[HeeluvaGuy]

Likely my last play of the day:

7:30 Marshall/Pittsburgh o63 (-105)
This number almost seems too good to be true. Marshall and Pitt are averaging 42.7 (T19) and 36 (#48) points per game this season, respectively. (Yes, Marshall racked up 62 points against Morgan St, so that average skews a little high.) Offensively, the Thundering Herd is averaging 10.1 yards per point (#7), and the Panthers are averaging 11.4 YPPT (T24). On the defensive side, the averages are 11.6 YPPT allowed for Marshall (#116), and 13.2 (T80) for Pitt. In other words, we have two high powered offenses going against two generous defenses. From a matchup perspective, Marshall looks to get it done through the air. Reports are that QB Chase Litton will be back for tonight’s game. He had 722 yards and 2 TDs in Marshall’s first two games before being injured in the 4^th quarter against Akron (concussion). Litton will face a Pitt defense that ranks next to last in the nation in passing yards allowed (361/game). The Pitt secondary has been susceptible to the big play – 4 UNC receivers had catches of 20+ yards last week – and Marshall has 9 receivers with catches of 20+ yards this season. When Pitt has the ball, the Panthers will likely look to run, thought last week Nate Peterman showed that he is capable of throwing the ball (and WR Quadree Henderson leads the nation in all-purpose yards). Marshall’s defense ranks 144^th in yards allowed, so whatever path the Panthers choose they are likely to encounter very little resistance. I have this game getting into the 90s, but we don’t need anything close to that. Others say: PW 73, PP 69, Steele 70, GS 57, Massey 69.

Good luck everyone!

[HeeluvaGuy]

I'm having an awful two weeks. 0-3 so far today. Apologies to anyone who tailed.

[HeeluvaGuy]

I have taken some time off from posting picks. Things just haven't been right this season. However, I thought I would post my model-based predictions now that I have some actual results to share. All lines are based off of he home team such that a "-" means the home team is favored. If the delta column is a negative number, that means my mode likes the home team ATS. If the number is positive, it likes the road team. The totals should be self-explanatory. If not, let me know. So, for example, my model likes Miami to cover tonight and the game to go under. The second link is to the results based on some splits I have looked at. Best of luck with your plays.

https://docs.google.com/spreadsheets...QY8RxjSgf8f0CY
https://docs.google.com/spreadsheets...BxJCQTbJ7fQgXc

[thekoreanmang]

Thanks for sharing this. Your insights are always appreciated.

If the delta is higher then the play is stronger? For example, BYU would be a much stronger play than Miami tonight?

[HeeluvaGuy]

Thanks for sharing this. Your insights are always appreciated.

If the delta is higher then the play is stronger? For example, BYU would be a much stronger play than Miami tonight?

In theory, yes. HOWEVER, my model does not take injuries, rivalries, weather, etc. into account. Where there is a large delta, there is often something the model is not accounting for. So you still need to do some due diligence.

My methodology is opponent power ratings based using three different ratings. I generate three lines from each set of ratings based on whole season performance, home/away performance, and recent performance. So I have 9 lines for each game. I then weigh those lines to come up with a final prediction for the games.

If you look at the results link, you'll see that the model has overvalued favorites this year, but has done nicely with underdogs (especially home dogs).

Oh, and I never bet on BYU. Just don't trust them.

[thekoreanmang]

I trusted in the Mormons last night. Thanks, HG.

[HeeluvaGuy]

I trusted in the Mormons last night. Thanks, HG.

Glad you picked a winner.

[HeeluvaGuy]

Updated projections for Week 9: https://docs.google.com/spreadsheets...QY8RxjSgf8f0CY
Updated results tracker: https://docs.google.com/spreadsheets...BxJCQTbJ7fQgXc

I typically update lines every night and then Saturday morning, but I didn't have time last night. So there may be some stale lines, and I haven't filled in the totals for all games yet. But the projections are all there.

I'll post later about the splits that are doing the best, and what those plays are this week.

[HeeluvaGuy]

I was able to update everything last night. If you take a look at my seasons results, you'll see that games where my model favors home underdogs ATS are 15-5 since Week 5. (Week 5 is the first week I use my model because I need at least 4 weeks' worth of games for it to work.) It's not only the tracker, but those games are also 10-10 ATS with an average line of +202.9.

The home dogs that qualify right now are:

(Fri) Utah St +6
(Fri) Fresno St +13.5
Iowa St +6.5 (would not qualify at +6)
Texas +3.5
South Carolina +13.5 (would not qualify at +12.5)

I have personally been waiting until game day to play these, but that's mainly to see if the lines move so much as to disqualify the play. The easiest way to identify these by looking at the model is to look for a negative number in the "Delta (from current)" column for the sides and a positive number in the "Current Line" column.

The model is also over 60% in games where it favors the under, and the difference from the current line is between 1 and 10 points.

Best of luck to everyone this weekend!

[doubledime]

I just want to chime in here.....

Since the beginning of the season, HG and I have been PMing each week and I can verify, this is the real deal and his win/loss numbers are spot on. Good luck to all!!

DD

[HeeluvaGuy]

Thanks DD.

I updated the current lines this morning. The link is in post #43, and will be the same link for the remainder of the season. You can use the tabs at the bottom to take you to the current week.

Also, I do my best to keep the results current during the weekend when I find myself in front of a computer.

Good luck this weekend everyone!

[doubledime]

Thanks DD.

I updated the current lines this morning. The link is in post #43, and will be the same link for the remainder of the season. You can use the tabs at the bottom to take you to the current week.

