I took some time today to make an update to the power rankings. One of the big issues with it, especially near the beginning of the season, is that it has been only considering the current season in the calculation ... which means that after one game there really is not enough data to resolve to anything realistic. I'm now using the previous season, slowly dropping earlier games, until we reach the season midpoint. The rankings will re-calculate tonight with this updated algorithm, so there will probably be some pretty significant movement.

While I was in there, I went ahead and put in the display for the odds in the schedule based on the power rankings. Again, tonight the numbers will be a little skewed, but after tonight's games it will be interesting to see what it says.

Doesn't seem to have made it better based on the new rankings this morning, at least from what I can see. Not sure what the best solution is, but after 4 games I don't think it is a good reflection of this season so far. Example, Carolina is 0-4, NO is 3-1, yet Car sits at 14 and NO at 18. I am not trying to be difficult with this or make it a big issue, just want it to make more sense when looking at it.

I personally think it should ONLY use the current season in the calculation (why factor in previous season data into a formula meant to reflect current season performance?), but maybe power rankings should not kick in or be displayed in until week 4 is done to have some data included after 25% of the season is complete?

However, I love the addition of the odds to the schedules. Thank you for that!

Really even 4 weeks doesn't give enough cross-league games to give a stable power ranking result, I feel like it really doesn't stabilize until week 8. The thinking for including the end of the previous season is that for the first part of the season you get a little grace for starting bad if you ended the previous season well, but by week 8 only the current season is being considered. Although I see what you're saying with it being odd-looking with NOS and CAR in this case.

The odds are interestingly huge, too, I may need to cap it at 14 points :-)

Honestly, I'd like to see these numbers without the cap. Would probably help refine the odds algorithm to see how accurate the 'ridiculous' numbers are.

I finds it interesting that Miami is a 10-point favorite over me since I have the better record. Sure, he has scored 4 more points than me over the season, but I've allowed 5 less.

I suppose this is because it's still considering last week. Or does the formula also consider stuff like strength of schedule?

Yes, it's considering strength of schedule. The algorithm is actually essentially the Excel 'solver' algorithm, getting as close as it can to produce the same scores that actually occurred, with a 3-point home field advantage (which, when I rolled this out, I did do some pretty extensive testing to make sure that is accurate as well).

Week 11 Game of the Week.

1-9 Denver hosts 4-6 Indianapolis.

You in charge of the TV Scedule`s too JDB? lol.

We finally won a game. Cleary America wants to see us ride this wave up.

Is there any way to have the odds / GOTW appear on the schedule, even if game results are hidden? They don't at current.