FanGM

There are a lot of complicated discussions digging deep into long-term contract option price calculations. Feel free to dig deep into those in comments here, but for starters we'll do a simpler poll regarding the basic foundation of how numbers are determined.

Right now, I believe we take the top 10 simulation contracts for annual salary (+ roster bonus) for players at the same position with roughly the same grade (plus or minus 2). The LTC option for that player with various choices is currently based upon an average of those salaries of top 10 players at same position with similar grade.

There can be cases on an individual basis where perhaps a contract restructuring or out-performed lowly rookie deals or huge over-pays to free agents may throw the odd contract or two into those top 10 salaries to average out. There's room to further explore tweaks to that formula with considering all salaries across all leagues for larger sample and/or throwing out the extremes on both ends or expanding/restricting certain salaries it includes in the average, but that gets into more complicated discussion.

One might say it sucks to have an oddity like another GM overpaying at a particular position or a couple older vets at that position being forced into contract restructurings screw your team or mess up the average, while others may say that's the kind of stuff that can happen in real life also with unpredictable agent demands or odd market forces doing things to individual values that aren't always from everything-has-to-always-be-fair-land and getting a bad LTC value you see as unfair might be similar to having an uncooperative agent demanding too much or on the flip side for deals seen as unfairly too low you also see that in reality sometimes with some players staying home for less. If we're going to go through the trouble of generating a value, though, I'll agree some care should be put into trying to have the most reasonable numbers.

I think there's probably also been some discussion about having grades be so vital to that determination, when some players might have a current grade more people see as either too high or too low that could be different a couple months later when new grades come out. I'd say the way we're setup, we'll always have to use grades as some kind of standard for both determining individual options as in this case or for impacting team performance on sim field. Although always some disagreements about grades, some objective standard is vital for how we operate. As teams have a fairly large window for when they can sign a player to an LTC even while still under contract a year away from free agency, I'd say if they think a player's grade is going to drastically change ahead that they'd be advised to use their thinking there to plan out the best time to LTC a guy, not unlike reality where a team might try to lock a young player up earlier than they have to hoping for a medium sized deal if they fear he may be too expensive once he gains more proven value down the line. I can't really see situations where we have contracts shift in the future if a player's grade rises later either as I believe was suggested, as that would be so difficult to individually manage every player's performance changes and goes against how contracts have been written ahead fixed. If a player's salary next year was impossible to know until his new grade came out, I don't see that as something workable or realistic.

For the current basic question of basing LTC values around either sim contracts, real contracts or franchise tag values, the easiest would seem to be using the published franchise tag figures. If we just did that, I wouldn't have had to program the current contract lookup and averaging scripts we used last year. There would just be one figure to look up and adjust from that somehow (such as one recent suggestion of for every grade point less being a percent off the franchise tag value for that position), instead of having the system gather a collection of similar grade/position contracts. We have done that leg work already, though, so it's not as bad now to just add in some tweaks to the calculations now that it already gathers all the related sim deals. We could also attempt to gather up real contract values from NFL for players of that grade/position instead of using our sim contracts, but while the access to that information is better now than in the past perhaps there are issues and added labor involved in keeping up with that.

Let's first try to see if some strong feelings toward either using our sim contracts or franchise tag values as the basis of any LTC value. If there are strong leanings one way or the other, then we can get into how to exactly adjust a formula or process from one of those base figures.


Some of the complexities involved in creating a utopia of perfect fairness will be beyond what we can really do, and we can't hold fake sim player individual negotiations to be on a perfect case-by-case individual situation basis for every player's unique situation. The best we can do is try to put some process/formula together that put our generably reasonable figures for most of the players and leave it up to the teams to voluntarily take them or not, but even if numbers are generated a team doesn't like there are still tons of other options and avantages to retain the player otherwise beyond this one option that's just one of many.

I will say this again. The way it's done is top 10 within a couple points of a grade.

So, there's 100 within the top 10 of a B+... maybe 50 of those are on rookie contracts

But, there's probalby only 10 within the top 10 of an A+... and 5 of those are on rookie contracts. So, all 10 are averaged in.


Those averages will never change. Rookie contracts are going to exist forever. They will always be lower than free agent contracts. So, we have rookie contracts being averaged into the BEST of the BEST players, but not being averaged into just a good player's value. There will always be MANY MANY more B+'s than A+'s. These aren't numbers that can possibly fix themselves over time. Infact, it can only get worse, because A+'s won't hit the market. They'll get LTC'd, and that will continue to lower the averages over the long run.


Another option would be to state that rookie contracts can't be averaged. However, that adds to the work load and complexity.

Jared A wrote:I will say this again. The way it's done is top 10 within a couple points of a grade.

So, there's 100 within the top 10 of a B+... maybe 50 of those are on rookie contracts

But, there's probalby only 10 within the top 10 of an A+... and 5 of those are on rookie contracts. So, all 10 are averaged in.


