P2Y nailed the picks in week 3. There’s almost literally nowhere to go but down from here…
WEEK 3 RECAP
Last week was the first time I used points squared per yard (P2PY) to try to predict game score and did surprisingly well. The picks along with the spreads are summarized below (ATS picks in bold italics). P2PY went 7-2-1 against the spread and 10-0 picking winners. This likely is not a sustainable win percentage but still a good sign going forward.
| Away | Home | Away Predicted Score (ACTUAL) | Home Predicted Score (ACTUAL) |
| Houston | Cincinnati (+7) | 26 (48) | 20 (16) |
| Florida St. | Louisville (+2.5) | 29 (20) | 38 (63) |
| Miami (FL) | Appalachian St. (+3) | 29 (45) | 14 (10) |
| Alabama | Mississippi (+9.5) | 39 (48) | 21 (43) |
| Oregon | Nebraska (-3) | 27 (32) | 37 (35) |
| Texas A&M | Auburn (-4) | 28 (29) | 22 (16) |
| Georgia | Missouri (+6.5) | 26 (28) | 22 (27) |
| Ohio St. | Oklahoma (+2.5) | 42 (45) | 22 (24) |
| Temple | Penn State (-9) | 20 (27) | 25 (34) |
| Baylor | Rice (+30) | 36 (38) | 11 (10) |
The Good
- Correctly picking all 10 winners (I forget if I mentioned this above).
- Correctly picking five underdogs to beat the number (Louisville, Texas A&M, Missouri, Temple, Rice.
- Picking the two underdogs to win outright (Louisville and Texas A&M).
The Bad
- Missing on Houston/Cincinnati and Alabama/Mississippi. I did say that I liked Houston better than the prediction would indicate but this is still a loss. I believed that the Alabama prediction was solid. This looked very bad early on, then good, then bad again…
Summary
Week 3 was a good first pass for the use of P2PY to pick games, but I also watched almost every one of these games and in a lot of cases, college football is fickle. If Oregon kicked extra points, they may have beaten Nebraska. If Penn State doesn’t turn the ball over at terrible times, they roll Temple. Those couple of things going in my favor make the result a little better than maybe it would’ve been otherwise. On the flip side, if Alabama secured an onside kick, they would’ve covered so I guess it goes both ways.
Season Long Results: ATS 7-2-1 (75%); Straight 10-0 (100%)
On to week 4…
Week 4 Picks
Method Notes
- Predictions are based on the same formula as demonstrated for week 3 using offensive P2PY and defensive yards allowed.
- Updated conference strengths-of-schedule rating from the same source as last week and P2PY is adjusted accordingly. Yards allowed are currently unadjusted.
- I am using a team’s average adjusted P2PY with an equal weighting for all previous games. Again, eventually I will start to weight early games less but I’m not comfortable to do this just yet.
- Game lines are current as of Monday (9/19).
The Picks
| Away | Home (Line) | Away Prediction | Home Prediction | ATS Pick | |
| Kent St. | Alabama (-43) | 17.8 | 35.3 | Kent St. | |
| Washington | Arizona (+9.5) | 40.1 | 14.5 | Washington | |
| Nebraska | Northwestern (+7.5) | 34.4 | 11.4 | Nebraska | |
| Wisconsin | Michigan St. (-5.5) | 21.2 | 13.5 | Wisconsin | |
| Georgia | Mississippi (-7) | 25.9 | 22.9 | Georgia | |
| Florida | Tennessee (-6.5) | 24.5 | 22.2 | Florida | |
| Oklahoma St. | Baylor (-9.5) | 24.5 | 27.0 | Oklahoma St. | |
| USC | Utah (-3) | 18.0 | 15.2 | USC | |
| Stanford | UCLA (+3) | 24.2 | 17.2 | Stanford | |
| Penn St. | Michigan (-18.5) | 27.2 | 42.7 | Penn St. | |
| Arkansas | Texas A&M (-6) | 25.5 | 30.4 | Arkansas | |
| LSU | Auburn (+3.5) | 20.4 | 20.5 | Auburn | |
| Louisville | Marshall (+25.5) | 41.3 | 20.0 | Marshall | |
| Florida St. | South Florida (+6) | 29.2 | 27.6 | South Florida | |
| Clemson | Georgia Tech (+9.5) | 20.4 | 18.0 | Georgia Tech | |
| Houston | Texas St. (+34.5) | 32.5 | 28.1 | Texas St. |
Week 4 Pick Notes and NGFT
- I legitimately think Kent St. has a real shot to beat the spread. Their defense isn’t terrible.
- Wisconsin, Georgia, Florida, and USC are predicted to win outright as underdogs. I think I like all of them.
- LSU and Auburn are basically dead even. I’ve got the Tigers all day long.
- Florida St. may come back strong or they may be totally sunk.
- If Texas St. can end up within four points of Houston, a reality TV personality can end up president of the US.
Main Photo
