BarleyNY Posted November 18, 2015 Share Posted November 18, 2015 Well, we are 5-3 in the conference, and they are 3-2. They have tougher games with two to the Raiders, and the Bills. The Browns could get back on track with McKown returning, and you never know with the Chargers twice as Rivers is a solid QB. You can see how this model is tracking if you look at not only KC and Buffalo, but also Oakland: KC: 4-5, 56% chance to make playoffs Buffalo: 5-4, 35% Oakland: 4-5, 7% The model is simply saying that the Bills have a tougher schedule (which they obviously do) and that currently KC is a much better team than Oakland, whom they still have to play twice. Things can change - injuries, development of players, hot streaks, etc. - but it's difficult to argue with those expectations today. Link to comment Share on other sites More sharing options...
GunnerBill Posted November 18, 2015 Share Posted November 18, 2015 You can see how this model is tracking if you look at not only KC and Buffalo, but also Oakland: KC: 4-5, 56% chance to make playoffs Buffalo: 5-4, 35% Oakland: 4-5, 7% The model is simply saying that the Bills have a tougher schedule (which they obviously do) and that currently KC is a much better team than Oakland, whom they still have to play twice. Things can change - injuries, development of players, hot streaks, etc. - but it's difficult to argue with those expectations today. I don't agree that KC are much better than Oakland. I think they are extremely closely matched and I think a split in those games is the most likely outcome. Link to comment Share on other sites More sharing options...
FloridaSnow Posted November 18, 2015 Share Posted November 18, 2015 I wonder how much those numbers would change if the game against KC was in Buffalo. That is a huge factor in the race, and the fact that it's in KC will make the Chiefs favorites. Link to comment Share on other sites More sharing options...
KD in CA Posted November 18, 2015 Share Posted November 18, 2015 I don't agree that KC are much better than Oakland. I think they are extremely closely matched and I think a split in those games is the most likely outcome. That is a nasty, old school rivalry. You never know how those Oak-KC will go. I agree a split is most likely. Link to comment Share on other sites More sharing options...
BarleyNY Posted November 18, 2015 Share Posted November 18, 2015 I don't agree that KC are much better than Oakland. I think they are extremely closely matched and I think a split in those games is the most likely outcome. That's certainly possible, even according to 538's model. But that model has KC (ELO of 1588) as a significantly better team than Oakland (1428) and thus gives KC a much better chance of finishing with the better record since they play twice and have similar schedules. Incidentally Buffalo's ELO is 1530, which is a lot better than Oakland but not quite as good as KC. If the Bills win they will improve their rating and odds. If they lose, it'll go the other direction. In that way I really like models like this because they force teams to earn respect. Emotion is minimized even if there is a lot of variance. Link to comment Share on other sites More sharing options...
GunnerBill Posted November 18, 2015 Share Posted November 18, 2015 But emotion and momentum are vital parts of team sports. I do see some validity in the modelling but at the same time the NFL is too much of an any given Sunday league to put too much store in it. Link to comment Share on other sites More sharing options...
BarleyNY Posted November 18, 2015 Share Posted November 18, 2015 But emotion and momentum are vital parts of team sports. I do see some validity in the modelling but at the same time the NFL is too much of an any given Sunday league to put too much store in it. Most of the gripes about modeling like 538's are from people who expect some nerd to be able to plug in a bunch of numbers and tell the future. If that doesn't happen they decry their work as bogus and completely write it off. It's just odds and expectations and only gets as specific as the model and available information allow. Plus modeling things with a lot of inherent variance - like football - is only ever going to be so accurate. It's a nice tool, not a be-all-end-all. Link to comment Share on other sites More sharing options...
GunnerBill Posted November 18, 2015 Share Posted November 18, 2015 Plus modeling things with a lot of inherent variance - like football - is only ever going to be so accurate. It's a nice tool, not a be-all-end-all. Yes I agree. Link to comment Share on other sites More sharing options...
Recommended Posts