Greetings. There has been discussion and trolling on which teams are overrated or not by claiming strength of schedule is better or worse by eyeballing records. For starters before this conversation is useful one has to make their case if they think the FEI is remotely accurate and if not what is. Do you really think the AP or the playoff committee is better? One of the things that I like about FEI is its attempting to measure the real caliber of a team minus luck. College football with 9-12 possessions and 12 games is going to have a higher variance than let's say basketball or the worse MLB. If you think FEI stinks interested why, and is there anything better.
What I think is interesting is there is this mantra of "you have not played anyone" If you look at the FEI page they have columns for records verses top10, 20, 30, 40. Of the top20 teams in the FEI, only 4 have beaten a top10 team. Part of this may be statistics. If your target team is a top10 team, threre are only nine others to play. Any team your team beats is going to have their statistics degraded by playing your team regardless of strength of schedule adjustments. The only teams in the top10 FEI with wins over other top10 teams are ND-->UM, Bama-->LSU, LSU-->Georgia. This low rate of play continues for the FEI top 20. The number of times a top 20 team has played another top 20 team is 34 times with the record being 17 wins and 17 losses. After 11 games the average number of times a top 20 team has faced another comparable team is a little over a game and a half.
What does this mean? If good football teams follow a normal distribution this makes sense. Most teams can't be great. If a group of good teams do get clustered, beating each other up at a rate of a .500 winning percentage will drag them all down.
In conclusion any power index system is a self fulfilling prophecy because of the nature of the low sample space. Good teams in a win are more likely to drop their defeated opponents. Even if that is mitigated by strength of schedule, there are not that many good teams. Random distribution says any team would only meet up to 2-4 even if the best teams are distributed in a few power conferences.
The reality of college football is that unlike the PRO game most of your time is spent beating up on overmatched opponents presuming you have a top10 team. Nobody in the top10 of the FEI has a great schedule as the majority of their games regardless of which team we look at, feasted on cupcakes. If anyone can name a team that has walked a college gauntlet to get where they are I'm interested in the counter example.
What I think is interesting is there is this mantra of "you have not played anyone" If you look at the FEI page they have columns for records verses top10, 20, 30, 40. Of the top20 teams in the FEI, only 4 have beaten a top10 team. Part of this may be statistics. If your target team is a top10 team, threre are only nine others to play. Any team your team beats is going to have their statistics degraded by playing your team regardless of strength of schedule adjustments. The only teams in the top10 FEI with wins over other top10 teams are ND-->UM, Bama-->LSU, LSU-->Georgia. This low rate of play continues for the FEI top 20. The number of times a top 20 team has played another top 20 team is 34 times with the record being 17 wins and 17 losses. After 11 games the average number of times a top 20 team has faced another comparable team is a little over a game and a half.
What does this mean? If good football teams follow a normal distribution this makes sense. Most teams can't be great. If a group of good teams do get clustered, beating each other up at a rate of a .500 winning percentage will drag them all down.
In conclusion any power index system is a self fulfilling prophecy because of the nature of the low sample space. Good teams in a win are more likely to drop their defeated opponents. Even if that is mitigated by strength of schedule, there are not that many good teams. Random distribution says any team would only meet up to 2-4 even if the best teams are distributed in a few power conferences.
The reality of college football is that unlike the PRO game most of your time is spent beating up on overmatched opponents presuming you have a top10 team. Nobody in the top10 of the FEI has a great schedule as the majority of their games regardless of which team we look at, feasted on cupcakes. If anyone can name a team that has walked a college gauntlet to get where they are I'm interested in the counter example.
Last edited: