Navigation: Jump to content areas:


Pro Quality. Fan Perspective.
Login-facebook
Around SBN: Knicks Beat Lakers With Familiar Strategy

TVP (Time Value of Players)

Since the Holliday trade there's been a lot of talk about the actual value of cheap prospects vs expensive MLB talent. So far, I've read a lot of good posts that give good qualitative points on the issue, but nothing that really boils the issue down to numbers that can be manipulated and compared directly. So, i decided to take a stab at creating a useful analysis tool for this situation.

Star-divide

I recently took an engineering economic analysis class and it partly focused on the concept of Time Value of Money (TVM). Basically, what this boils down to is that money gained or lost in the present is worth more (or has more power) than that gained or lost in the future. This is based upon an annual Minimum Acceptable Rate of Return (MARR), which is usually based on what you could get for putting the money in low risk investments.

While I'm fairly new to advanced baseball stats - Vorp,WAR and such - I thought that this concept could be modified to express the Time Value of a Player, which would help put the whole Prospects vs MLB Talent debate into a clearer perspective. I've created a spreadsheet and It basically breaks down like this:

  • For each of years 0 through 5, (0 is considered the current season, which is the time at which the present worth comparisons are "scaled" to and compared) Salary, WAR and the seasons $/Win (found by estimating a $400,000 increase/yr from the values found at fangraphs) are entered in.
  • The Future Worth (called this because it is in the future from the time when the values are to be compared) is calculated as WAR*($/Win)-Salary. (This value isn't shown on the spreadsheet).
  • The Present Worth is then calculated with a simple formula based on the year and the MARR.
  • All the Present Worths are added together to give a total which represents the current value of the player over the entire analysis period.

Now, the confusing part in all this is how you define the MARR. Minimum Acceptable Rate of Return doesn't make a whole lot of sense for this analysis. The way I see it, MARR (from now on called SVal) in this situation is how much you value the current season over the next season. The way you would calculate this is

MARR = SVal = 1-(Value_of_Next_Season/Value_of_Current_Season)

So if you value winning this season twice as much as winning next season you would have 1-(1/2)=50%

Teams currently in contention over the last few years would have a large SVal, while teams in rebuilding years would have a small or possibly negative SVal (meaning gains and losses in the future actually have MORE weight than present gains and losses).

 

Anyway, the whole thing is a fairly crude analysis tool, depending on how much weight you put into WAR and how accurate you feel your projections are.

 

Problems:

- Salary might need to be an adjusted value compared to a replacement player (figure this out on your own and plug it in and it will work just the same)

- Because a single SVal is given to represent a large period of time, it can only represent a teams general philosophy for that whole period. The spreadsheet could be reworked to allow for a changing mindset over that time (a team in contention one year might be rebuilding the next) but would require a lot more work than I wanted to do at the time)

 

Anyway, just wanted to get your thoughts on whether or not you think this would be useful.

Also, the values that are currently in the sheet are complete guesses. I know next to nothing about projecting WAR and how minor league (and major league) salarys work. Just fill in what you want.

Comment 15 comments  |  5 recs  | 

Do you like this story?

Comments

Display:

Export the doc to excel or other spreadsheet program to mess around with it

I’m having trouble figuring out how to let people mess with it, without it permanently changing the document.

by Frozen_Beltran on Jul 28, 2009 4:27 AM EDT reply actions  

god

i hate sabr stats, feels like im back in trig…and i feel like this arguement is beating a dead horse….enjoy the trades or dont. we have no control over it. that and so far the trades are raking it in at a major league level and not in the minors

Pujols takes out "I" in BIG and "A" in MAC, previously considered to be an unyielding, consonant threat

by DESTROYER on Jul 28, 2009 8:29 AM EDT reply actions  

if you hate stats so much

why bother commenting to a stats post?

"But listen, and understand: more Molinas are out there. They can't be bargained with. They can't be reasoned with. They don't feel pity, or remorse, or fear." - THT

by Yadi2Second on Jul 28, 2009 8:36 AM EDT up reply actions  

You obviously must be trolling,

or venting, or baiting, or something, because your logic leaves much to be desired.

"I'll be glad to have Ryan [Braun] help if he wants to. I'll give him a badge and he can be my deputy." - Doug Melvin

by all4tookie on Jul 28, 2009 8:57 AM EDT up reply actions  

This is a great start..

With much further applications that you may or may not realize. I was thinking about just this same thing about Holliday vs. Wallace on my commute last night. For instance:

1. You could apply WAR projections on an individual basis, factor in salary, and run an analysis on 40-man similar to a discounted cash flow model to determine and compare total franchise values (and determine whether a trade was accretive by running the analysis before and after)

2. From 1, you could give a clearer picture of ROI from certain players. Usually when we see a list of the best signings it is something along the lines of “Player W has contributed X WAR over Y seasons for $Z million” – this isn’t technically accurate because the timing of the WAR contributions do matter.

3. I just got excited as I realized this while typing 2, but there could be a pretty straightforward CAPM-type valuation here. If we could somehow determine a risk-free rate of return (I think this would be replacement level contributions/cost) throughout baseball, we could determine players’ betas based on ZiPs projections and come up with an expected return required for a given price (I assume this projection would be a good enough proxy for beta, might need to adjust for injury risk). And the benchmark performance is simple enough – just determine what WAR costs on the FA market. So, for instance, when signing a FA to a long-term deal it would be much, much more accurate to view his contributions based on the timing of said contributions, his beta (likelihood of under/over performing baseline rate), and time-valued cost rather than simply a total WAR/$ average.

