New Free Agency Idea

[Ulrich82]

I know we have to wait for the NFL labor deal to be settled before we put in place our future free agency rules, but I have an idea that might improve free agency bidding in the future. Of course, a major change like this would take some discussion, improvement, and testing, but I think the general idea would improve this game.

There has been some discussion about free agency and the lack of the ability to offer contract extensions to your own players. I also remember a discussion from several months ago about how to determine the best contract offer to a player. In my mind, one of the biggest faults is that contracts just go to the highest bidder despite the fact that in the NFL many other factors come in to where a player signs.

I propose two main changes. First of all, each contract offer is assigned a rating based on several factors (most of which would be the total money). For example, any or all of the following could be included:
1.) Total money offered
2.) Signing bonus offered
3.) Roster bonus offered (or signing + roster)
4.) Total contract length - maximum value for 4 year offers, 3 or 5 year deals are slightly less preferred, etc. (The idea is players like 4 year deals because they offer job security but are still short enough that the player will have a chance to score another signing bonus).
5.) Resigning - slight preference given to teams trying to resign their own player (so a player might resign with his current team for slightly less money than a new team)
6.) Average team wins over a three year period - players like to play for winners (yes, this puts a small disadvantage on rebuilding teams, I'll address this at the end of this post)

The idea would be to make a weighted average of all these variables to come up with a total score for each contract offered to a player. Of course, the money offered would make up the majority of the weight (probably 70 to 85%), we'd have to test the system a bit to figure out a fair way to set the weights (we don't want players passing up great contract offers to sign with their current team for pennies, but I think with a little testing, would could make sure this doesn't happen).

My second idea would be to use a random distribution to actually sign the player. This would take some discussion, but my first thought would be the following: We keep the same system in place. People offer contracts to free agents. The money of the highest contract offer will be listed on the page, but you won't know which team made the offer. Once the highest offer has not been outbid for 24 hours, an algorithm will run to determine who signed the player. The difference is, the highest bid won't necessarily win. Instead, the score of each contract offer will be used to construct a probability for each offer. Then, we can use a random number distribution to pick the winning contract (similarly to how the NBA draft lottery is done). This would help to simulate the real world situation where some free agents might choose to sign with a team he wants to play for for less money than for another team.

Yes, if we used the average winning percentage of teams to calculate the score, it would make a slight disadvantage for rebuilding teams to sign free agents. However, the actual weight given this parameter would probably be small. A rebuilding team could still beat out the super bowl winner by bidding more money (which probably mimics real life). Also, the random aspect of the selection would also keep this from ruling the market. Rebuilding teams would also be able to improve by solid drafting, trading, and identifying under the radar free agents. If anything, this would be a disincentive for teams to tank a year for a better draft pick.

IF we decided to investigate such a system, I'd be happy to pitch in on developing the math and code to make it work. It also would be helpful to have some archived information on free agents to experiment with what variables to consider and how to weight them.

Also, if we made the switch, it would have to be decided and announced a few seasons in advance so people can plan their future draft pick usage.

-Daryl

[redsoxfan31x21]

I like the idea, I'd definately back it.

Someone knows their math (stats) alittle bit there huh?