Assessing proper value of a FA offer

[QuikSand]

Okay -- if you've played much FOF at any level, you have undoubtedly come across this issue. If you've played multiplayer FOF 2004, even more so.

It's just darned tough to program a simple formula that properly weighs in the variables involved with a free agent contract offer.

FOF does a respectable, but certainly not infallible job of this. Maybe some of the good minds here (though everyone is welcome, snicker snicker) can come up with something that works better.


I think an ideal formula for the appeal of a given offer should reflect:

-the amount of "guaranteed" money involved
-the amount of total money involved
-the amount of money involved before any rapid escalation of salary
-the rapidity of the player seeing free agency again
-the degree to which the contract is over or under market value
-the age of the player and his likely prospects after more time
...and probably more things that I'm not considering at the moment.


I don't even have a first stab at this in mind yet... so I'm at square one as well, I confess.

[cthomer5000]

This is probably a bit too techincal for me (in terms of coming up with an actual formula), but I think a few more things need to be involved:

- a player's loyalty to his current squad
- the proximity of teams to his hometown
- the proximity of teams to his college (this is something I'd like to see introduced in FOF, or I'd at least like to see the game-generated players stay closer to home more often)

[gottimd]

Are you talking about creating a Third Party Mod for the game? I would be happy to help out with that.

[gottimd]

His current and Potential value in comparison to current roster and other available Free Agents

Also, Frequency of injury. Although I don't know how to obtain this, unless in a newer version of FOF (if ever) has this. That way you can determine risk on money. You don't want to offer top dollar to a great receiver, if he is injury prone.

[gstelmack]

I'd think that calculating a "present value" of the dollars is a good way to start out on the financial side of it. Money in later years of the contract should not count nearly as much as money in the bonus and first year or two. That would help with backloading issues.

There should also be something in here about the player's perceived value, which I guess is where your market value factor comes in. What I like about this is if teams stop paying exhorbitant salaries, this number might actually come down, reflecting the true market value of the players.

[gottimd]

There should also be something in here about the player's perceived value, which I guess is where your market value factor comes in. What I like about this is if teams stop paying exhorbitant salaries, this number might actually come down, reflecting the true market value of the players.

Do you mean adding the factor in of available Cap Space for each team, and another variable of each teams need?

[QuikSand]

I don't know if this would necessarily become a game modification, but I guess that's possible. In a MP league, the commissioner could set something up to evaluate the various offers submitted outside the game, and then only enter the one that tops the list according to the formula.

I agree with the stuff about adding loyalty and location issues -- but I also see those as pretty secondary. I think our biggest problems lie with the game's actual recognition of numbers, and the inabilityt o propery distinguish "real money" from "empty money."

[cthomer5000]

A complete shot in the dark at evaluating money in a weight/ratio form:

bonus money: 24
1st year money: 18
2nd year: 12
3rd year: 9
4th year: 7
5th year: 5
6th year: 4
7th year: 3


any thoughts? My thought here is that the money becomes less and less valuable in future years, but once you get into year 4 and beyond... the dropoffs aren't nearly as harsh as they are from years 1 to 2 , 2 to 3, 3 to 4. Bonus money is obviously huge. I don't think it's unreasonable to think that 1 million in bonus might be worth as much as 8 million in salary 7 years down the road. Or am I over-valuing bonus money a little (or a lot)?

[albionmoonlight]

This may be hard from a programming point of view, but I would love it if you could also factor in the team's situation. I know in IHOF, some players were complaining because they had a lot of cap room in the current year, and players were not accepting contracts with high first year salaries, preferring instead to push more non-guaranteed $$ into the future (this may have had more to do with renegotiations than FA, but to me the two are related).

It would be nice if a player could see that you had lots of current cap room--or lots of cap room in the next two years--and understand that you are trying to give him the $$ that he wants--just in a way that helps your team.

I also think that players should appreciate current year salary much more. It is still cash in hand, much like a bonus. The only down side to first year salary instead of bonus is that it makes it a bit easier for the team to cut you in the future, but I think that FOF does not appreciate it enough.

Finally, I wish that players understood better the typical NFL contract:

BIG BONUS
Year 1: Modest salary
Year 2: Modest salary
Year 3: Modest salary
Years 4-7: The GNP of China

In the real world, everyone knows that this is a three year deal that will be renegotiated in year 4. In FOF, you have a dual problem. First, players won't accept such deals easily. Second, they put too much stock into late season salary that will never be recognized (meaning that they put any stock into it whatsoever).

