PR after Week 14

The system did decently last week. It went 11-5, which isn't too impressive considering that I am only picking winners.

BAL over HOU (835) Hit
ATL over CAR (781) Hit
CLE over BUF (754) Hit
PIT over CIN (624) Hit
GB over DET (535) Miss
PHI over DAL (467) Hit
KC over SD (409) Miss
TB over WSH (267) Hit
NYJ over MIA (238) Miss
NYG over MIN (222) Hit
NE over CHI (129) Hit
DEN over ARI (124) Miss
IND over TEN (123) Hit
JAC over OAK (009) Hit

For whatever reason, I mistakenly left out 2 games from last week. This is the result

SF over SEA (306) Hit

NO over STL (133) Hit

Here are the new Power Rankings:

32	CAR	0.009
31	BUF	0.064
30	ARI	0.091
29	HOU	0.099
28	DEN	0.145
27	CIN	0.147
26	SEA	0.168
25	MIN	0.211
24	DAL	0.249
23	TB	0.259
22	WSH	0.319
21	DET	0.330
20	JAC	0.409
19	TEN	0.432
18	SF	0.437
17	CHI	0.533
16	OAK	0.542
15	MIA	0.549
14	SD	0.565
13	NYJ	0.589
12	NYG	0.639
11	IND	0.705
10	STL	0.751
9	CLE	0.764
8	ATL	0.775
7	KC	0.776
6	PIT	0.854
5	PHI	0.884
4	NE	0.927
3	BAL	0.948
2	GB	0.949
1	NO	0.955

Green Bay's run at 1st ended for the first time so far. Interesting. Let's see the spreadsheet's picks for next week, as always, from highest confidence to lowest:

ATL over SEA (793)
CLE over CIN (608)
OAK over DEN (559)
MIA over BUF (361)
PIT over NYJ (342)
CHI over MIN (262)
PHI over NYG (213)
DAL over WSH (201)
ARI over CAR (129)
TEN over HOU (102)
KC over STL (88)
NE over GB (85)
JAC over IND (59)
DET over TB (26)
BAL over NO (10)
SD over SF (1)

You may have noticed since last week that the spreadsheet isn't just simply picking the team with the higher rating to win the game. What I have done is skew the teams performance numbers more heavily in it's recent games, but only for the picks, not for the power rankings. So, power rankings are determined by the whole years performance, every single game being of equal importance. But the picks view the more recent games as more important. Let's see how it goes this week. We know we are at least 1 - 0 so far. The system picked SD in the Thursday night game with probably the lowest confidence ever (1!), but it worked out.