New YPPT Method

hey roughshod. you started quite an epidemic here. keep up the great work guys

after you've made your adjustments, why not go back and re-predict scores for week 6 instead of waiting for week 7?

this is the craziest shit i have ever seen!!

you guys need to go get laid...

insanity.

BOL

Love this thread but not too confident in my work, I put everything in excel to do the calculations.  I am wanting to make sure I am pulling the correct numbers from statfox.. if someone could take the time to review this that would be great...thanks..

Your numbers look right to me. I came up with

Tulsa: 30.78

UTEP: 22.08

To me it looks too close to call but it may not be a bad idea to take the home dog. These are my calculations, but roughshod lost me with some of the changes.

Roughshod - I have a few concerns when it comes to college.

1) home field is way too important in college football.

2) To many junk games. 1-AA teams or just bad teams like Washington St.

3) Rankings are based too much off of perception and because of the number of teams in the country it can be way off. In the NFL they are more accurate since there is only 32 teams.

4) I think just taking into consideration their conference apponents would be a good way to go. That way the same level of competition is taken into consideration. Notre Dame would be a team that comes to mind that this may not work for - but my suggestion would be to only fade them. (that system has worked okay for me over the past 4 seasons).

Love the thread and great job with this...please keep updating us on this.

i think you guys are doing it wrong, cuz i got

tulsa 36
utep 19

do this "sagarin power rankings/sagarin schedule"
this is probably more accurate, because CFB has a wider variety of competition. Think about it, UTEP lost to Memphis and Buffalo and still managed to upset Houston. Yet the schedule average (according to sagarin) is only 6 points for UTEP. So there is a big flaw with the SOS multiplier when it comes to CFB.

the way you guys are determining the SOS multiplier is more useful for the NFL, because there is a much smaller and more similar teams.

also, i agree on HFA, but its not THAT important unless its a division game. the better teams come to win out of division more times than not.

also there is much faster ways to do this if you only want the projected score. but this formula is cool because it lets you determine who has could have a better run game (which is very important in 1 - 3 pt spreads) and who will have a defensive impact. its much more specific.

It probably has to do with some other numbers. I was using the Sagarin Schedule ranks. Tulsa 62.3 and UTEP - 68.77

Let me know what you mean by dividing the number.

Thanks,

I dont know if this is possible but is there any actual book with this system in it?   I would very much like to read a book with this information in it.  I sort of understand from what has been posted but I think if there is a book it would be very helpful because I would expect every little part of the formula to be broken down so it was easier to understand at first.  Im sure after you do it a bit getting the right numbers to put in certain spots isnt that hard but just looking at it right now I'm a bit lost and would love to go buy any book that you guys would know of that has this in it.  If there is no book then maybe a website? 

I figured it would be more helpful then me asking a thousand questions.

PM me or leave any info in the thread either works for me I really appreciate you sharing this and trying to break it down for us slower guys.

I rode the short bus and I am not familiar with all the jargon and lingo so if someone can break it down into a FOR DUMMIES format it will be appreciated greatly.

 

Piece of Cake!..lol

I just tried my first spreadsheet for the week 7 game between sdo at kcy.  I tried pasting from statfox but it gives you too much info.  I have to do the calculations one by one for each game.  Then again, I am in the early stages of spreadsheets...lol  Heres what I did so far...

 

 

	57.6	93.4	-35.8	284	214.1	69.9	13.8	16.1	-2.3
	130.3	120.3	10	249.5	225.5	24	15.8	15.6	0.2
187.9	213.7	-25.8	533.5	439.6	93.9	29.6	31.7	-2.1
	94.0		-12.9	266.8	46.95	14.8	-1.05
	394.8	13.75		22.88	40.68	1.12	444.05	12.03	36.9
	17.8	20.34	0.88	35.4

 

I cant really tell what is going on with that formatting but I have

KC

wow I cant get the formating either...but I have SD-38 KC-13

I got the same thing ncsutodd.

I like the system. The methodolgy seems sound. I think what I'm going to do is bet on games where the projected spread is significantly different than the actual spread. Last week's picks where the variance was 6 or more points went 6-2. So, based on that, I'm taking:

SD -5 1/2

NE -15 (might buy up to -14)

SF +3

Atl +4

Ari +7 (unless weather is horrendous)

Phi -7

The only bad thing is that these are all road teams. But, looking over Week 7, there are lots of bad teams at home this week - Car, KC, STL, Cle, TB, Oak, Was. So, I guess it just works out that way.

By the way, the college games that are looking good to me are:

Arkansas +6

Auburn +7.5

Oregon St. +21

Again, all road teams. But, these all had variances of at least 13 points without HFA added.

Two more that are actually home teams:

Purdue -10

Mississippi St. +23

5 point variances without HFA. With, it likely creeps into double digits.

NFL Week 7 (Original system with statfox stats. 9-5-0 last week)

San Diego 37
Kansas City 17

Indianapolis 30
St Louis 8

Chicago 20
Cincinnati 19

Green Bay 20
Cleveland 16

Minnesota 29
Pittsburgh 21

New England 53
Tampa Bay 8

San Francisco 34
Houston 18

NY Jets 20
Oakland 8

Buffalo 23
Carolina 13

New Orleans 30
Miami 26

Atlanta 32
Dallas 13

Arizona 35
NY Giants 16

Philadelphia 26
Washington 9

Let's see how it does this weekend.

VG,
The new england is slightly over exaggerated because of the snow game with Tenn 59-0.

The prediction I got is 42-10

systems don't take into consideration how the game was played, and in what condition.

