I know what you're thinking-

'Finally! Joe will produce his charts and graphs and pretty things that tell me what's going to happen! It's about time!'

Except I'm not.

Statistical models are based on the fundamental assumption that this year will be more or less like the last few years, and that we can draw conclusions from yesterday about what will happen today. That's wrong right now, in ways both big and small.

We don't know which players will take the field on Saturday for either team. COVID and its associated quarantines could strike any player at any moment and take them out of the game. It's likely to strike position groups in swaths, as these players tend to spend a lot of time together. Even though the game is being played, we don't have much visibility into if it's being played with a second string offensive line or secondary on one side or the other.

We don't know how these teams have been prepared or coached. Every step of the process could potentially be disrupted by COVID, from what drills they do in practice to what they eat to how well a player sleeps or focuses. We don't know which coaches or teams will get hit the worst, because we don't know that much about how they practice or prepare.

We don't know how these restrictions and challenges will impact different teams. Some coaches could take COVID protocols very seriously, others could blow them off. Some counties or states could have different protocols that impact teams differently. We don't know which ones, because we don't have a good grasp of what the local restrictions are or how they're being enforced.

We're seeing that play out across the country, as teams lose to opponents they have no business losing to and scores are wildly off from what we'd normally predict.

In light of the chaos we've seen and continue to expect to see across college football, the sense of security provided by analytical models is more false than usual. I'm not going to use my typical projections. You might as well ask me to predict who would win if the football teams had a pickup basketball game. It's not just a different year- it's basically a different sport.

Still! You are here for a pick and a pick I shall make. We should have something like full rosters for Saturday's game. We have enough data to make a prediction, but without the level of certainty we can usually feel. It's especially weird because the Trojans have given us two games of data, but it's unclear how reliable that is because we know as little about their opponents as anything else.

USC is favored by a couple of points today, and it's easy to see why prognosticators have more faith in the Trojans- they've actually watched them play. They have performed decently well against the two Arizona schools, showcasing an explosive offense that has nevertheless lacked efficiency (19 scrimmage plays of 10+ yards per game, good for 11th nationally). If Utah can contain their big plays, they should be able to go toe to toe.

Utah's secondary is untested, and we know more about what they've lost than what they have- two NFL DROY candidates in Jaylon Johnson (Bears) and Julian Blackmon (Colts). Those are going to be big shoes to fill, and that doesn't bode terribly well for Utah's secondary in their first live game in almost a year.

Sit back, relax, and enjoy the game. Don't hold these players to your usual expectations, and don't be surprised if Utah looks like a mess. Under normal circumstances, I'd pick Utah to win this game by 2. However, with cancellations and COVID I'm less confident than usual and picking the Trojans.

USC 31, Utah 28