Welcome to Hogville!      Do Not Sell My Personal Information

New Formula for predicting point spreads

Started by arkfanchip, December 29, 2007, 04:03:18 pm

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

winbe

Quote from: arkfanchip on December 31, 2007, 10:05:09 am
I got your PM so I'm going to post my entire Texas Tech/VA formula here for everyone to look at. 

Texas Tech Offense
Run                   
17-53
7-10
17-73
12-39
Total 53-175
                                                       Texas Tech Run/Pass Ratio is .18/.82
Pass
72-420
48-466
59-490
63-431
Total 242-1807

Texas Tech Defense
Run
34-106
62-283
35-91
44-217
Total 175-697
                                                   Opp. Run/Pass Ratio is .54/.46
Pass
45-302
30-266
47-191
27-132
Total 149-891

Virginia Offense
Run
36-97
39-130
25-82
36-94
Total 136-403
                                                  Virginia Run/Pass Ratio is .48/.52
Pass
28-144
25-288
43-225
51-275
Total 147-932

Vir. Defense
Run
46-131
39-95
50-140
30-84
Total 165-450
                                              Va. Opp. Run/Pass Ratio is .58/.42
Pass
26-299
21-145
26-175
46-347
Total 119-966

Texas Tech Formula is
Run:  175 + 450 = 625 * 1.5= 937* .18= 169*.58= 98
Pass:  1807+ 966= 2773 * .82=2274* .42= 955

955+98= 1053/4 263

Virginia Formula
Run: 403+697*1.5*.48*.54= 428
Pass: 932+891*.52*.46= 436

428+436=864/4=216

Texas Tech 263= 37
Virginia      216= 20


Actually, I am using the Off/Def %.  I ended getting the numbers you got for TT/VA, but using the same spreadsheet with the AR/MO game I got totally different results that had AR winning big.  I put my numbers on here so you could see.  I have rounded what I typed below just for ease to follow.

I get AR with 323 and MO with 249.

AR Rush
1824 * 1.5 = 2736 * .7175 = 1963 * .5472 = 1074

AR Pass
1699 * 1.0 = 1699 * .2825 =  480 * .4528 = 217

AR Total
(1074 + 217) / 4 = 323 (Off of your scale)

MO Rush
1063 * 1.5 = 1595 * .4923 =  785 * .4651 = 365

MO Pass
2316 * 1.0 = 2316 * .5077 = 1177 * .5349 = 629

MO Total
(365 + 629) / 4 = 249 (app 32 POINTS)

Where do we differ?

arkfanchip

Not sure???  I'm sure I probably added wrong or something.  Are you using the LSU game?  I'm not betting on Arkansas so I really didn't pay that much attention to the game.  I'm still pissed at Clemson for giving me a loss last night. 

 

arkfanchip

Ok I just did the last four games for each team except I didn't use LSU b/c of the overtimes.

I made a huge mistake the score I posted was what the LSU/Arkansas score was when I testing the formula at the very beginning.  So 41-31 is what the score should have been between Ark/LSU. 

I have Arkansas 42
Mizzou              33

I just redid it twice so I hope it's right. 

arkfanchip

My scores for the day along with my scores from the past few days.

Wake Forest -3 24         Actual Score   24
UCONN             12                              10

Penn State -5.5  24                             24
Texas AM           19                             17

Alabama -4        32                             30
Colorado            16                             24

Ok. State -5       49                             49
Indiana              31                             33

Clemson -2.5      24                             20
Auburn               10                             23

Boston Col.         23                             24
Mich State          20                             21

Orgeon St.          24                             21
Maryland            16                             14


My scores for today
Arkansas +4  42       (Not betting on Arkansas just a rule I have)
Mizzou          33


Florida -11     41   (According to the formula this one should be BLOWOUT city)
Michigan        10 


Texas Tech -5.5  37       (I'm anxious to see how the huge passing numbers skew the formula)
Virginia              20


Wisconsin +2.5   30         (This one is like the Clemson game last night...my heart tells me no but the formlua says yes) 
Tennesee            14



Feralhog

Quote from: arkfanchip on January 01, 2008, 09:17:25 am
Not sure???  I'm sure I probably added wrong or something.  Are you using the LSU game?  I'm not betting on Arkansas so I really didn't pay that much attention to the game.  I'm still pissed at Clemson for giving me a loss last night. 
I've got the spread sheet working if you want to give me the raw data for Ark and Mizzou
Seer, Sage, Soothsayer and former Computer repairman for Hunter Biden......Feralhog the Magnificent

arkfanchip

Arkansas (First number is attempts second is yards.)
Run
54-293
58-541
36-127
45-206

Pass
18-144
11-109
29-136
18-223

Run Defense
34-162
32-129
43-157
31-80

Pass Defense
34-182
51-364
25-128
51-421

Mizzou
Run
32-98
43-151
36-143
49-203

Pass
39-219
50-368
41-284
35-352

Run Defense
40-166
22-42
32-150
39-133

Pass Defense
26-209
49-349
41-256
28-247

Feralhog

January 01, 2008, 11:00:39 am #106 Last Edit: January 01, 2008, 11:02:22 am by Feralhog
Ark - 275.6167  = 44
Mizzou 203.3919 = 10

Obviously there's a mistake in one of cell formulas on the Mizzou's side of the ledger. 
Seer, Sage, Soothsayer and former Computer repairman for Hunter Biden......Feralhog the Magnificent

arkfanchip


Hoofer

so how do the remaining games look? anybody run the numbers? rough day today but that was kind of expected after the 7-0 run.

arkfanchip

Yea it was a bad day....but I still have faith in the formula.  Maybe it will work better when teams have only a week to prepare for their opponents instead of a month.  I don't know???  I'm going to do the NFL playoffs this weekend and see how it works. 


arkfanchip

The formula said WVU would win but after yesterday I was embarrased to post it....

arkfanchip

Has anyone done the NFL games this weekend yet?  I'm going to do it tom. but I'm not using week 17. 

 

Singha Hog

This is pretty darn cool.  I do not gamble at all but I am absolutely fascinated with mathematical formulas and software programming (yes, I am a serious geek).  Combine that with my love of college football and I think I have found something to make the off season go by a bit better.

One thing I think I would change is the "multiply the rushing number by 1.5".  That may be fine in most cases, but when you are looking at a team like Texas Tech or Hawaii that rely much more on the pass than the run this might not be accurate.  I'm not saying switch it to multiplying their passing yards by 1.5; but that might be a place to tweak it a bit.  Maybe you can use a bit of a sliding scale for a multiplier value based on that team's preference to pass or rush.

rzrbacknTX

Could someone who has FeralHog's spreadsheet for this send it to me?

JHGiunta

I would also like this EXCEL spreadsheet, if someone has it handy.

josephgiunta@centurytel.net