Model Plays: Week Ten
Let's take a look at the (CFBDm) Model and how it performed in week nine. Then, we'll run these games through the simulation machine to see which sides it favors for a loaded week ten.
To be completely transparent in this (which is the goal from the start), I have been waiting for the Model to take a beating. One really bad week to knock us back. While the UNIT plays have been holding even recently after a hot start, we continued that trend in week nine. But, the “spray the board” method would have been a big success for us. In week nine, the Model went:
61-44-1 (+1 UNIT) = 57.5%
Here is the weekly further breakdown:
The ‘OVER’ lean in week nine was the biggest boon, going 19-10 in those overall (29 plays), compared to 11-13 on the 25 UNDERS we played. It still only resulted in a +1 UNIT advantage on the suggested plays. It went 16-games over the .500 mark on P4 vs. P4 matchups and 25-22-1 in G5 vs. G5 tilts.
Here are the full results rolled up or the 2025 season:
It’s nice to be back on the top of the roller coaster ride with a 57.5% week, but we are cautious of getting too overconfident in the performance, considering the way things have gone this season (and, last). It’s bound to have an off week, but that’s four straight weeks at or above .500 and now we are sitting at a solid 53.4% win rate for the season. That should not go very far in any direction at this point with 928 results included in that number.
WEEK TEN
We need to first take a look at how we are doing early on in the week with six games already in tow. Depth+ subs got early access to the simulation results, which had three ‘suggested’ UNIT plays, all single plays on three totals. It’s not going all that well out of the gate here in week ten, going 1-2 in those plays and 4-8-0 in everything thus far.
We have one play in tonight’s (Friday) three-game slate:
Memphis (-14) at Rice - 2 UNITS
The model continues to love the Tigers, and it came through for us last week in their comeback victory (and cover, of course) against the mighty South Florida Bulls. This time, they’ll need to avoid a letdown on a short week, having salvaged their season just a week ago with an effort that was sure to take a lot out of them. Rice has not been good, but they still run a quirky offense that could cause some issues and keep things close. However, the Tigers have a LOT to play for and should eventually come through, and that talent will overwhelm the undermanned Owls. Looking at the Game Card, the most significant advantage is the explosive nature of the Memphis offense (ranked 22nd) compared to the Owls’ inability to limit such plays (ranked 124th). That means, quick scores — so even if things appear close and this makes us uneasy early on, the model sees the much more talented team pulling away by more than two scores.
SATURDAY…
Texas (-2.5) vs. Vanderbilt
One of the features of the model is what we call “quirks” based on specific social factors that apply an adjustment that data alone cannot pick up. One of those is a “gauntlet” effect that could happen after big wins or a series of them. That’s precisely what the Commodores are going through right now. They have played a string of close, hard-fought battles, and it will eventually catch up to them. Add in the fact that no matter how ‘off the rails’ this season has gone for the Longhorns, they still have a LOT of talent. Even if Arch Manning is unable to play, the Model still likes them to pull away and win this by more than a field goal in Austin. We saw what a good defense can do to Diego Pavia last week when Mizzou brought the wood and made him look human. Texas has that same ability. Here is what the model simulation came up with even having Manning at 50/50 chance to play:
Illinois vs. Rutgers OVER 61.5
What happened to this Rutgers offense over the past few weeks? Well, they played better defenses for one (Washington and Oregon), and they have fallen so far behind in games that they were unable to have any semblance of balance to their attack. Illinois has proven that they are not a very good defense. Their latest escapade was allowing 42 points to Washington in Seattle last week. Those injuries have piled up, and they can’t seem to get their footing. This could be just the matchup to get Athan Kaliakmanis back on track. It will definitely help to have his favorite target back, as Ian Strong remains very questionable to play. But that’s just a bonus. Throw in a truly horrific Rutgers defensive unit — especially against the run — and this has all the makings of a back-and-forth affair that lights up the scoreboard in Champaign.
UConn (-11.5) vs. UAB
Every Model has a favorite. This Model loves the Huskies. They have three losses, and all three have been in overtime. UAB is off a bye, and the last time we saw them, they were beating that Memphis team we drooled over earlier in this article. Alex Mortensen either found a magic potion, or this team is actually a lot better than they played under Trent Dilfer. The Model will always favor the larger sample size, and that has the Huskies as a much better team than the Blazers. Throw in that it will be at home, likely with some windy conditions, and UConn’s superiority on the line of scrimmage offensively, and we have ourselves a blowout potential in Storrs. UAB is also quite banged up, and even if they get some of these players back, it’s unlikely to be enough to keep this one close.
Added Saturday morning…after a 6-0 Friday evening. The Model is really feeling itself right now. Can it carry over?
Air Force / Army OVER 49.5
The model is definitely leaning more to the overs these days and it seems to be for good reason and these are not your typical service academy offenses. Both have modernized in some fashion and have had success. The primary reason though that the simulation gets to 58 points is that Air Force defense. It’s been really bad this season and we don’t see that changing just because of one bye week. Look at their points allowed this season against FCS-level opponents: 49, 49, 44, 34, 51 and 21 (week eight against Wyoming). I mean, three of these are right there for the full-game over alone! It does also see the game staying close and sets up one of those fun back-and-forths. Weather should not be an issue at last check, so…
Buffalo (+2.5) at Bowling Green - 2 UNITS
The love for the Bulls remains despite so many letdowns this season. We’ve said it many times here that this computer simulation doesn’t have feelings. That’s one of it’s best features. You see, if you keep getting burned by a certain team, you tend to avoid them. This Model says, “wait a minute, there is just too much talent here to jump off now.” It doesn’t actually talk either, but you get the point. And Bowling Green has been really difficult to figure out. They beat Toledo and then lost to Central Michigan and KENT STATE. So, who do we have here? They do seem to play better at home, but there are some key injuries piling up for the Falcons, including at the all-important quarterback position as their top three are dealing with some sort of injury: Drew Pyne (out), Lucian Anderson III (questionable) and Baron May (out). Leaving an unproven redshirt freshman named Hunter Najm as their likely starter.
Louisville / Virginia Tech UNDER 53.5
Our first (and only) suggested under play comes in the form of Louisville at Virginia Tech. I’m not entirely sure where this comes from but when we look at the ‘Game Card’ it may just be the pace driving this one. Or, maybe it is that both teams ave some impact offensive players on the injury report. Ah, I see. It’s that underrated Louisville defense led by Ron English. The Model just doesn’t see a way for Virginia Tech to hold up their end and then Louisville leans on that run game to close things out. If that Hokies defense can just limit the big plays and keep things in front of them — tackling effectively, of course — then this one drags out and stays well under this number.
Other UNIT suggested plays:
Louisiana (+4.5) at South Alabama | Game Card >
California / Virginia OVER 52.5 | Game Card >
San Diego State (-10.5) vs. Wyoming | Game Card >
Here is the full “Card” as it stands. Note that these are moving slightly as we get availability reports and make playing time adjustments all the way up to kickoff.
Thank you for your support of our efforts and have a GREAT week ten!