Superbowl XLIX Projection, Week 15, Final

New England over Seattle

Yesterday I got a comment asking about the ridiculous looking ranking of having Dallas below Houston, despite the 3-game difference in record, and Dallas having actually beaten Houston.

To address the second point first, Dallas won by 3 in overtime, so that's practically a tie in terms of using head-to-head to determine who is better.

As for the Strength column, that's determined purely by points scored and points allowed, which is basically Bill James' baseball formula for determined a team's Pythagorean Winning Percentage. The theory goes that a team that allows about as many points as it scores will be about .500 in its future games, regardless of if it managed to pull off a bunch of small wins and few huge losses, or vice versa. That's why 7-7 San Francisco is listed at .460, but 7-7 Miami is at .554, and 7-7 Houston is at .601. A team could even be 1-13 at this point, but if they won a game 62-0 and lost the rest 2-0, they'd be near the top. The actual formula I use is somewhat complex, involving exponents, league-wide points per game, and regressing to .500 based on how far into the season it is, but there is a simplification that can be made for purposes of comparing 2 teams: the higher a team's pf/pa ratio, the higher their strength will be, where pf = points for, and pa = points against.

Houston has scored 324 point against 277 allowed. Dallas has scored 381, but allowed 328. That gives Houston a slightly higher ratio of pf/pa, and the corollary to that is that their strength is at .601, where Dallas is at .596. Houston's points have not been distributed as favorably for them, though, so they've won 3 less games than Dallas. The meaning of those strengths in my projections is that Houston would win 50.52% of the simulations if they played again. That's just barely an advantage.

One caveat to mention is that this is supposed to represent an average of the teams over their last 14 games, so the average Houston lineup over the last 14 games would be slightly favored to win over the average Dallas lineup. This doesn't mean the Houston lineup with the 3rd string QB would be favored to win this week against the mostly-intact Dallas lineup. It does mean that Arizona, even at 11-3, is likely even worse than their 11th-best ranking, since they're now on QB number 3 as well.

Hopefully all that made some sense. I realize I did a little hand-waving to get from pf/pa to strength, but I didn't want to post a full screen worth of LaTeX equations to show my work.

TEAM	Record	Strength	Div	Plyff	WC	Qtr	Semi	Champ
NE	11-3	0.744	+	+	98.917	61.429	37.150	22.659
BAL	9-5	0.706	26.987	73.909	44.150	21.811	11.748	6.630
SEA	10-4	0.703	49.701	94.796	80.685	48.578	28.434	14.621
GB	10-4	0.681	51.029	96.997	78.036	44.581	24.509	11.958
IND	10-4	0.679	+	+	57.805	28.714	13.966	7.448
DEN	11-3	0.656	+	+	98.504	50.432	22.159	11.254
KC	8-6	0.649	-	35.861	18.771	8.091	3.710	1.855
BUF	8-6	0.630	-	17.482	8.350	3.611	1.588	0.763
PHI	9-5	0.616	40.604	45.447	21.706	10.103	4.782	2.000
DET	10-4	0.606	48.971	94.094	69.010	33.329	14.891	6.091
ARI	11-3	0.604	50.299	+	82.116	36.824	16.705	6.794
HOU	7-7	0.601	-	4.283	1.937	0.684	0.306	0.138
DAL	10-4	0.596	59.396	68.666	36.497	16.368	7.354	2.936
PIT	9-5	0.589	33.707	72.212	32.967	12.365	4.864	2.127
MIA	7-7	0.554	-	0.052	0.020	0.006	0.003	0.001
CIN	9-4-1	0.548	39.307	78.341	31.733	10.784	3.800	1.510
SD	8-6	0.520	-	17.859	6.846	2.073	0.706	0.262
STL	6-8	0.487	-	-	-	-	-	-
NO	6-8	0.482	57.403	57.403	19.430	6.513	2.203	0.653
ATL	5-9	0.462	28.830	28.830	9.234	2.929	0.938	0.263
SF	7-7	0.460	-	-	-	-	-	-
NYG	5-9	0.456	-	-	-	-	-	-
MIN	6-8	0.455	-	-	-	-	-	-
CLE	7-7	0.446	-	-	-	-	-	-
CAR	5-8-1	0.363	13.767	13.767	3.286	0.774	0.184	0.038
CHI	5-9	0.304	-	-	-	-	-	-
WAS	3-11	0.283	-	-	-	-	-	-
TB	2-12	0.281	-	-	-	-	-	-
NYJ	3-11	0.244	-	-	-	-	-	-
TEN	2-12	0.214	-	-	-	-	-	-
JAC	2-12	0.195	-	-	-	-	-	-
OAK	2-12	0.193	-	-	-	-	-	-

First to 10% chance of Superbowl Win: Cincinnati (Week 3, Sunday)

First to 80% chance of Playoffs: Cincinnati (Week 3, Sunday)
First to 90% chance of Playoffs: Indianapolis (Week 7, Sunday)
First to 99% chance of Playoffs: Arizona (Week 11, Sunday)
First to 99.9% chance of Playoffs: Denver (Week 14, Sunday)
First to clinch Playoffs (sort of): Arizona (Week 15, Thursday)
First to clinch Playoffs (for real this time): Arizona, New England, Indianapolis, Denver (Week 15, Sunday)

First team eliminated: Oakland (Week 11, Sunday)

[Week 1]
[Week 2, Thursday]
[Week 2, Sunday]
[Week 2, Final]
[Week 3, Thursday]
[Week 3, Sunday]
[Week 3, Final]
[Week 4, Thursday]
[Week 4, Sunday]
[Week 4, Final]
[Week 5, Thursday]
[Week 5, Sunday]
[Week 5, Final]
[Week 6, Thursday]
[Week 6, Sunday]
[Week 6, Final]
[Week 7, Thursday]
[Week 7, Sunday]
[Week 7, Final]
[Week 8, Thursday]
[Week 8, Sunday]
[Week 8, Final]
[Week 9, Thursday]
[Week 9, Sunday]
[Week 9, Final]
[Week 10, Thursday]
[Week 10, Sunday]
[Week 10, Final]
[Week 11, Thursday]
[Week 11, Sunday]
[Week 11, Final]
[Week 12, Thursday]
[Week 12, Sunday]
[Week 12, Final]
[Week 13, Thursday]
[Week 13, Sunday]
[Week 13, Final]
[Week 14, Thursday]
[Week 14, Sunday]
[Week 14, Final]
[Week 15, Thursday]
[Week 15, Sunday]