Those averages will never change. Rookie contracts are going to exist forever. They will always be lower than free agent contracts. So, we have rookie contracts being averaged into the BEST of the BEST players, but not being averaged into just a good player's value. There will always be MANY MANY more B+'s than A+'s. These aren't numbers that can possibly fix themselves over time. Infact, it can only get worse, because A+'s won't hit the market. They'll get LTC'd, and that will continue to lower the averages over the long run.


Another option would be to state that rookie contracts can't be averaged. However, that adds to the work load and complexity.

I like the current system, however following up a bit based on Jared's post here, I think it would be fair to not count guys who are on their first contract. Those are almost always under 1 million dollars. By doing that it should raise the LTC figure by creating a bit of a fence for how low of a salary is allowed to count. That may create more work for Troy though and we are trying to decrease the work load so It may not work.

I'm was going to try not to wade too much into the complications until I get a better sense of how the majority want to move toward (either staying with sim contracts as base or moving to franchise tag as base, etc.) but reading a lot about rookie contract impacts and wanted to clarify per my understanding.

The only way a rookie contract should show up within a LTC value calculation currently I believe (unless I'm off as it's been a while since we did those last off-season) is if that low rookie contract salary amount happened to be one of the top 10 highest salaries of all players at that position with a similar grade. If it's a position on the field that's generally reasonably well paid with roughly 10 teams or so having a similar veteran starter at that position making above rookie minimums (which is just 1/3 of the league), then rookie contracts shouldn't show up at all for that player's LTC calculation since it just looks at top 10 highest salaries for similar position/grade. If lots of rookies were showing up in a players LTC calculation of top 10 highest paid similar players, to me that might mean the league isn't paying guys at that position with similar grade values very much so arguably a lower LTC may be justified if few are paying anything higher.

I don't know that it would be reasonable to say on the whole that no rookie contracts can go into LTC calculations at all, because some teams may want to extend an LTC to a lower graded player they have high hopes for in the future if they don't have any higher graded LTC eligible guys and perhaps most of his similar peers might be similarly low paid guys on younger player contracts. We also don't currently designate one contract as any different from any others in the data to say this contact is a players first contract ever and cannot be used in the calculations at all without going back in and manually adding that kind of classification to all past deals one at a time. We could say that for any salary to be included in the averaged salary calculation it has to be above $x more easily knocking out any lower rookie deals, but that almost defeats the purpose checking what players generally make for lower graded players and would end up that no lower paid player could ever have his contract extended only the higher graded higher paid ones. The idea originally was you could extend any player so long as willing to pay the reasonable going rate for someone like him around the league, not just only higher paid guys.

Especially if we end up sticking with pulling in similar sim contracts as before, I'd probably want to look more at specific examples and find some of the areas to tweak some of the worst cases.

One small tweak could be to not only look at the top 10 highest salaries for similar players at same position, but also pull in any higher contracts for players at that position with a lower grade than grade range being checked so that we're getting a larger sample and more complete picture of what the market should be for players at that position at their general skill level.

Another small tweak if we noticed that elite players were getting too low of LTC offers (I don't remember if that was the case or not) might be having some minimums put in place at the highest grade levels to prevent some of those unwanted situations if the math tended to come out funny for the highest groups if a lot of rookies ended up with elite grades at that position. May have been suggested before, but perhaps something simple like any A rated players would have to have their LTC at least equal franchise tag amount or something as some kind of floor set for particular groups if the LTCs weren't generating as reasonably for them.

Although it's a little trickier, it may also be possibe to eliminate some of the restructured contracts from the averaged calculation if the system picks up on the usual restructured contract format.

Before we did the work to compare sim contracts, I was probably leaning more toward the franchise tag as a baseline to use for LTC value adjustments. That would be easier (although much of the work already done for sim contract comparisons even if needing tweaks) and would also be more stable without some of the funny sim market influences or odd contracts that may come from sim bidding. So I could see a reasonable shift to that as a baseline instead also.

But if we stay on track the way things are calculated now, there may be some issues with that but probably some small adjustments we can also make to avoid some of the biggest problems ahead and generate reasonable numbers on the whole.

I voted for option 2, but I am not sure if that was the right option. I really like the idea of extending it to the top 10 contracts of similar position players rated at the player's grade +2 or below.

I think the problem with rookie contracts comes from grade scarcity at some positions. It may be the case that there are just not enough players in a +/- 2 point window around a players grade. For example, players with high 90 rated grades are rare. Let's say Jason Witten's contract was expiring and his GM wanted to offer him a LTC. Witten is a 96 rated TE so his contract would be based on the top 10 paid TE's between 94 and 98. Right now, this group includes Rob Gonkowski, Jimmy Graham, and Aaron Hernandez all of whom are on cheap rookie deals. I don't think there are 10 other TE's with 94-98 ratings, so these 3 contracts would significantly drive down Witten's price. If we changed the formula to the top 10 highest paid TE's with 96 or less rating, the larger contracts of Antonio Gates (86) and Vernon Davis (91) would set a more fair market price. Even 79 rated Jacob Tamme's 1.6 million salary (in the CFFL) would be included instead of the <500,000 values from Gronk, Graham, and Hernandez.