Only final stats.

Regardless, system saying Patriots cover, and that's all that matter.

I wouldn't put too much weight just because it predicted a "blowout".

For me, it's just a cover.

By the way, I don't know how you got a 42-10 prediction. If you are using the same formula as the original post, we should ALL get the same results.

This is the breakdown in that game.

New England

OFFENSE
RUSH: 116.30 rush/gm, Oppnts defense: 113.60, DIFF: +2.70
PASS: 288.30 pass/gm, Oppnts defense: 222.40, DIFF: +65.90

DEFENSE
RUSH: 112.80 rush/gm, Oppnts offense: 129.90, DIFF: -17.10
PASS: 182.80 rush/gm, Oppnts offense: 205.60, DIFF: -22.80

YPPT
New England: 14.90 yppt,Oppnts defense: 16.80, DIFF: -1.90
New England's Defense: 19.50, Oppnts offense: 16.30, DIFF: +3.20


Tampa Bay

OFFENSE
RUSH: 99.50 rush/gm, Oppnts defense: 128.30, DIFF: -28.80
PASS: 178.20 pass/gm, Oppnts defense: 179.00, DIFF: -0.80

DEFENSE
RUSH: 171.70 rush/gm, Oppnts offense: 126.90, DIFF: +44.80
PASS: 198.50 rush/gm, Oppnts offense: 214.60, DIFF: -16.10

YPPT
Tampa Bay: 18.70 yppt, Oppnts defense: 15.10, DIFF: +3.60
Tampa Bay's Defense: 13.20, Oppnts offense: 16.10, DIFF: -2.90

Sagarin Ratings (SOS):

New England: 22.31
Tampa Bay: 15.70

(Formula 22.31 + 15.70 = 38.01 / 2 = 19.01)
New England's: 1.174 (<--- 22.31/19.01)
Tampa Bay: 0.826 (<---15.70/19.01)

NEW ENGLAND'S ADJUSTED RUSH = 196.76

116.30 + (+44.80) = 161.10
171.70 + (+2.70) = 174.10

161.10 + 174.10 = 335.10 / 2 = 167.60

167.60 x 1.174 = 196.76


NEW ENGLAND'S ADJUSTED PASS = 314.98

288.30 + (-16.10) = 272.20
198.50 + (+65.90) = 264.40

272.20 + 264.40 = 536.60 / 2 = 268.30

268.30 x 1.174 = 314.98




TAMPA BAY'S ADJUSTED RUSH = 68.72

99.50 + (-17.10) = 82.40
112.80 + (-28.80) = 84.00

82.40 + 84.00 = 166.40 / 2 = 83.20

83.20 x 0.826 = 68.72


TAMPA BAY'S ADJUSTED PASS = 139.35

178.20 + (-22.80) = 155.40
182.80 + (-0.80) = 182.00

155.40 + 182.00 = 337.40 / 2 = 168.70

168.70 x 0.826 = 139.35


Do same formula with (YPPT)



New England: 511.74 adjusted yards / 9.62 projected YPPT = 53.2

Tampa Bay: 208.07 adjusted yards / 26.41 projected YPPT = 7.9

FINAL SCORE

NE 53
TB 8

(With taking average of "double nagatives" or "double positives", the prediction is 48-9)


(If there are two "negatives" or "positives", I don't take the average of the two. It seems logical if a team rushes 30 yards less (-30) than average on offense, and the opponent they are facing are allowing 30 yards less (-30) on defense, you would just average this (-30 + -30 = -60 / 2 = and get -30 as result.)

I don't do that. Last week I checked both predictions, and the difference were not that big. A few points here and there.

Hi guys, been on covers for a while and decided to get post just now. But regarding this system, I tried to decipher the formulas just by reading this thread and I attempted to create a excel spreadsheet with formulas and everything.

After seeing the numbers VGPOP got for week 7, I noticed that my numbers are slightly off.  For example, I got San Diego 34 Kansas City 13 or San Fran 20 Texas 30.

I don't know where I went wrong. Is it possible for one of you that have the excel functions plugged in already take a look at my spreadsheet or let me take a look where I went wrong.

Thanks a lot.

Let me correct myself and give breakdown in this case people I know people will make a mistake here.

NEW ENGLAND PROJECTED YPPT = 9.62

14.90 + (-2.90) = 12.00
13.20 + (-1.90) = 11.30

12.00 + 11.30 = 23.30 / 2 = 11.65

11.65 x [0.826 (Tampa Bay's SOS number)] = 9.62



TAMPA BAY PROJECTED YPPT = 26.41

18.70 + (+3.20) = 21.90
19.50 + (+3.60) = 23.10

21.90 + 23.10 = 45.00 / 2 = 22.50

22.50 x [1.174 (New England's SOS number)] = 26.41

wow thanks a lot vgpop.

but here is a question. Would it be too big of a problem if I were just to round all the yardage to whole numbers? I would still be keeping the strength of schedule numbers the same.

I just feel like .2 of a yard here or there won't make that much of a difference.

I've been thinking myself to take average of two negatives/positives in YPPT calculations. (instead of adding/subtracting them)

These are the most crucial part in this system.

The system can project a team to get 500 yards on offense, but if the YPPT is 30, team would only get a 16.67 projected score (500 / 16).

But if system projects a team to get 250 yards on offense, and the YPPT is 10, that's 25 points (250 / 10)

YPPT calculation is the single most important aspect here in this system, IMO.

VGPOP what's your system's record from week 1 to 6?