Two years ago when I launched this public project, I went ahead and did bowl predictions for every single damn bowl game, both for me and for my computer. I didn't do it last year and thought about making that public this year but due to time, decided nah fam. This year for my bowl picks, I've exclusively went with just my regression model from this year (meaning zero input from me no matter what), which is good and bad. We are currently standing at 19-15, which is already one win better than what my computer model had two years ago (didn't track last year), so it seems we have a slight model improvement with the change in formula. But at the same time we're still shooting a measly 55%. Unfortunately, we hit 3 of those losses in one, very sad day where each of those three games came down to the final possession. But whatever, let's break this down a bit.
So again the regression model I used is only tracking how many wins a team should be getting based on the numbers they are putting up. This is a good measurement to track what a team is capable of long term, but we need to consider other factors such as potential schedule (see Central Florida). This is also a bad thing, as a long term forecast will not do a great job of short-term forecasting in the event of injuries, inactives, weather, etc. Which is why I'm bummed at myself for not doing a better job of tracking games as the season progressed. However, I will say that when you retroactively apply the final standings in the regression model to the entire season, we're getting about 80% accuracy in all of the 1000 some games played (which is how I determined the bowl picks). Not bad. But again, not great when what we're trying to accomplish is a perfect predicting machine, especially if you're trying to do it in the moment. (perfection is impossible, but we can get damn close to it right?)
Which is why I'm happy and sad at our current 19-15 bowl record. This model is not the greatest at predicting short term outcomes (or being fed off-the-field information), so we either need to make further adjustments, or use some human information when making these picks. But why I am happy, is that we've only had a handful of bad misses (6 out of 15. Compared to two years ago we ended up with a total of 11 bad misses. Mind you we still have a handful of games to go. A bad miss being defined as a pick losing by 14+ points), 2 of our losses could have been prevented with some information (Oregon not having Royce Freeman, and Northern Illinois not having their starting QB). The rest of the losses have been by one possession, which is frustrating because we could be looking at a 26-8 record if those broke that way (one can argue that the one possession wins could have also broke for losses, you hush with that. Positive vibes only). Next year I would like to do a better job of keeping track of this model as it runs through the season (again we'd probably have to start at around Week 3 in order to have a large enough sample size for the standings to reflect properly), and I would like to also compare it to the model I used to run, which in my opinion, did a better job of predicting on a game by game status. But that is also the beauty of bowl games, random shit happens because different teams have different motivations, and we are likely to see some random upsets.
Either way I'm still somehow in the top 20% of the nation for bowl picks, so we take those?
We're Talking Playoffs?:
Anyway, the real reason for this post is that the 4th year of the College Football Playoff is upon us, and every year it is the same shit about which teams should be in it or not. My immediate answer to this is "Stfu, no matter how you do it, some team is going to be pissed off". In the BCS era, it was the top 2, with #3 getting left out. In this era, we have top 4 with #5 getting left out. So I asked myself, is there possibly a way to satisfy top teams without diluting the privilege of playing for the national championship? Let's look.
Hypothetical Situation #1: So this year; we have Clemson (ACC Champ), Oklahoma (Big 12 Champ), Georgia (SEC Champ), and Alabama (SEC at-large), leaving out Ohio State (Big 10 Champ). So let's say we expand to 8 teams: Hypothetically this year, we might've seen:
The above five
Auburn (SEC runner-up)
Wisconsin (Big Ten runner-up)
and either USC (Pac-12 Champ) or Miami (ACC runner-up)
So that last point right there already brings up one problem. We still have a fight for that 8th spot just in that hypothetical situation alone. In my personal opinion, I think both teams suck and shouldn't be in, but if we need 8 teams, we need 8 teams. Who should it be? And what about Central Florida? Even at 8 teams, they're still getting snubbed (again if I had it my way, it would probably be UCF over both USC and Miami, but I never hold the popular opinion).
Hypothetical Situation #2: So let's try to get UCF into the playoffs. What about 16 teams? That's a nice even number:
The above nine
Central Florida (Undefeated, AAC Champ)
Penn State (Big Ten at-large)
Washington (Pac-12 at-large)
TCU (Big 12 runner-up)
Notre Dame
Stanford (Pac-12 runner-up)
???
So now we have a similar, but still new, problem. Who should the 16th team be? We have a lot to choose from, and to be honest, I think it's a stretch having even Stanford in, but again we need to fill spots and I think a conference runner-up should be in (this is implying Washington is better than Stanford and don't tell me otherwise. Or show me the score of that game). You can already see that we're diluting the national championship by giving these conference runner-ups and at-larges the chance. What if by some magic a 9-win Stanford team runs the table to win it all? What the fuck is what I would be thinking. And then even still, we need to fill that 16th spot. Memphis, Boise State, Troy, South Florida, and Toledo are the only other teams with 10 wins not mentioned, and if we inserted a generic 9-win team (let's say Michigan State because they are ranked 16th), then that opens up the can of worms of deciding who is truly the best 9-win team in the nation. We'd need our own damn playoff to figure that one out.
Hypothetical Situation #3: So let's try the the basketball approach. Good ol' March Madness has a nice system of giving every conference winner an automatic bid to the tournament, but shit even at 68 teams we're still throwing bitch fits about why #69, 70, 71, or 72 didn't get in. But for the sake of argument and curiosity, here is my proposed "Champs first" 16 team playoff:
Clemson (ACC)
Oklahoma (Big 12)
Georgia (SEC)
Ohio State (Big Ten)
USC (Pac-12)
Central Florida (AAC)
Florida Atlantic (C-USA)
Toledo (MAC)
Boise State (Mountain West)
Troy (Sun Belt)
Alabama, Auburn, Wisconsin, Penn State, Miami, TCU (runner-ups/at-larges)
In this instance, we are replacing Washington, Notre Dame, Stanford and whoever our 16th mystery team from above, with conference champions (FAU, Toledo, Boise, and Troy). Now again, this seems like it would work out, but is that fair to the above teams to be left out? And who gets to decide what? Penn State has the same status as Washington being an at-large team, why did I arbitrarily put them in instead (in a world where the Fiesta Bowl hasn't happened yet)? Do we really want to give these other four minor conference champions a chance even though they didn't even go undefeated (much less in their own conference)? In my opinion, this becomes the closest solution as we are respecting conference champions as a whole, but we are indeed allowing some lesser teams have a shot, at the sacrifice of a team that could probably beat them, and still asking ourselves the question, whom of the at-larges, should truly be in the playoff?
Bottom line, no matter how you stack it up, unless our society as a whole is ok with letting some lesser teams have a shot like the basketball tournament, we just need to suck it up. Someone is not going to be happy. An 8-team playoff could work, barring we don't have any Central Floridas. But we still struggle with trying to find out who that 8th team would be without the #9 team getting all pissy. 16 teams does sound like a lot, but the NCAA loves their money so they probably wouldn't be against it, but that is allowing a lot of teams that may be undeserving a shot at the title, despite being conference champions. But boy it would be funny if Troy beat Alabama too.
Anything more than 16 teams is just getting out of hand so don't even go there.
