Predicting every major college football game in 2017

Last August, I published this blog about how I would be predicting every major college football game for last season. I'm doing the same again this year, via a new-and-hopefully-improved spreadsheet for prediction-tracking.

As I wrote this time last year, the goal of this is to predict each individual game in the 2017 season -- this year, I'm predicting every Power Five conference plus Notre Dame:

When the Associated Press released its pre-season Top 25 rankings a couple weeks back, I started to think about what process I would go through if I was an AP voter. It was interesting me to think about how I'd project the season as a whole -- not really a pre-season Top 25, but a projection of the season's final Top 25 poll. So I built a way to do this.

There weren't really any mid-major programs that interested me enough to include this season (though South Florida would have been tops on the list), but most anyone with some past experience in Excel should be able to take my spreadsheet template and expand it to include the Group of Five teams as well if they want. I will warn that putting this sheet together is quite time-intensive, though. It's a labor of love and madness for me. That's why I wound up with only Notre Dame in the "extras" column -- I felt that if I included South Florida, I immediately had to include at least a half-dozen other teams who could be as good as Charlie Strong's Bulls.

Here's the public version of this year's spreadsheet, for those who are familiar with this and want to get right to that.

If you didn't use this spreadsheet last year, here's a quick rundown of how it works. You put a "W" or "L" for each team's game in their season to project whether they'll win or lose that game. As you go, make sure that the inverse result goes in the corresponding field for that team's opponent in each game -- the spreadsheet has no way to know if you make an error by putting a "W" or "L" in both instances of the same game.

The spreadsheet automatically compiles your projected record for each team below their schedule, as well as at the bottom of the sheet beneath their entire conference results. This helps you organize your standings for each division in each conference. From there, you figure out your conference championship game participants and winners; that helps you figure out a rough Top 25 for the end of the season, and from there you should be able to project your Playoff match-ups.

As the season goes on, you write in the real result of every game on the spreadsheet with a "W" or "L" in a separate column. The "Correct?" column automatically takes the data you put into the "Result?" column to let you know if you were correct or not. (Note: This was added after the sheet was originally published, on Aug. 9, 2017.) The spreadsheet compiles each team's real-life record, as well as your record in the accuracy of your predictions for that team.

Your predictions also compile at the conference and national level further down in the spreadsheet.

Again, here's the link to this year's public spreadsheet -- follow the instructions in the first line to create your own personal copy of the spreadsheet so you can get predicting.

A couple of other notes before I talk about my picks...

  • Compared to last year, this year's sheet has a row for August 26, which ... isn't Week One ... but it kind of is Week One if you ask the schedules on major websites? Kinda weird how that works.
  • It also has a row to reflect each team's bye week. One of the most annoying parts about counting results last year was some confusion around why different teams had a different number of games filled in -- this year, every week of the season is represented.
  • One change that I thought about making: I wanted to figure out how to make the spreadsheet automatically update the inverse of your prediction if possible. So, if you picked Alabama to beat Florida State in the very first game on the sheet, the box for Florida State's game against 'Bama would automatically be updated to their corresponding "L." This would only work if you filled out the sheet from top-to-bottom, left-to-right. It also made it so that all of the boxes with this formula filled out (so, the second instance of each game, from top-to-bottom and left-to-right) would start out initially with a "W" already in that field, rather than being blank. This would give the appearance of the sheet behind half-filled-out when you first showed up.
  • For these reasons, I thought it would be too confusing to explain ... even though I wound up explaining it anyway. I'm interested in hear if anyone would like this feature added next season. Maybe I'll explore a way to have this functionality on a website rather than a Google Sheet ... though that is something that seems way over my head in terms of web-building knowledge.

Anyway ... here's where I netted out in my picks. To get things started, I marked Alabama and Ohio State to go undefeated. From there, I filled out picks for the other teams I figured would be strong contenders -- Florida State, Penn State, USC, Oklahoma, Washington, etc. Here is my full, filled-in spreadsheet, and here are some conference-by-conference notes of where I see things netting out:

Florida and Alabama meet in the SEC title game for the third straight year. Unfortunately, the Gators lose to the Tide for the third straight year as well in my predictions. I think UF will take a good step forward in Jim McElwain's third season, but I don't think they're good enough yet to beat Alabama. Ultimately, Florida finishes 11-3 after losing to LSU and FSU in the regular season plus 'Bama in the SEC title game, and beating Louisville in the Orange Bowl. We'll get to Bama's fate in a moment. As for the other SEC players, I had LSU and Arkansas at 9-3, and I had Georgia and Auburn at 8-4.

At long last, the ACC gets the Florida State vs. Miami title game they originally figured would be a regular occurrence, with the 'Noles winning. It's funny to me that Miami might finally win its first division title in the ACC this year. Truth be told, I tried my hardest to figure out why they shouldn't win their division, but the middle class of the ACC is tough to trust right now. I have N.C. State taking a good step forward at 9-3, but they're in the same division as Clemson, Florida State and Louisville. I have these guys beating up on each other a bit, with Clemson losing to Louisville and N.C. State but beating Florida State; Louisville, meanwhile, only loses at Florida State.

