Islander
Footballguy
I use Gray's strength of schedule and find it valuable. For my LBs for example, this season so far there is 28% correlation between the predicted SOS right before each game and the actual number of points scored by my LB in that game. Not bad. Of course the higher the correlation, the better or the more accurate Gray's SOS is.
I think there would be a way to improve the accuracy of the SOS.
Would it be possible to break out DEs and DTs from the DL SOS? Break out CBs, SS, and FS from DBs. Break out MLBs/ILBs from the OLBs, and maybe have a separate category for WLBs and SLBs... for that one it's not as clear what should be done.
I believe some teams run more up the middle so the matchup could be good for a MLB but not as good for a WLB, whereas other teams might be sacked often by a rushing WLB.
For DLs, some matchups are good if you're a DE because the offensive tackles are weak... but that same matchup might not be as good if the team is strong in the middle of the offensive line. To be fancy, you could even separate LDE from RDE because one of the two tackles might be very good and the other one mediocre. Obviously most FFLers know these things, but it's nice to put them in a numerical way whenever possible.
Another suggestion: the SOS should vary whether the team plays at home or on the road. I am assuming that LBs tend to score more when they are on the road because the home team is likely to run more offensive plays. Therefore, if playing against team X represents 26.0 in the SOS, maybe instead of plugging 26.0 everywhere for teams playing against team X, maybe it should be 25.0 vs 27.0 for home/road games. For DEs it might be the opposite, I don't know. If the road team is the underdog, plays from behind, and has more pass attempts (so more sack opportunities), it might overcome the lower number of plays for DEs. I guess that's more questionable than for LBs.
Thoughts?
I think there would be a way to improve the accuracy of the SOS.
Would it be possible to break out DEs and DTs from the DL SOS? Break out CBs, SS, and FS from DBs. Break out MLBs/ILBs from the OLBs, and maybe have a separate category for WLBs and SLBs... for that one it's not as clear what should be done.
I believe some teams run more up the middle so the matchup could be good for a MLB but not as good for a WLB, whereas other teams might be sacked often by a rushing WLB.
For DLs, some matchups are good if you're a DE because the offensive tackles are weak... but that same matchup might not be as good if the team is strong in the middle of the offensive line. To be fancy, you could even separate LDE from RDE because one of the two tackles might be very good and the other one mediocre. Obviously most FFLers know these things, but it's nice to put them in a numerical way whenever possible.
Another suggestion: the SOS should vary whether the team plays at home or on the road. I am assuming that LBs tend to score more when they are on the road because the home team is likely to run more offensive plays. Therefore, if playing against team X represents 26.0 in the SOS, maybe instead of plugging 26.0 everywhere for teams playing against team X, maybe it should be 25.0 vs 27.0 for home/road games. For DEs it might be the opposite, I don't know. If the road team is the underdog, plays from behind, and has more pass attempts (so more sack opportunities), it might overcome the lower number of plays for DEs. I guess that's more questionable than for LBs.
Thoughts?
