By Josh Knust, Never Say Wombat
In all of our leagues, the standings that determine the playoffs only rely on two metrics, record and points scored by your team. This is fine for playoff seeding because all that really matters are wins and losses; ugly wins are still wins, just ask Dave-O in Week 2. For week-to-week predictions and discussion, I wanted to go a little deeper and create a Power Rankings for each of the teams in the league.
The goal of this whole experiment is to create a ranking system that better shows the true power of a team by incorporating coaching efficiency and league-wide breakdown in addition to base scoring and record. The rankings probably won’t be that much different than the ESPN standings, but under this system, a good team with a poor record won’t be dismissed because of some close losses or outlier weeks by their opponents.
For this first iteration,
base record is worth 25% weight,
average record against the field is worth 25% weight,
average points scored per week is worth 35% weight,
and coaching rank is worth 15% weight.
• Base record is calculated as win percentage times twenty-five to give a score out of twenty-five.
• Average record against the field is your win percentage against each other score for that week times twenty-five to give a score out of twenty-five.
• Points scored is calculated by taking your average points scored and dividing by the highest average point total (giving a percentage), then multiplying by thirty-five to give a score out of thirty-five.
• Coaching rank is calculated by taking your average points scored and dividing by the average “ideal lineup” score (giving a percentage), then multiplying by fifteen to give a score out of fifteen. The “ideal lineup” score is determined by going back and creating the top-scoring lineup for your team on a given week. This is the way that leaving points on the bench is incorporated into the ranking.
These for different scores are totaled and give you a Power Score out of 100. Now, the reason you are here, the rankings:
Lincoln sits atop the President’s League Power Rankings, and deservedly so. He has been impressive so far and is the lone undefeated team after Bianchi was toppled by an outstanding week from Austin Bamrick and The Mirthmobile. The Power Rankings look much like the ESPN Standings, but my gut says that as records start to diversify and with a larger sample size, we’ll see more differences.
How your Power Score ranks against the rest of the league is your Power Ranking (1-12). Keep in mind that even though you may be adjacent to someone on the Power Rankings, you can look to that Power Score to determine how close it stands. In the rankings from this week, look at the teams with 2-1 records. Between Bianchi and Knust, they have the same record, but Bianchi’s Power Score is significantly higher than Knust’s. Also note that Stewart is ranked higher than Denton and Jameson is ranked higher than Masker despite having worse records.
The math behind these rankings is subject to change, and I again want to welcome any suggestions. A lot of these metrics are based off of averages, but I want to look at ways to incorporate consistency with standard deviations of scoring. For example, when looking at T-Mike versus Dave-O (T-Mike is more highly rated under the current format), T-Mike has a population standard deviation of 22.3 of his weekly scoring totals whereas Dave-O’s is 12.1; this means that Las Tortugas is a much more consistent team, and the power ranking may favor that consistency and flip the order. Incorporating something like this would likely make the rankings more accurate.
Another thing that I want to look at incorporating is the idea of a ceiling for your team. For example, Sullivan manages his Phuket Rippers well, and gets 94.3% of his possible points from real score to ideal score. He just edges out Bianchi who gets 93.6% of his Hairy Lumberjack’s possible points. However, the Rippers only have the tenth highest ceiling, or points under an ideal lineup, whereas the Lumberjacks have the highest ceiling. I’m not sure at this point what this means, but it might be worth incorporating. Some coaches, like Sullivan, are great at managing bad teams, where some coaches are bad at managing good teams (Jameson’s Dictators rank 2nd in ceiling, but 11th in coaching efficiency). I don’t mean to dog on anyone here, just stating the results of my calculations.
If people like this content, I will continue to do it, and would be more than happy to help get the other leagues started on a Power Ranking system. If you want to see my Excel sheet, just let me know and I can email you with what I have. If you want to talk about the theory behind the rankings, message me and we can set up a time to talk.
2018 Assist Raffle Tracker
T. Shannon - 24
J. Edwards - 1