I do remember one problem from last year, but I don't remember if we ever came to a fix. I saw a specific case where two RB's with the same grade had significantly different LTC prices (I think it was Tolbert and someone else). The difference came from Tolbert being listed as a FB/RB while the other player was listed as RB/FB. As a result, the system averaged the top 10 FB contracts for Tolbert and the top 10 RB contracts for the other player. Its a big job to go through and clean up all the player positions (not to mention deciding if Tolbert is currently really a FB or a RB). I think the better answer is to form position groups. Specifically along the lines of: QB, FB/RB, WR, TE, or (WR/TE all lumped together), Offensive line (like the franchise tag is computed the same for C or T), front seven, and DBs. I am in favor of a front seven position group since for the sake of the sim, the player grades are treated equally. Also, anything else gets extremely complicated as we have DE/DT, DT/DE, DT/NT, NT/DT, DE/OLB, LB/DE, and other positions. Lumping DB's together removes the complication of players switching between CB and S (sounds like a rarity, but I've had two players do it in the past two years).

Goodell wrote: One small tweak could be to not only look at the top 10 highest salaries for similar players at same position, but also pull in any higher contracts for players at that position with a lower grade than grade range being checked so that we're getting a larger sample and more complete picture of what the market should be for players at that position at their general skill level.

I fully agree with this. and think it would be a good safe guard.

Yeah I have to agree that is a very good safe guard. Guys ratings fluctuate to much not to look into it. Another thing on the LTC, we shouldn't put this to a vote. If we truly want to mimic the NFL contract wise, an automatic change to the LTC must take place. It cannot stay the same as it was. Having no change just keeps the rich, rich and the poor, poor. I'd like to throw out the old system all together and then find perhaps 3 options all concerning a different way to include making the LTC as realistic as possible. Right now, in DFFL Revis is the only corner with his grade in a bit of a range. With current figures and a 10% increase that leaves a possibility that he may only be paid 2.5 mill a yr under current LTC rules. I'd love to have him and pay that, but this league is supposed to mimic NFL, and LTC change is very necessary.

tino38 wrote: Right now, in DFFL Revis is the only corner with his grade in a bit of a range. With current figures and a 10% increase that leaves a possibility that he may only be paid 2.5 mill a yr under current LTC rules. I'd love to have him and pay that, but this league is supposed to mimic NFL, and LTC change is very necessary.

The way it's currently being discussed, though, there would no longer a small grade range window. For him, it would be any CB with any grade below him so top 10 of pretty much all CB contracts in the league so probably would be a decent LTC generated from looking at all CB salaries.

A couple of the small potential tweaks discussed could catch most of the extreme cases of unfairness to competitive balance and sim player worth, though. The biggest loophole that would need to be looked at for me would be if the LTC represented a paycut. I can't really see that the league in all fairness overall should force elite players with high demand on the market to have no choice but accept paycut extensions with discounted LTCs below their last market-driven value. An LTC shouldn't be a pay cut, it should be paying a little more to keep the player from getting bids on an open market. LTCs should be intended for the rising guys you want to keep and want to prevent bids from exploding their price and you're willing to pay them more for taking that freedom from them, not a tool to dish out paycuts plus keep them away from the open market too.

We had much discussion about LTCs (probably the most discussed topic this off-season) across multiple threads.

I had posted a link last week (which everyone can get to from team rosters) to check some of the discussed tweaks made to the LTC values generated and let me know if big issues noted. Not much response to that.

As mentioned, the biggest change to LTC values and seemed to be the most supported in discussions would be broadening the window of contracts used in the averages. Last year it was just a tiny grade window between +/- 2 grade points and left more room for odd results. This year seemed positively supported in discussion to look at all salaries of players at that position for all grades below also not just in a small window of grade ranges. Seems like we'll get more consistent averages there looking at larger samples of a group of players a sim agent might compare to his player's value.

A vote on using our sim contracts, real contracts, changing to an entirely different system (suggestions asked) or using franchise tag amounts instead was strongly supported to continue using our sim contracts as basis for LTC values. Whether to keep it exactly the same as last year or tweak it a little was pretty split vote. In the end, I think doing what seemed most supported in feedback and votes -- pretty similar to the LTC process we put in place last year but some minor tweaks to avoid the biggest of problems. Primarily getting rid of small sample odd outcomes by expanding the averaged salary calculations which seemed most supported of ideas in discussions. There was also some talk about LTCs not being a pay cut and ensuring players already paid more than the LTC average as elite player not being forced to take a pay cut and avoid free agency market. As I setup LTC options for this year, looking back on last year's tools, that was already part of the instructions/rules implemented the first year if not totally implemented in programming fully but should be now to prevent LTC pay cuts.

LTCs are a big issue. We could spend the next decade discussing them in the smallest of details with mixed opinions and debate, but probably can't delay free agency for 2013 until 2023. It's definitely something to keep an eye on and keep those suggestions/proposals coming toward future rule change improvement discussions. I think we'll have to keep monitoring the results and finding some of the cases where the outcomes aren't quite as intended and keep modifying the calculations going forward in future seasons.

I 100% believe it should be ALL salaries at, or bellow the player... top 10 averaged.