I've done it! I've cracked it! I think I've finally come up with a way to rate fantasy player performances that reflect the most important fantasy attribute, consistency. Before I get into the nuts and bolts of this rating system let me review my preferred brand of fantasy football:
1) Partial points
2) Individual defensive players (IDPs)
3) 1/2 PPR (a full point for a catch is a bit much for my tastes)
4) Points for return yards (C. Patterson, Percy Harvin, Antionio Brown are all legit fantasy players)
5) TDs are worth 6 points for every player
6) Season ends Week 16 not Week 17 (if you disagree you obviously didn't have Jamaal Charles this year)
Okay, now that that's that (isn't the English language great?) let me explain Fant SPR.
The key to this rating are two statistical measures, average points per week and the Consistency Index (CI). Judging fantasy players based on total points can be very misleading. Looking at last year you would have thought David Wilson was better than Darren Sproles. If you still think that then maybe you should stop reading. If you're concerned about consistency a more functional point metric for players is how many points per week they average. That gives you a better idea on what to expect from a player and it puts seemingly large total point differentials into the proper perspective. The difference between a QB that finishes the year with 500 points vs. a QB with 450 is only 3 points per week. That's not a big difference especially if QB1 dropped 75 in week 8 and averaged 27 points in the other weeks.
The next rating used for Fant SPR is the Consistency Index or CI. This ones a bit complicated, but essentially it's the standard deviation of the average weekly points scored. The lower the the CI the more consistent the player is from week to week. Here's the twist. I give weight the CI based on how a player performs in the playoffs. If your league has a roster and points configuration like mine then a total team score north of 220 points is usually good enough to win a game in any given week, especially the championship game. Given this total value you can calculate position values that will get you to that total (28 points or higher for a QB, 10 points or higher for kickers, and 14 points or higher for every other position in my league). So if a player in weeks 14-16 scores at or higher than those thresholds (28 or 14, I don't bother with kickers) then I give their CI a boost (the number is lowered). Likewise if a player at any point during those final three weeks scores below a certain threshold (20 or 10) I drop their CI (the number is increased) because a poor performance in the playoffs at a position you depend could mean a playoff loss. Take A.J. Green vs. Megatron this year. Their point total and point averages were 248.5 to 278.2 and 16.6 to 18.55 respectively. Big difference huh? Who wouldn't want Megatron as their number one guy? Certainly not before they took Green right? Wrong! Green's playoff numbers (16.2, 13.8, 25.2) were way better than Megtron's (6.7, 14.8, 5.8), so much so that the later probably lost you a playoff game if you were counting on him for his average 15+ points. Based on my CI formula Green had a CI of 2.31 (2nd best in fantasy) and Megtron had a CI of 22.98 (awful!).
So, the final step in calculating Fant SPR is simple. Divide the weekly point average by the CI and multiply by 100. I like to think of Fant SPR as a percentage (or probability if you don't multiply by 100) of what you can expect from a player's average in a given week. Anything SPR over 80 is elite and any negative value is terrible. SPRs between 79 and 45 usually means a good player that may have failed to meet expectations in the playoffs. These are the guys you need to take a closer look at and see if in the playoffs they simply had slightly lower weekly points (i.e. 14, 10, 12.5) or if they completely crapped their pants in one game (i.e. 22, 5.3, 17.8).
If you choose to use Fant SPR then you'll have to decide the playoff threshold values for each position and how much of a bonus and how much of a penalty you're going to give players based on their week 14-16 performance. The value of this system is that it does a better job of telling you the players that are going to bring it home for your team and what players are fools gold. Based on my latest calculations the pool of elite fantasy talent is VERY small (this year at QB was particularly brutal). I'm talking 5 or less elite players at every offensive position.
I've already given you some insight on the players I like for next year so I won't change much of what I said based on this new system I've come up with (QBs are pretty much spot on, RBs need some tweaking). Stay tuned for my SPR rankings at each position and the players that are on my watch list for 2014.
