Would a .5 WAR player earn $2.5M their second year in arbitration? Probably not.
Beyond the Boxscore Not a member? Sign up.
by Sky Kalkman on Jul 16, 2009 12:52 PM EDT up reply actions 0 recs
The spreadsheet is very cool.
Here’s a fanpost I put up on RR about some the guys the Royals could look at trading.
He can get 4, NOT 5.
by Warden11 on Jul 23, 2009 11:20 PM EDT reply actions 0 recs

by 










BtB on Facebook
