Obviously the challenge is going to be determining the required rate of return for different teams, but it is probably decently correlated with average payroll, attendance, etc (as long as causation issues are avoided). At the very least you can determine the ballpark rate for a few teams just by looking at recent transactions and their IRRs.

There are countless other applications that you could extend your basic TVP principle to, and it is so obvious I find it hard to believe this work isn’t out there. Front offices would be crazy not to run these types of models.

Great work for a fanpost – this should be something that stirs up some interesting thoughts and discussion.

"I'll be glad to have Ryan [Braun] help if he wants to. I'll give him a badge and he can be my deputy." - Doug Melvin

by all4tookie on Jul 28, 2009 8:55 AM EDT reply actions  

I'm pretty certain

Billy Beane and the Athletics do just this. They probably have to.

by Michael_68_1999 on Jul 28, 2009 12:29 PM EDT up reply actions  

Hmmm...
(I assume this projection would be a good enough proxy for beta, might need to adjust for injury risk)

Wouldn’t you have to adjust for injury risk? Especially if you are using replacement level contribution and cost to determine a risk free rate of return?

"I just wish that the late Harry Caray were still around so I could hear him mispronounce 'Kosuke Fukudome' every fukun' night" -- Dennis Miller

by fourstick on Jul 28, 2009 7:13 PM EDT up reply actions  

Injury risk definitely needs to be accounted for,

And it depends on if you think the projection system accurately represents this risk. ZiPS and PECOTA only reflect injury risk insofar as it bases playing time assumptions on past time missed. If there is reason to believe a player will miss more time in the future than he has in the past due to change in use or out of position or something, that would need to be adjusted.

I’m thinking you could make base case, downside case, and upside case to get a pretty realistic idea of value.

"I'll be glad to have Ryan [Braun] help if he wants to. I'll give him a badge and he can be my deputy." - Doug Melvin

by all4tookie on Jul 28, 2009 7:22 PM EDT up reply actions  

Does this have legs?

I’m pretty interested in this obviously. I think its a great idea with some insightful applications.

"I'll be glad to have Ryan [Braun] help if he wants to. I'll give him a badge and he can be my deputy." - Doug Melvin

by all4tookie on Jul 31, 2009 8:24 AM EDT up reply actions  

Yes

I would love to see this experimented with further and refined. It might not shake out in the end, as there are a lot of issues to account for, but it would be lightyears ahead of the “__% of 1st round draftees contribute at the major league level” type musings we’re seeing currently. I don’t have the free-time (or math, or intellect) for something like this… but all the best to you if you’re willing.

-- Aidan Sonoda
R.I.P. Nick Adenhart - 4/9/09
In necessariis unitas, in dubiis libertas, in omnibus caritas.

by Aidan Sonoda on Jul 31, 2009 11:39 AM EDT up reply actions  

Excellent idea for further statistical analysis....

Furthermore, if you actually could come up with a model for this that you were happy with, you could even expand this idea out to include the worth of compensation draft picks, by figuring out what percentage of prospects pan out.

Of course, that’s a HUGE next step, but if we had a model that could do that, we could really make a much greater case for why the Holliday/Wallace trade was either an acceptable risk or foolishly squandering away prospects.

by mtalken on Jul 28, 2009 1:09 PM EDT reply actions  

I think it was Victor Wang over at Hardball Times...

who pinned the value of compensation picks at ~$5MM for Type A and ~$2.6MM for Type B

by mopack on Jul 28, 2009 3:53 PM EDT up reply actions  

Victor Wang did a similar analsyis at THT

He finds that the average prospect has a certain surplus value depending on their rank. He also looked at draft picks. Sky at BtB has a nice summation of the two:

http://www.beyondtheboxscore.com/2009/7/20/950254/which-is-better-compensation

F%#& Billy Beane. Actually... I kinda like Holliday

by vivaelpujols on Jul 29, 2009 12:35 AM EDT up reply actions  

good stuff

I looked into his paper on prospect value and he was using a little tvm stuff. However, he doesn’t real go into depth on how the value was derived other than “someone else used it”.

by Frozen_Beltran on Jul 28, 2009 9:14 PM EDT up reply actions  

Comments For This Post Are Closed


User Tools

The Internet's #1 St. Louis Cardinals blog.

FanPosts

Community blog posts and discussion.

Recommended FanPosts

N1046613005_8392_small
Our 2010-2011 strays
649494__1__small
Hall of WAR: Part 2

Recent FanPosts

Hahaha_small
These were a few of my favorite things (fink reminisces about the 2011 regular season)
Dsc01844_small
Cardinals take the Governor's Joplin Challenge, will help build 35 homes for torando victims
St-louis-cardinals-script_small
Best Cardinals of All-Time - Relief Pitching Edition
St-louis-cardinals-script_small
Best Cardinals of All-Time - Starting Pitching Edition
Small
Two Trades That Set the Cards Back in the 70s
Nyc_small
Cardinals Offense vs. Reds Offense - 2012
Nyc_small
Cardinals Rotation vs. Reds Rotation - 2012
St-louis-cardinals-script_small
Best Cardinals by Position - Center Fielders

+ New FanPost All FanPosts >


Managers

Jack_benny__1__small DanUpBaby

Editors

Bendermad_small azruavatar

Trigun_001_small the red baron

Images_small tom s.

Authors

1989_bgh_cropped_small bgh

Valverde_medium_small vivaelpujols