I did this, began preparing for it back in 2004 and we went live in 2005 and are now entering our 6th season.
I went with 12 teams because I thought it was the best option for a number of different things, including division setup, scheduling, playoff format, and how many starters you get per week with it.
Some things I aimed for:
* Offense is most important, but IDP defense is not trivial.
* Favor elements that give skill a chance to differentiate itself from luck.
* Wanted more than redraft.
* Didn't want a drastically dominant position like RB in normal leagues (which is part of favoring elements that draw on skill).
* I want a lot of interaction between owners. I want a lot of trading, so a setup that causes teams to have to deal with each other to fill some of their needs.
I did a few tests of scoring systems, dumping the results into VBD to see how it affected player value, etc. I asked around on the boards about other leagues... and Bob Harris was kind enough to give me a copy of his league rules which incorporated a lot of ideas for contracts which I liked.
End result:
12 team. Hard salary cap of $500. 45 to 55 player rosters. Start 1 QB, 2 RB, 1 flex QB/RB. 4 WR, 2 TE, 1 flex WR/TE. 2 DT, 2 DE, 3 LB, 2 CB, 2 Safeties. 1 place kicker. 1 punter. 1 team head coach. 1 team kick return unit. Scoring is mostly standard, -2 for turnovers, and staggered PPR of 0.25 RB, 0.5 WR, and 1.0 WR. The flex QB/RB should almost always be a QB, and the flex WR/TE is mostly a WR.
The result of the lineup is that there is approximate parity across the top of the positions. My projections for the first season had the top 4 players in order of value as Peyton, Gonzo, Moss and LT. The positions all fall off in value differently so there's quite a lot of room for a skilled owner to maximize his team.
Auction for vets, draft for rookies. Players get a 3 year contract at their winning bid. Rookie first round salaries we take all the vets, sort by salary, and take the bottom half of fantasy starters (so RB we take RBs 13-24 ranked by salary), average that group and that's a 1st round contract, with a minimum of $6 for offense and $3 for IDP. Price drops by 1/3 each round until 4th round everyone is $1.
After year 2 you can give a contract extension that adds another 1 or 2 years, at the cost of a raise that goes into effect immediately in year 3 (raise is greater of 20% for 1 year or $5 or 40 or $10 for the 2 year). When a player's contract is up you can use 1 franchise tag or 2 transition tags to retain them... they get average of top 5/10 players at their position or a 20%/10% raise. The Tag stays on the player through his 3 year contract, it doesn't "refresh" every year. (Neither does the NFL's... what they do is they remove the tag before signing the player to a long term contract or else the tag stays on him like ours does in fantasy). Tagged players are put in an auction and other teams can bid on them. At the end the original team can match the winning bid and keep them, or can accept compensation (two 1st round picks for franchise player, a single 2nd for Transition player).
Playoffs are like an NFL conference playoff format, 6 teams total making it with the top 2 seeds getting a first round bye. Regular season schedule, first two weeks are triple headers against division teams. Next 8 weeks are doubleheaders against out of division teams. Then 2 more weeks of triple headers. Result is you play your division 4 times and out of division twice. Regular season done week 12 and playoffs weeks 13-15, done by the time that NFL teams start resting people. The doubleheaders cause the final results to be more in line with how a team really did... it's less likely you ran into the high scoring teams every week that way, etc.
Rookie draft order for playoff team is worst to first, except that consolation games you need to win your game to pick ahead of your opponent, so there is incentive to win. The 6 non-playoff teams have draft order set by a combination of regular season order of finish and by order of finish in the consolation bracket. Regular season counts just over twice as much as the playoffs do.
Punters are a great addition to fantasy football. Let me say that again, punters are a great addition to fantasy football. If you want skill to matter more than luck. The top ten punters normally read like a who's who of the worst offenses and teams in the NFL. If you can predict those, you can predict who will be a good punter. We only give points for gross punting yards, as points for inside the 20 would lessen their predictability I believe.
Head coaches, same thing. We give a few points for winning their NFL game, and more points for margin of victory (3 for win and another 1/3 fantasy point for each point in margin of victory). Predict which NFL teams will win and you do well in fantasy. If you're good at picking winners you can even just play matchups with below average coaches if you miss out on a good coach. We use the Offense category on MFL for this, as it has the stats we need and the actual coach category they have uses named individual while we just want a "team" head coach.
Kick return unit I like as well. If you can predict which teams have good returners (i.e. Hester, Cribbs), and/or predict which teams will give up a lot of scores, you'll generally do well in selecting a kick return unit. So again, lends itself well to skill over luck.
Waivers, I wanted to have players be cut with their contract intact for a new team to pick up, unless they clear waivers in which case they would get a new contract if picked up. But there were some issues in getting the website to handle all the details, so we simplified it and now a player's contract is done whenever he's cut. His new contract is whatever he's won for in blind bidding waivers... or $1 in FCFS waivers between the last bidding window and start of games. So teams will often cut overpriced players during the season and try to pick them back up for cheaper. Sometimes this has had big ripple effects as other teams win the bid and drop bunches of worthwhile players to pay for him.
So sort of in summary, it's a league with very big rosters (26 starters each week, up to 55 players on a roster). People have to research rookies and know all the players on the teams, not just the top 20 or 30 at each position. You have to balance a player's value with his price, and to do well it helps to be willing to cut a good player whose salary is too much. With using so many players there is a scarcity issue at each position down near the bottom of the starters... very few teams have solid starters at every position plus have depth. And even when you do it's very hard to maintain because players contracts roll over or you have to give them raises to keep them, needing a constant influx of young players at cheaper draft prices.
The positions all have value. You can't just focus on getting RBs when there are 11 offensive players starting each week. With starting a full 4-3 defense, you can get burnt if you ignore it. One team has done very well with getting strong defenses that have made up for being weaker on offense (though I think that's a tough strategy to do consistently). Offenses tend to outscore defenses in VBD value about 60-40 which was very much by design... I tweaked the points for tackles and such until I got the ratio I wanted.
Everyone in the league who has given me feedback on it has said it's the best setup they've ever played in. Most may have some feature they might prefer another way... one guy hates the double headers, but overall people seem to like the setup and it is really putting people's skill to the test. The owners are pretty much all people I've played with in past leagues who did well... and it's become very apparent which of those owners had just learned the Stud RB theory and thrived off of it, compared to which know how to figure out what the value of players is in any league setup and can adapt.
16 teams every single day of the year. Its the perfect balance for a league if you have solid owners.
