Guess the NFL Football Scores!

Is it allright if other people make predictions in here as well? I'm new to CBS sports so I am not sure if this is just for you.

Guess the NFL Football Scores!

Absolutely - feel free to post your predictions as you would, but don't make the posts too long. I put up the NFL team logos, so that you can use them in the predictions if you like, rather than having to type out team names. I encourage participation from everyone

Week 1 Predictions

Week 1 Predictions:

(ALL TIMES ET)

Thu, 8:30

23 @ 27

Sun, 1:00

20 @ 26

  6 @ 27

30 @ 23

20 @ 27

40 @ 16

20 @ 23

34 @ 30

20 @ 44

27 @ 12

Sun, 4:15

23 @ 33

16 @ 20

20 @ 27

Sun, 8:20

20 @ 27 (*)

Mon, 7:00
16 @ 33

Mon, 10:15

29 @ 20

Guess the NFL Football Scores!

All picks made against the line

-4  over Mes thinks miami overachieved last year not this year
-2.5 over this one tough wouldnt be surprised if carolina takes this one
-13 over    wow k.c. not good                                                   
-5 over        Palmer  yes            Orton  Ummmmmmm?
-4 over   Minny laying only 4 in Cleveland, Was 7 until Favre signed, should be more like 14

+ 4.5 over   First upset     Sanchez Smanchez  jets just better overall team

-7 over   Stop Jones Drew...check   Stop who else? lol who knows

-13 over   13 a lot in the nfl, leos could use another 13, dont be fooled, these cats dont change their stripes

+5.5 over     after this one jerry jones calls to see if charles rogers or mike williams available, I mean roy williams has had     such a solid nfl career lol
-6.5 over this game usually tight, not this time, gmen destroy rmen

Guess the NFL Football Scores!

Okay where's your blog for Week 2. Ready to make some picks

Guess the NFL Football Scores!

Should be up by tonight, actually. I almost finished yesterday, but I got sidetracked. I think the projections should be up by roughly around 7:30 ET

Week 2 Predictions

Week 2 Predictions:

( ALL TIMES ET)

Sun, 1:00

16 @ 24

47 @  16

23 @ 30

23 @ 27

27 @  24

23 @ 20

34 @ 33

16 @ 20

13 @ 34

Sun, 4:05

20 @ 27

20 @ 17

Sun, 4:15

20 @ 19

20  @  24

27 @  17

Sun, 8:20

27 @  23 (*)

Week 2's Featured Matchup! (*)

Mon, 8:30

23 @  13

Guess the NFL Football Scores!

Okay where's your blog for Week 2. Ready to make some picks A little late perhaps, but I posted my picks nonetheless. I will do the analysis portion later. I expect fewer surprises this week than in week 1. I really do see the Raiders beating the Chiefs at their turf - that might raise a few eyebrows. I'm expecting Matthew Stafford to pull off something similar to Jake Delhomme against the Vikings purple people eaters - INT TD for the Vikings is very possible and that's why I have them down for 47 points. I think Saints @ Eagles and Steelers @ Bears will be extremely close toss-ups, but I picked the away team each time. The rest requires very little explanation, but quite a few details. I'm liking the excitement in the matchup Giants @ Cowboys , but I think the Giants' D will cause an uproar. (Hey, let's see if any of the punters can hit that big ol' screen thing in the game. They will take aim at practice, I suppose)

Well, here are my official picks! Love them or hate them, they're all here!

Accuracy Formula

Accuracy Formula - What is it, anyway?!

The accuracy formula I use is actually pretty simple. It's as easy as 1-2! (No '3' required.)

Team A vs. Team B

Let:
Ap = Team A predicted points
As = Team A scored points
Ax = Team A prediction accuracy
Bp = Team B predicted points
Bs = Team B scored points
Bx = Team B prediction accuracy
Tx = Total prediction accuracy

STEP 1. Find the accuracy of each prediction:

100% - % error

100% - [(|Ap - As|) / (21 + As)] = Ax
100% - [(|Bp - Bs|) / (21 + Bs)] = Bx

STEP 2. Find the total accuracy:

(Ax + Bx) / 2 = Tx

---

Let's take some examples from last week's games.

