Abstract: our standings are too luck based. A good standings system would ignore the points scored by opponents, which suggests points scored as the indicator, but should also track how well a team was managed each week, rather than looking only at points scored in total. Essentially, points scored should be analyzed each week, then converted to win probability to account for the possibility that each team could have played any other team during any given week, and achieved a radically different final record.
I'm posting this cause I'm a goddamn math teacher with too much free time.
I think it's fair to say that we've all felt screwed by the FFL system at some point or another - whether it's that week you have the second best points total in the league but you played the top scorer, or you're just one of the poor saps who lost to Herman. Needless to say, the wins/losses standings system leaves a lot to be desired, as it is very susceptible to luck playing a huge factor. Take our end of year standings (bold made the playoffs):
Ronnie Vella Fanti Hagood George Joe Niki Tommy Santangelo Caponi Bourandas Cohen Ropke Herman
Personally, when I see those I just can't accept that they're right. Sure, we know they're going to be pretty good, I mean the same people pretty much always do well every year, so we can't blame it all on luck. But consider Bourandas, who scored more points this year than half the playoff teams, but got raped in the games he lost. He was a decent manager, but got shitty matchups. Completely out of his control. To hammer this point home, my team could have finished 12-1 had I had proper matchups, and that'd just be fucking wrong.
Looking at total points scored should clear things up a bit.
Ronnie Vella Joe Fanti Bourandas Niki Tommy Hagood George Ropke Santangelo Caponi Cohen Herman
Look for who dropped in the standings now, they got really lucky some weeks and benefited from low scoring opponents while they were not doing that well either. There isn't too much movement, but a few people move quite a bit, and there's definitely a different playoff picture.
Still, this doesn't say it all. As alluded to above, it's possible that a team could have the second highest point totals every week, yet play the top scoring team every time, and end up 0-13 despite having a great team and just shitty luck. Also, a high scoring team could feasibly score 250 points for a few weeks and then be completely mismanaged and fall apart: high point totals does not necessarily indicate a good manager. Luck is still heavily involved and, more importantly, the season shouldn't be looked at considering just these single, total numbers. Every week presented each manager an opportunity to set their optimal lineup, so it would be wise to consider point totals by week, so we can see better how those points were distributed.
So what I set out to do was to basically consider win probability for a given week. I ranked each team by the points scored each week, assigning the highest scoring team 1 point and the lowest scoring team 14 points. In this way, the lower your points over the course of a season, the more likely you were to win over all. This would account for sad scenarios like the made up one about a team scoring the second best points every week but still losing: that team would still be ranked very highly under this system.
Which brings me to the power rankings. Check out the results:
Vella (66) Ronnie (69) Joe (71) Fanti (73) Niki (84) Bourandas (85) Hagood (95) Tommy (96) George (106) Ropke (109) Santangelo (109) Caponi (120) Cohen (135) Herman (147)
I think this paints a vivid picture of how the league really looks. For all the shit we gave Ronnie behind his back for being overrated, every week he put himself in a good position to win. But not as good as Vella.
Also we see real stratification here: a block of four pretty dominant teams, two decent ones, two more a little weaker...and then we see everyone who wouldn't make the playoffs with over 100 points.
This makes intuitive sense, too. A completely average team (a team in the middle of the pack every single week) would have 97.5 points at the end...everyone better than average made the playoffs in this system, everyone worse, didn't. You can't say that for the other rankings.