Also, I do my best to keep the results current during the weekend when I find myself in front of a computer.

Good luck this weekend everyone!

Thanks for updating.

[HeeluvaGuy]

Saturday morning update: https://docs.google.com/spreadsheets...it?usp=sharing

The only two sides that qualify as home underdog plays right now are:

Texas +4
South Carolina +14.5

Best of luck with your action today!

[HeeluvaGuy]

Oops. Double post.

[HeeluvaGuy]

Home dogs went 3-1 this week with 2 SU winners (Texas +157 and S. Carolina +700). Iowa St covered too, but wasn't an official model play because of the line move to +6. Wyoming was a play at +15, but I think it closed at 14.5. Regardless, I won't count those when I do my updates over the next few days.

[doubledime]

Home dogs went 3-1 this week with 2 SU winners (Texas +157 and S. Carolina +700). Iowa St covered too, but wasn't an official model play because of the line move to +6. Wyoming was a play at +15, but I think it closed at 14.5. Regardless, I won't count those when I do my updates over the next few days.

Sweet!!!

[doubledime]

HG

Thanks for sharing your program with the forum. Very impressive results. If anyone clicks your link in post #43 they will see a lot more "green" than "red" to the tune of 57-39 (quick look this morning some games not yet colored in). I think your on to something here!!!!

Thanks again, DD

[HeeluvaGuy]

Thanks DD. I just finished updating everything.

Here is the link to Weeks 5-9: https://docs.google.com/spreadsheets...it?usp=sharing
Here is a link to the results (with splits): https://docs.google.com/spreadsheets...it?usp=sharing

Through 531 plays (sides and totals), the overall results are 54.2%, but using the splits it should be pretty easy to see where the model is outperforming this season (hint, hint, underdogs).

I should have Week 10 numbers posted Tuesday night.

[doubledime]

Just to be clear, home dogs at the opening line are 71% and at the closing line are 75%, correct?

[HeeluvaGuy]

Just to be clear, home dogs at the opening line are 71% and at the closing line are 75%, correct?

Sorry, no. Home dogs with any delta (as little as a .1 point favorite according to my model) are 71%. Where my model likes them by more than 1 point, they are 75%. All ATS.

I do my best to track these using closing lines, but like yesterday I got busy and wasn't able to see what happened with the Wyoming game. In that case I go with my most recent update. Not perfect, but I doubt there are more than a few games that would have been affected.

[doubledime]

Sorry, no. Home dogs with any delta (as little as a .1 point favorite according to my model) are 71%. Where my model likes them by more than 1 point, they are 75%. All ATS.

I do my best to track these using closing lines, but like yesterday I got busy and wasn't able to see what happened with the Wyoming game. In that case I go with my most recent update. Not perfect, but I doubt there are more than a few games that would have been affected.

Just like I like it, not too many games with a high win percentage.

[freeVICK]

Awesome work HG, keep it up bro!

[HeeluvaGuy]

Week 10 here: https://docs.google.com/spreadsheets...QY8RxjSgf8f0CY

NOTE: I was a little tired yesterday doing these, so I plan to double check them later this week. Most of what I do is automated now, so I don't expect any changes. But I'll post if there are any.

[doubledime]

Thank you

[HeeluvaGuy]

Thank you

I hope you're thanking me on Sunday.

[doubledime]

HG if you can are you going to post the plays, since some may not quite understand how to read your numbers?

[HeeluvaGuy]

Yeah, of course. Just waiting to see if there are any big line moves. First play looks like Thursday night. I'll post tomorrow night with explanations.

[HeeluvaGuy]

The home underdog plays (as of right now) are:

(Thurs) Iowa St. +21 (good to 19.5)
Kentucky +2.5
UL-Monroe +13.5 (no lower)
North Texas +20.5
SMU +3
Northwestern +7

Past results don't guarantee future success, but these home underdogs have gone 18-6 ATS (75%) since Week 5, which is the first week my model "went live" so to speak. These plays have also gone .500 SU at an average of +240.5.

Here's a link to my model numbers: https://docs.google.com/spreadsheets...it?usp=sharing
Here's a link to a tracker showing some splits: https://docs.google.com/spreadsheets...it?usp=sharing

For tips on how to read the model, see post #38. If that doesn't help, feel free to ask.

NOTE: My model does NOT take injuries, weather, etc. into account.

[HeeluvaGuy]

Just realized Georgia St +4 also qualifies.

[HeeluvaGuy]

Barring some late line movement, it looks like there will be two home dog plays tonight:

Georgia St +5.5 (qualifies as long as Ga St. is a dog, so get the best line you can)
Iowa St +20.5 (qualifies down to 19.5, but no lower)

Best of luck tonight and this weekend!

[HeeluvaGuy]

1-1 last night on home dogs. None qualify tonight. Looks like we'll have several on Saturday if the lines hold pretty close to where they are.

[Renegades]

That Ga St team is painful to watch. The qb looked like one of those tear away referee dolls with the flimsy arms. GL tmrw

[HeeluvaGuy]

That Ga St team is painful to watch. The qb looked like one of those tear away referee dolls with the flimsy arms. GL tmrw

They really are bad. The blocked FG for a TD was the worst part.

Thanks for checking in, and good luck to you as well.

[Renegades]

I didnt even get why they were trying a 50 yard fg in the first place. Odds of him making that were very low

[Codyodom]

I didnt even get why they were trying a 50 yard fg in the first place. Odds of him making that were very low

Yup..messed up my 1st half GS +3 bet big time