PR After Week 12

First let's take a look at this week's prediction results. I instantly noticed that my 1 and 2 both lost without looking at the numbers. I was a bit disappointed, but did find some diamonds in the rough.

I will order Week 12 games by the discrepancy of their Power Ranking Score (PRS). This would reflect the confidence that my system has in it's pick. The pick is just assumed to be the team with the higher score.

PIT at BUF (Confidence = 802)
The system has PIT as a winner. This pick almost went awry, as you would have noticed if you heard about Buffalo's comeback heroics. PIT was still a winner so this was a CORRECT pick.

CAR at CLE (782)
The system picked CLE as a winner. Despite the close game, again, the system picked the CORRECT winner.

TEN at HOU (657)
The system picked TEN as a winner, but they were shut out, so this was a terribly INCORRECT pick.

NO at DAL (638)
The system picked New Orleans CORRECTLY.

STL at DEN (577)
CORRECT

NE at DET (576)
New England wins, CORRECT pick.

CIN at NYJ (538)
Another CORRECT pick.

TB at BAL (533)
Baltimore wins, CORRECT pick.

KC at SEA (452)
Kansas City, CORRECTLY picked.

PHI at CHI (332)
Chicago wins convincingly, the system picked PHI, its second favorite team, but it was the INCORRECT winner. You may also notice the dramatic drop in confidence. The system is getting to its more uncertain games.

GB at ATL (319)
The system's favorite team was beaten, meaning both its top 2 teams went down. Atlanta wins, making this an INCORRECT pick.

SD at IND (316)
The system picked Indianapolis, which lost. Yet another INCORRECT pick.

JAC at NYG (309)
The system CORRECTLY picked New York as the winner

MIN at WSH (309)
The system picked Washington to win INCORRECTLY.

SF at ARI (149)

The system was not very confident in its San Francisco win, but they won convincingly, making this a CORRECT pick.

MIA at OAK (51)
The system had the least confidence when picking Oakland to win, and they lost. Making this an INCORRECT pick.

You may have noticed that as the confidence rating dropped, so did the success of the system. The overall performance was 10-6.

For this week at least, the important cut off seemed to have been at 0.400 confidence. The system had a +400 confidence in 9 games, of which it picked correctly 8 times.

So without further ado, here are the updated power rankings (excluding HOU v PHI on Thursday)

32 CAR 0.010
31 ARI 0.054
30 BUF 0.058
29 MIN 0.159
28 HOU 0.173
27 DAL 0.180
26 CIN 0.183
25 SEA 0.232
24 DEN 0.237
23 TB 0.264
22 DET 0.295
21 JAC 0.313
20 SF 0.352
19 WSH 0.383
18 OAK 0.390
17 MIA 0.458
16 NYG 0.468
15 SD 0.573
14 TEN 0.653
13 CHI 0.686
12 ATL 0.708
11 IND 0.716
10 NYJ 0.800
9 CLE 0.816
8 KC 0.820
7 STL 0.831
6 NE 0.838
5 PHI 0.856
4 BAL 0.896
3 PIT 0.912
2 NO 0.915
1 GB 0.956

And here are the Picks for Week 13 and confidence values, from most confident to least:

STL over ARI (778)
NO over  CIN (732)
PHI over HOU (683)
GB over SF (604)
KC over DEN (583)
IND over DAL (536)
ATL over TB (444)
CHI over DET (390)
CLE over MIA (359)
TEN over JAC (341)
SEA over CAR (222)
SD over OAK (182)
MIN over BUF (101)
NYG over WSH (85)
NE over NYJ (37)
PIT over BAL (17)