Example 1: DET @ NO

20 @ 44 [Final: 27 @ 45 (W)]

Ap = 20
As = 27
Bp = 44
Bs = 45

STEP 1

100% - % error

100% - [(|Ap - As|) / (21 + As)] = Ax
100% - [(|Bp - Bs|) / (21 + Bs)] = Bx

STEP 1.1

100% - [(|20 - 27|) / (21 + 27)] = Ax
100% - [(|-7|) / (48)] = Ax
100% - [(7) / (48)] = Ax
100% - [.1458] = Ax
100% - 14.58% = Ax
Ax = 85.42%

My Team A prediction accuracy is 85.42%. (It's a ballpark average for acclaimed experts at predicting football games. This is more a high-end prediction for me.)

STEP 1.2

100% - [(|44 - 45|) / (21 + 45)] = Bx
100% - [(|-1|) / (66)] = Bx
100% - [(1) / (66)] = Bx
100% - [.0152] = Bx
100% - 1.52% = Bx
Bx = 98.48%

My Team B prediction accuracy is 98.48%. (Which, by the way, is remarkable - talk about prediction marksmanship!)

STEP 2

(Ax + Bx) / 2 = Tx

(85.42% + 98.48%) / 2 = Tx
(183.90%) / 2 = Tx
Tx = 91.95%

My total prediction accuracy is 91.95%. For purposes of practicality, we will round it off to 92.0% (This was an EASY one! Let's get to the hard ones, though... things will get ugly.)

FINAL RESULT:

20 @ 44 [Final: 27 @ 45 (W), Accuracy: 92.0%]

---

Example 2: STL @ SEA

20 @ 27 [Final:   0 @ 28 (W)]

Ap = 20
As = 0
Bp = 27
Bs = 28

STEP 1

100% - % error

100% - [(|Ap - As|) / (21 + As)] = Ax
100% - [(|Bp - Bs|) / (21 + Bs)] = Bx

STEP 1.1

100% - [(|20 - 0|) / (21 + 0)] = Ax
100% - [(|20|) / (21)] = Ax
100% - [(20) / (21)] = Ax
100% - [.9524] = Ax
100% - 95.24% = Ax
Ax = 4.76%

My Team A prediction accuracy is 4.76%. (Good lord, what an absolutely horrible prediction! It is possible to have a negative prediction accuracy, by the way.)

STEP 1.2

100% - [(|27 - 28|) / (21 + 28)] = Bx
100% - [(|-1|) / (49)] = Bx
100% - [(1) / (49)] = Bx
100% - [.0204] = Bx
100% - 2.04% = Bx
Bx = 97.96%

My Team B prediction accuracy is 97.96%. (I have redeemed myself from my first prediction, in a big way!)

STEP 2

(Ax + Bx) / 2 = Tx

(4.76% + 97.96%) / 2 = Tx
(102.72%) / 2 = Tx
Tx = 51.36%

My total prediction accuracy is 51.36%. For purposes of practicality, we will round it off to 51.4%. (I may have won the battle and the game, but I lost the accuracy war!)

FINAL RESULT:

20 @ 27 [Final:   0 @ 28 (W), Accuracy: 51.4%]

---

Example 3: JAC @ IND

34 @ 30 [Final: 12 @ 14 (L)]

Here's the total accuracy of not only a game where I incorrectly predicted the winner, but also a game with very controversial implications.

Ap = 34
As = 12
Bp = 30
Bs = 14

STEP 1

100% - % error

100% - [(|Ap - As|) / (21 + As)] = Ax
100% - [(|Bp - Bs|) / (21 + Bs)] = Bx

STEP 1.1

100% - [(|34 - 12|) / (21 + 12)] = Ax
100% - [(|22|) / (33)] = Ax
100% - [(22) / (33)] = Ax
100% - [.6667] = Ax
100% - 66.67% = Ax
Ax = 33.33%

My Team A prediction accuracy is 33.33%. (Seriously, though - I'm not THIS bad at predicting games, in general! Let's admit it - we ALL make mistakes.)

STEP 1.2

100% - [(|30 - 14|) / (21 + 14)] = Bx
100% - [(|16|) / (35)] = Bx
100% - [(16) / (35)] = Bx
100% - [.4571] = Bx
100% - 45.71% = Bx
Bx = 54.29%

My Team B prediction accuracy is 54.29%. (Close game predicted - that did happen. But I miscalculated BADLY.)