This winds up with Florida State in the ACC title game due to the head-to-head result of their game with Louisville; they wind up in my playoff because their only losses are to Alabama in Week One and a 10-2 Clemson in midseason. Chaos elsewhere allows this to happen; we'll get there in a moment.

I don't really trust Miami a ton in the Coastal since I've never seen whoever their quarterback will be, but I somehow wind up with them going 10-2 in the regular season because I also don't really trust Georgia Tech, Virginia Tech, Pittsburgh or North Carolina this year. The 'Canes get the nod because their defense should be nasty.

Ohio State rectifies last year's slip-up against Penn State with an undefeated B1G run, beating Wisconsin to cap things off. Wisconsin's schedule is extremely easy this year, but I have them dropping two to Iowa and Michigan near season's end. They still win the West.

The East is stacked by comparison -- Ohio State finishes 13-0 and they're in the Playoff. Meanwhile, Penn State finishes 11-1, with only a loss to Ohio State, so ... they're in the Playoff as well. We're gonna get to explaining why in a second, okay? Michigan finishes 9-3 and one of those losses is to Florida, yay!

The Pac-12 delivers the championship game we desperately wanted last season, with USC topping Washington. Here's where the chaos happens, though -- I predicted this because this is my world and in my world we can have a lot of fun. Washington gets through the regular season undefeated at 12-0, winning the North. Stanford is good too in this division, losing only to USC and Washington. In the South, USC runs away with the division despite dropping games to Washington State and Arizona State for a 10-2 regular season. I just can't see them going undefeated or losing one game right now, so we'll see how much this burns me.

By the time the Pac-12 title game rolls around, though, USC is rolling again, just like they were at the end of 2016. They beat Washington, but their two losses aren't very impressive, so they're left out of the Playoff. Washington gets compared to Penn State, which also didn't win its conference title, for the final Playoff spot. Both teams have one loss, but Washington lost to Stanford while Penn State lost to only undefeated Ohio State, so the Nittany Lions sneak in.

The Big 12 is left in the cold again, despite Oklahoma beating Oklahoma State in the return of the Big 12 title game. The Sooners get revenge for a regular-season loss to the Cowboys, but it leaves the Big 12 without a good Playoff contender. Oklahoma lost to Ohio State and Kansas State as well in the regular season, and Oklahoma State loses to Pittsburgh and Texas in addition to their conference title game loss.

The Playoff rankings look like this following conference title game week:

  1. Alabama (13-0, SEC champ)
  2. Ohio State (13-0, Big 10 champ)
  3. Florida State (11-2, ACC champ)
  4. Penn State (11-1)
  5. USC (11-2, Pac-12 champ)
  6. Louisville (11-1)
  7. Oklahoma (10-3, Big 12 champ)
  8. Washington (12-1, Pac-12 runner-up)
  9. Florida (10-3, SEC runner-up)
  10. Clemson (10-2)
  11. Wisconsin (10-3, Big 10 runner-up)
  12. Oklahoma State (10-3, Big 12 runner-up)
  13. NC State (9-3)
  14. USF (11-1)
  15. LSU (9-3)
  16. Miami (10-3, ACC runner-up)
  17. Stanford (10-2)
  18. Notre Dame (9-3)
  19. Arkansas (9-3)
  20. Michigan (9-3)
  21. Texas (9-3)
  22. Northwestern (9-3)
  23. Auburn (8-4)
  24. Wash State (8-4)
  25. Kansas State (8-4)

Ohio State beats Florida State in the Rose Bowl, which is a game I will despise watching. Alabama beats Penn State in the Sugar Bowl. We once again get Ohio State and Alabama in the title game, but this time Alabama wins. Wow. 'Bama wins it again. Again, here's my filled-in spreadsheet for reference.

Predicting every single game makes it a lot more difficult to throw out a statement of "I think X team will have Y wins and Z losses." Like, I don't think both Ohio State and Alabama will finish undefeated, but I also don't feel like assigning either of them a loss this year. I don't think Miami will go 10-2, but I trust them more than anyone else in the division. Ultimately, it's just a lot easier to predict something like "Miami will go 8-4 this year" without backing that statement up with a rundown of who will serve those losses, and how those games shake out in the bigger picture of the ACC. Et cetera. 

So that's how I wind up with 'Bama and Ohio State meeting for a rematch. Last year, I got 'Bama and Clemson correct at 1 and 2, but I had Stanford in third and Houston in fourth ... so I went wacky with the third and fourth picks this year. In 2016 I finished with a pick percentage of .716, so I guess my goal will be to top that this go-round.

I hope some people enjoy this crazy spreadsheet this year. Here's to another great college football season.