I know that this does not help us get much closer to a formula, but it's my 2c.

[QuikSand]

An example, to make some of my point a little clearer:


Player is a 5th year free agent, fairly promising -- let's say he's a 50/60 MLB. The minimum salary for his age is currentl $600,000. A solid starting player at MLB currently makes about $1,500,000 and the top salary at the MLB position is $5,000,000.

If he gets these six offers, how should he feel about them?

Offer Sheet A
4yrs, $10 million
Bonus $2,000,000
Sal1: 1,000,000
Sal2: 1,000,000
Sal3: 2,000,000
Sal4: 4,000,000

Offer Sheet B
4yrs, $10 million
Bonus $4,000,000
Sal1: 1,000,000
Sal2: 1,000,000
Sal3: 2,000,000
Sal4: 2,000,000

Offer Sheet C
4yrs, $10 million
Bonus $1,000,000
Sal1: 1,000,000
Sal2: 1,000,000
Sal3: 1,000,000
Sal4: 6,000,000

Offer Sheet D
2yrs, $6 million
Bonus $2,000,000
Sal1: 2,000,000
Sal2: 2,000,000

Offer Sheet E
2yrs, $6 million
Bonus $2,000,000
Sal1: 1,000,000
Sal2: 3,000,000

Offer Sheet F
2yrs, $6 million
No bonus
Sal1: 3,000,000
Sal2: 3,000,000

[gottimd]

I think the players agent value should be weighed into this as well if you are trying to determine a "fair deal" with the Free Agent. Obviously the player with an agent who is known as a stubborn negotiater will demand more money, and you maybe able to find a player with similar characteristics who is willing to sign for less.

[Buzzbee]

This may be a little bit of a sidebar, but probably falls into the "other factors" section. It may also help give some shape to a formula for FA offers. I was doing some testing for IHOF regarding free agent signings to see if we felt Free Agency needed to be handled out of game or if the game did a sufficient job of handling competing offers. My testing basically involved setting up all 32 teams and making the same offer to the same free agent to see what team the FA would sign with. From memory, here are some of the things that I found. (The threads are buried somewhere on the IHOF forums if you want the details.)

- When all offers are the same, the FA will select the same team repeatedly, without variation. (I'll refer to this as the "chosen team")

- If the "chosen team" offers less money, the player will still sign with that team, up to a point.

- The point (salary) at which a player chooses a different team does not vary. In other words, if the Chosen Team offers $10,000 below that point, the player will sign with Team B every time. If Chosen Team offers salary at that point, player signs with Chosen Team every time. NO RANDOMNESS.

- I was not able to distinguish any strict correlation to "loyalty" or "play for winner" in a cursory review of which player signed with which team. However, I didn't really set up the test to track this, nor did I perform any in depth study. It was mostly just gut feel that high loyalty didn't result in the former team being the chosen team.

- I didn't do any research on the mix of the offers (bonus vs. salary). All offers were one year contracts, IIRC.

- While there was no randomness in what a player would accept, there was some randomness in when the player signed. Same offer simmed 10 different times would result in player signing in different stages of FA. Larger/better offers "seemed" to induce players to sign quicker, but I only tracked when they signed to determine IF they would sign at different times, not if bigger contracts led to earlier signings. Therefore after just a few trials, I quit tracking.

With that, I'm guessing that there is some sort of weighted formula to "grade" each offer. The highest grade gets the player.

Factors include those things you mentioned above, and I'm guessing factors such as loyalty and play for winner. Past injustices may also play a role.

Hope this info is helpful.

[cthomer5000]

my take on the contract offers:

B > D > E> F> A > C

[cthomer5000]

and my reasoning on the evaluation of offers:


B - largest signing bonus of any deal, and effectively 5 million guaranteed (bonus+1st year salary).

D - As big a bonus as any offer except B, and of a short length. 4 million guaranteed with a very good chance of earning the year 2 salary.

E - again, second biggest bonus. 3 million guaranteed, and a good 2nd year salary - although the chance of being cut is greater than in the 'D' package.