STEP 2

(Ax + Bx) / 2 = Tx

(33.33% + 54.29%) / 2 = Tx
(87.62%) / 2 = Tx
Tx = 43.81%

My total prediction accuracy is 43.81%. For purposes of practicality, we will round it off to 43.8%. (I could settle with 87.6%, but 43.8%?! Cut me a break, please - I'm actually less than half accurate! )

FINAL RESULT:

34 @ 30 [Final: 12 @ 14 (L), Accuracy: 43.8%]

---

Example 4: NYJ @ HOU

20 @ 23 [Final: 24 @    7 (L)]

Now here's a game where, while I incorrectly predicted the winner, I attained a better accuracy than one of the games in which I did correctly predict the winner. This is the second best accuracy attained, of these examples. (Anybody notice that Ap = 20 in 3 of 4 of these examples so far?)

Ap = 20
As = 24
Bp = 23
Bs = 7

STEP 1

100% - % error

100% - [(|Ap - As|) / (21 + As)] = Ax
100% - [(|Bp - Bs|) / (21 + Bs)] = Bx

STEP 1.1

100% - [(|20 - 24|) / (21 + 24)] = Ax
100% - [(|-4|) / (45)] = Ax
100% - [(4) / (45)] = Ax
100% - [.0889] = Ax
100% - 8.89% = Ax
Ax = 91.11%

My Team A prediction accuracy is 91.11%. (Whew! Finally got that break I deserved!)

STEP 1.2

100% - [(|23 - 7|) / (21 + 7)] = Bx
100% - [(|16|) / (28)] = Bx
100% - [(16) / (28)] = Bx
100% - [.5714] = Bx
100% - 57.14% = Bx
Bx = 42.86%

My Team B prediction accuracy is 42.86%. (Karma ensures that the good balances the bad. I'm cool with that.)

STEP 2

(Ax + Bx) / 2 = Tx

(91.11% + 42.86%) / 2 = Tx
(133.97%) / 2 = Tx
Tx = 66.99%

My total prediction accuracy is 66.99%. For purposes of practicality, we will round it off to 67.0%. (This one actually wasn't that bad, but I usually do better than this. The fact I incorrectly predicted the winner does provide solace, though.)

FINAL RESULT:

20 @ 23 [Final: 24 @    7 (L), Accuracy: 67.0%]

---

So here are some examples. I hope you all understand, or at least have an idea, of how the accuracy thing works. I was a bit harsh on myself for choosing some bad picks for the most part, but I think it's fitting, since we learn more from our losses than our gains.

And in case you're wondering, I don't arbitrarily do these calculations step-by-step myself every time! I use Microsoft Excel to do my accuracy calculations. I just displayed the arbitrary method in my example to show all of you how this gizmo works. I double-checked the results in my examples using the arbitrary method, with the Excel calculations INDEPENDENTLY... and the results matched .

Also, in case you're wondering what the '21' is for, in [(|Ap - As|) / (21 + As)] and [(|Bp - Bs|) / (21 + Bs)], it is meant to provide levity in predicting results of low scoring games (where accuracy per point differential tends to decrease) AND to avoid situations of dividing by 0 (which can neither be done nor allowed), when teams don't score. I chose '21' in particular, because historically, that is the average amount of points NFL teams score per game - overall, the purpose of that number is to provide a counterbalance for comparing my predictions to the actual results of NFL games, where the actual teams average roughly 21 points per game. (The teams averaged 20.3 points in week 1. In my predictions for week 1, the teams averaged 24.3 points - a bit too high. I obviously underestimated D-Fence!) This also means that the accuracy per point differential is TWICE as lower when the points scored is 21 (2.38%), than what it is when the points scored is 0 (4.76%). The system is just as fair as it is simplistic. Personally, I believe this design is foolproof.

If you have any questions or comments regarding my accuracy formula, feel free to ask!

---

Accuracy formula:

Let:
Ap = Team A predicted points
As = Team A scored points
Ax = Team A prediction accuracy
Bp = Team B predicted points
Bs = Team B scored points
Bx = Team B prediction accuracy
Tx = Total prediction accuracy

Tx = ([(|Ap - As|) / (21 + As)] + [(|Bp - Bs|) / (21 + Bs)]) / 2

Guess the NFL Football Scores!

The only thing I know is you did not pick the 49ers again. I hope we piss you off this week to!  Just kidding...good luck