If you like what you read don't be afraid to become a follower and comment. Cheers!
1) Partial points
2) Individual defensive players (IDPs)
3) 1/2 PPR (a full point for a catch is a bit much for my tastes)
4) Points for return yards (C. Patterson, Percy Harvin, Antionio Brown are all legit fantasy players)
5) TDs are worth 6 points for every player
6) Season ends Week 16 not Week 17 (if you disagree you obviously didn't have Jamaal Charles this year)
Okay, now that that's that (isn't the English language great?) let me explain Fant SPR.
The key to this rating are two statistical measures, average points per week and the Consistency Index (CI). Judging fantasy players based on total points can be very misleading. Looking at last year you would have thought David Wilson was better than Darren Sproles. If you still think that then maybe you should stop reading. If you're concerned about consistency a more functional point metric for players is how many points per week they average. That gives you a better idea on what to expect from a player and it puts seemingly large total point differentials into the proper perspective. The difference between a QB that finishes the year with 500 points vs. a QB with 450 is only 3 points per week. That's not a big difference especially if QB1 dropped 75 in week 8 and averaged 27 points in the other weeks.
The next rating used for Fant SPR is the Consistency Index or CI. This ones a bit complicated, but essentially it's the standard deviation of the average weekly points scored. The lower the the CI the more consistent the player is from week to week. Here's the twist. I give weight the CI based on how a player performs in the playoffs. If your league has a roster and points configuration like mine then a total team score north of 220 points is usually good enough to win a game in any given week, especially the championship game. Given this total value you can calculate position values that will get you to that total (28 points or higher for a QB, 10 points or higher for kickers, and 14 points or higher for every other position in my league). So if a player in weeks 14-16 scores at or higher than those thresholds (28 or 14, I don't bother with kickers) then I give their CI a boost (the number is lowered). Likewise if a player at any point during those final three weeks scores below a certain threshold (20 or 10) I drop their CI (the number is increased) because a poor performance in the playoffs at a position you depend could mean a playoff loss. Take A.J. Green vs. Megatron this year. Their point total and point averages were 248.5 to 278.2 and 16.6 to 18.55 respectively. Big difference huh? Who wouldn't want Megatron as their number one guy? Certainly not before they took Green right? Wrong! Green's playoff numbers (16.2, 13.8, 25.2) were way better than Megtron's (6.7, 14.8, 5.8), so much so that the later probably lost you a playoff game if you were counting on him for his average 15+ points. Based on my CI formula Green had a CI of 2.31 (2nd best in fantasy) and Megtron had a CI of 22.98 (awful!).
So, the final step in calculating Fant SPR is simple. Divide the weekly point average by the CI and multiply by 100. I like to think of Fant SPR as a percentage (or probability if you don't multiply by 100) of what you can expect from a player's average in a given week. Anything SPR over 80 is elite and any negative value is terrible. SPRs between 79 and 45 usually means a good player that may have failed to meet expectations in the playoffs. These are the guys you need to take a closer look at and see if in the playoffs they simply had slightly lower weekly points (i.e. 14, 10, 12.5) or if they completely crapped their pants in one game (i.e. 22, 5.3, 17.8).
If you choose to use Fant SPR then you'll have to decide the playoff threshold values for each position and how much of a bonus and how much of a penalty you're going to give players based on their week 14-16 performance. The value of this system is that it does a better job of telling you the players that are going to bring it home for your team and what players are fools gold. Based on my latest calculations the pool of elite fantasy talent is VERY small (this year at QB was particularly brutal). I'm talking 5 or less elite players at every offensive position.
I've already given you some insight on the players I like for next year so I won't change much of what I said based on this new system I've come up with (QBs are pretty much spot on, RBs need some tweaking). Stay tuned for my SPR rankings at each position and the players that are on my watch list for 2014.
If you like what you read don't be afraid to become a follower and comment. Cheers!
No comments:
Post a Comment