F - no bonus, but effectively 3 million guaranteed. The team risks nothing in cutting the player, so there is some serious risk beyond year 1.

A - 2 million dollar bonus, 3 million 'guaranteed.' Not a terrible deal, but a player could basically get the same in offers E or F without being tied up for so long. I think we would universally agree that players would rather earn 4 million in 2 years and be free versus making 4 million in 2 years and still being under contract for 2 more years.

C - 1 million bonus, only 2 million guaranteed. Also, there is no way year 4 is seen without re-negotiation.

[cthomer5000]

seeing if my system (above) is worth a damn, here's how it sees the values of the offers (numbers in millions)


B 158
A 124
D 108
C 105
E 102
F 90

B,A,D,C,E,F
my human ranking:
B,D,E,F,A,C

my system needs to be re-tooled. I think it's overvaluaing later years (since a longer deal has an inherent advantage over a shorter one). How about if I divide the point total by the number of years?

B 158 /4 = 39.5
A 124 /4 = 31
D 108 /2 = 54
C 105 /4 = 26.25
E 102 /2 = 51
F 90 /2 = 45

old ranking
B,A,D,C,E,F

new ranking
D,E,F,B,A,C

my human ranking
B,D,E,F,A,C

getting closer I guess.
<!-- / message --><!-- sig -->

[Buzzbee]

F - no bonus, but effectively 3 million guaranteed. The team risks nothing in cutting the player, so there is some serious risk beyond year 1.

Not totally true, is it? Salaries are prorated, so cut the player early in the year and he doesn't get a sniff of that $3 million. Correct?

[cthomer5000]

F - no bonus, but effectively 3 million guaranteed. The team risks nothing in cutting the player, so there is some serious risk beyond year 1.

Not totally true, is it? Salaries are prorated, so cut the player early in the year and he doesn't get a sniff of that $3 million. Correct?
what you're saying is true. But from the sound of the player (as described by Quik), there would effectively be no chance of the player being cut in year 1. That was my basis for making that sort of assesment.

[Buzzbee]

Valid point.

[Chubby]

Finally, I wish that players understood better the typical NFL contract:

BIG BONUS
Year 1: Modest salary
Year 2: Modest salary
Year 3: Modest salary
Years 4-7: The GNP of China

In the real world, everyone knows that this is a three year deal that will be renegotiated in year 4. In FOF, you have a dual problem. First, players won't accept such deals easily. Second, they put too much stock into late season salary that will never be recognized (meaning that they put any stock into it whatsoever).

I know that this does not help us get much closer to a formula, but it's my 2c.
I totally agree with this and I too wish more people would realize it. This is one of the main reasons why things are coded like they are and I don't see it changing much.

Players don't accept such deals easily but they are possible with practice. I do agree too with cthomer in that it'd be nice for players to recognize when a team has free cap space NOW and wants to move some of that money in later years up.

[QuikSand]

cthomer, I see your starting point (and the logic behind it) but I think there's something inherently flawed about any kind of blind weighting that isn't context-sensitive.

For instance -- if an offer looks like this:

4yrs, $10 million
No Bonus
Sal1: 1,000,000
Sal2: 1,000,000
Sal3: 2,000,000
Sal4: 6,000,000

Then I don't think you want to put much of any weight on that last figure at all -- it's a total fantasy. Whether it's $6m or $26m, it's not going to actually happen, so ideally, it basicaly shoudln't be weighted at all. But that's tough -- it's difficult to find a formula or routine that properly identifies these "cliff effects" and correctly drops salary values that occur after them (even more so when there is little or no bonus involved).

[Pyser]

ive had a bit of an opposite experience in the eNFL.

many times, players are asking for 5 yr, $40-mil. and i have seen at least 4 times, a player with those demands signing 2 yr, $6-mil deals. over, in fact, 5 yr, $30-mil dollar deals.

now, i dont know the breakdown of the $30-mil offer, but the 2 year deal the players sign always have a flat salary. so:

2 yr, 6 mil deal
$3-mil bonus
$1.5-m first year
$1.5-m second year

im completely baffled.

also, last year, a player turned down a 5 yr, $25-m contract with $7-m signing bonus, for a 4 yr, $16m deal, no bonus, flat $4m per year. that should NEVER happen, i think.

[randal7]

I have a slightly different take on the salary proposals: D-E-F-B-A-C. Reason being: my perception of my own future value. With a player rating of 50/60 (especially an MLB, who is probably in the top 10-15 guys for potential with this rating), the shorter contacts give me the maximum cash over the two years duration and allow me to be a free agent as my abilities peak. I will gamble the extra million that first year against the mega-contract 2 years down the road. Now if I am a 20/30 FB, I take B.

[MIJB#19]

Something I'd like to add to the discussion:

When players sign long term deals with big bonus money involved, doesn't that increase the chance they will be longer on the roster? Who's going to cut a guy with a 15M bonus on a 5yr deal in the second year. The dead cap space of 9M in the third year would be too much to cut him and the only option will be to renegotiate. With renegotiating, there's usually the cap out offer, where the year 2 salary turns into bonus money (guaranteed money) as well. And the same will happen in year 3. So in fact, of the 5 yr deal, the "sure thing" are not just the bonus money, but also the salaries in year 1 to 3 (or maybe better is x-2, where x is the total length of contract.) Of course, this only applies if the bonus money exceeds a certain percentage of the cap figure.

[Buzzbee]

But that's tough -- it's difficult to find a formula or routine that properly identifies these "cliff effects" and correctly drops salary values that occur after them (even more so when there is little or no bonus involved).

Perhaps the inverse of the ratio of salary/prior year salary, or some modification thereof. In your last example it would be:

year 1 to year 2:
1 / (1,000,000/1,000,000) = 1

year 2 to year 3:
1 / (2,000,000/1,000,000) = 1/2

year 3 to year 4:
1 / (6,000,000/2,000,000) = 1/3

Perhaps maybe even sqaring the result to really reduce the weight if there is a big "cliff".

Just tossing it out as an idea.

[QuikSand]

I have a slightly different take on the salary proposals: D-E-F-B-A-C. Reason being: my perception of my own future value. With a player rating of 50/60 (especially an MLB, who is probably in the top 10-15 guys for potential with this rating), the shorter contacts give me the maximum cash over the two years duration and allow me to be a free agent as my abilities peak. I will gamble the extra million that first year against the mega-contract 2 years down the road. Now if I am a 20/30 FB, I take B.

I think this gets to an important underlying point in these considerations -- the value of a contract is deeply case-specific. A promising young player who is likely to keep getting better should value a shorter contract, while an aging veteran or a mediocre player should probably value more guaranteed money and perhaps a longer-term deal.

[henry296]

I think this gets to an important underlying point in these considerations -- the value of a contract is deeply case-specific. A promising young player who is likely to keep getting better should value a shorter contract, while an aging veteran or a mediocre player should probably value more guaranteed money and perhaps a longer-term deal.

It seems quite the opposite in Real life though. The younger players are getting the super long term contracts like McNabb for 10 years. I think Portis' contract was long as well.

Todd

[dixieflatline]

Ok I'll take a shot with a formula that seems to produce some reasonable numbers to me. Hopefully, this will be a decent place to start and we can modify from there. I'm going to use Quiksand's example to run some numbers for what the formula gives. This formula will try to determine a rough estimate of the chance that the player will make the team in each of the years of the contract to calculate a expected money earned. Beware, I'm going to try to explain all the parts of the formula that I am going to use and it will get math intensive. Feel free to skip to the bottom and just look at the results.

Player is a 5th year free agent, fairly promising -- let's say he's a 50/60 MLB. The minimum salary for his age is currently $600,000. A solid starting player at MLB currently makes about $1,500,000 and the top salary at the MLB position is $5,000,000.

Offer Sheet A
4yrs, $10 million
Bonus $2,000,000
Sal1: 1,000,000
Sal2: 1,000,000
Sal3: 2,000,000
Sal4: 4,000,000


The only small adjustment I'm going to make to make the numbers a bit easier is lower the minimum salary to $500,000 or half a million. I'm also going to use offer sheet A to run through some numbers.

I'm also going to use our best guess as to the true player ratings from other thread.

AR = Apparent rating (1-100 scale)
ER = Effective Rating (375-625 scale)
and therefore ER = AR * 2.5 + 375

Which means that ouy two guys, with flat rtaings of 40 and 80 wouldn't have "effective ratings" that are in a 2:1 ratio, but rather something like this:

AR 40 --> ER of 475
AR 80 --> ER of 575

I'm going to define cap number for a particular year as signing bonus/length of contract + base salary for that year. So for the first year of offer sheet A we have 2/4+1 or $1.5. This represents how much cap room the team will have invested in the player. The more cap room the player takes up the more likely it is he will be replaced by a cheaper player.

Next is the cut number which will be how much money the team will save in the up coming year if they cut the player. This is base salary for the year - (signing bonus/length)*(years reamaining on contract). So if the team would cut the player after signing him to offer sheet A before the year started they would "save" 1-(2/4)*(4) or $-1. The team would lose cap space by cutting the player because the signing bonus would all count against the cap the first year. The last year of the deal though the team would save a bunch of money 4-(2/4)*(1) or $3.5. This number will be important because if the team can save a bunch of money by cutting the player he will have less job security.

Player goodness is a rating for how much better a player is than a potential replacement player the team could sign off the street. The better the player is the less likely he is to be cut. I'm going to use the 375-625 scale for this. Also using the info from Jim's default data file:
Here's a list of the frequency Overall Player Ratings are assigned in the default data file:

600-625 - Rare Superstar - 0.3%
575-599 - Superstar - 1.4%
550-574 - Near Superstar - 1.8%
525-549 - Elite Starter - 1.7%
500-524 - Excellent Starter - 3.5%
475-499 - Very Good-Plus Starter - 4.3%
450-474 - Very Good Starter - 6.0%
425-449 - Good Starter - 16.2%
400-424 - Replacement-Level Starter (decent backup) - 27.2%
375-399 - Roster Filler - 37.6%

I fit this distribution with an exponential and got:
%of player = e^(11.1-0.019*ER)
The fit isn't perfect. It actually looks like this is two linear distrbutions put together. One from 375-475 and the other from 475 to 625. This would have been a mess to deal with though so I am going to use the exponential to define player goodness. I'm going to define a replacement player to 400 which may be a little low but makes the numbers look nice. This corresponse to a AR of 10. Plugging that into our exponential and we get %33 percent. Our player goodness formula then will be replacement level divided by the player's level.
goodness=33/e^(11.1-0.019*ER)
So our player who is rated 50/60 rates out at a goodness of 33/e^(11.1*0.019*500) or 33/e^1.6 or 33/4.95 or 6.66.

Notice that I'm not adjusting upwards or downwards on the players rating as the contract runs into each successive year. If we had a nice formula to tell us what we expect the player's rating two years down the line that we could then recalculate "goodness" for each year down the road.

Ok now we have all the parts let's put it together: Player's chances of making the team are:

(Minimum salary * player goodness)/(cap number + cut number)

So as the player's salary increases beyond the minimum his chances for getting cut go up. This is counter balanced by the player's goodness but if the player gets hurt and comes back a shell of himself the chances of being cut increase. If the percentage goes above 100% the palyer is a "lock" to make the roster. Testing the lower bound a replacement player signed to a contract with no signing bonus at the minimum salary would have a %50 chance of making the team. Not perfect but pretty reasonable. Let's run this on Offer Sheet A from before.


Offer Sheet A
4yrs, $10 million
Bonus $2,000,000
yr base cap cut %chance
1 1 1.5 -1 666
2 1 1.5 -.5 333
3 2 2.5 1 95
4 4 4.5 3.5 41


So this is a backloaded contract with a smaller signing bonus. The player appears pretty safe until the last year of the contract which he could become a cap casualty.

To find the expected money earned take the signing bonus and then add each base year pay multiplied by the % chance of making the team. This means that Offer Sheet A has an expected value of $7.54 and the player should treat it as a 4 year $7.54 million dollar offer.

Looking at the other offers

Offer sheet B is 4 years $9.66
Offer sheet C is 4 years $5.66
Offer sheet D is 2 years $5.66
Offer sheet E is 2 years $4.66
Offer sheet F is 2 years $3.33

If you would like to take this one last step we can divide the total expected salary by the number of years to get an annual expected salary we get:

Sheet A $1.89/yr
Sheet B $2.42/yr
Sheet C $1.42/yr
Sheet D $2.83/yr
Sheet E $2.33/yr
Sheet F $1.66/yr

So this formula would say:D,B,E,A,F,C which matched up pretty well with what people were thinking. Assuming, of course, that I got all the math right.