This week we have SEVEN previously unranked teams that have jumped into the top 25. Some of them, like Cincinnati or Mississippi State, are foreseeable. Others, like Syracuse and Kansas, are less so.
The rise of those seven teams came courtesy of seven teams who lost in shocking upsets. Of the seven teams that dropped from the rankings, five (Washington, Baylor, Oklahoma, Minnesota, Florida State and Pittsburgh) were favored on average by nine points. They proceeded to lose by an average margin of 15 points.
So, in other words, it was a chaotic weekend for college football. And ZeLO thrives on chaos.
Sort of. When ZeLO is compared to FPI, in weeks where there is predictability, FPI and its heavy level of certainty is a massive advantage. ZeLO and FPI could pick games at the exact same clip, and FPI could still easily blow ZeLO out of the water.
In weeks of parity, however, like this past weekend, ZeLO picked better than FPI and had more accurate picks overall (as measured by Brier Points). ZeLO went 45-18, scoring 323.2 Brier Points, while FPI scored 44-18 and 296.4 (FPI did not pick the Boise State game this weekend).
As we have progressed through the season and have begun to get actual results, I have been decreasing ZeLO’s weighting of preseason projections and increasing the actual statistics from this season. Given ZeLO’s success over the past two weeks, I think it is calibrated just right and it now has a firm understanding of what teams really are.
So now that ZeLO has a better understanding of what teams actually look like (and to some extent what Notre Dame actually looks like), I thought it would be worthwhile to revisit the schedule and talk about which games will be tricky for the Irish and how that has changed from preseason expectations. Instead of listing them out by difficulty, I decided to make a tier list.
Games Notre Dame is still expected to lose
No. 5 Clemson (34.5%)
No. 16 BYU (39.5%)
The expected news first: Notre Dame’s play has seriously decreased their chances against some more formidable opponents.
Since the preseason, the model has loved BYU and had the Cougars as the best independent team. That was reflected in the Irish’s win percentage, which in the preseason was 44%, but is now down 4.5 points. I think this is a weird game for ZeLO to predict. This is a neutral site game where BYU has a distance advantage, but Notre Dame has fans that travel well. I am willing to admit that ZeLO may be overlooking the Irish here. Only time will tell.
A game that does not get the same excuse is the Clemson game. Despite being the home team, the Irish were still underdogs (at 40.9%). The only significant change was that Notre Dame is now even more of an underdog. I would expect Clemson to have an even better chance of beating the Irish by the time the game rolls around. Clemson will face BC, a team they should easily beat. They also face Florida State and Syracuse, teams they will likely be heavily favored to beat.
Preseason wins that are now projected losses
No. 22 Syracuse (38.6%)
No. 6 USC (38.9%)
This is by far the most exciting category for me. Syracuse, a team I doubt even the most delusional Syracuse fan had starting 5-0, has done exceptionally well this season. In the preseason, Notre Dame was a 57% favorite, despite being on the road. Now, the Irish are the underdog when they head on the road. This projection could easily change. Syracuse plays No. 14 North Carolina State and No. 5 Clemson before they take on Notre Dame, so a lot could change.
USC is different from Syracuse. While ZeLO was reasonably skeptical of USC early on, the Trojans were expected to be good and after their performance thus far through the season, ZeLO has now swung the other way. Originally, the Irish were a fairly comfortable road favorite at 66.2%, but with the two teams trending in opposite directions, the Irish dropped 27.3% to be an underdog.
Decreased win chances, but still projected wins
Luckily, the news is not all doom and gloom for the Irish. Notre Dame still has four games that they are heavily favored in. These are games I would be shocked if they lost.
Stanford and BC were projected to be two of the easiest games for Notre Dame this season. While that has not changed much, Notre Dame’s odds of winning against the Cardinals have dropped by nearly nine percent. The Irish gave up less than 5% against BC. Notre Dame should still win these, but it might be closer than you think.
UNLV also saw a 9% increase in winning chances against Marcus Freeman’s squad, though I do think this could change somewhat as the two teams play out the rest of their season. I expect the same thing of Navy. Notre Dame should look better as the game approaches, but the slow Irish start has dropped their win percentage by about 2%.
Week 6 Picks
No. 8 Tennessee @ No. 25 LSU
And ZeLO starts with a flaming hot take. After having LSU as a mid to bottom-tier SEC team, the model has swung the other way, favoring LSU to upset Tennessee. Now, LSU is 4-1 and has had strong showings against Mississippi State and a road win against Auburn. And if LSU had not had inferior special teams play, they could easily be 5-0. So, while it feels crazy, the Tigers are home and on a roll.
No. 17 TCU @ No. 19 Kansas
You read that right, TCU and Kansas are not only playing in a top-25 game, but both are also undefeated and both sit atop the B12 Conference standings. ZeLO has TCU as a slight favorite, just 56.6%, and this game, as weird as it is to read (let alone write it), could have important conference ramifications.
No. 11 Utah @ No. 18 UCLA Bruins
UCLA has been seen by ZeLO as a strong team and a dark horse to win the PAC-12. With Washington’s stumble from last weekend, the current PAC-12 Championship matchup is UCLA and USC. This game could very easily change that projection, as the Utes, whom ZeLO favors winning this game (just slightly) at 52.3%, is on the outside looking in. This game has massive conference and playoff implications.
A win for Utah would keep them as a 1-loss team, which keeps them as a potential playoff team (emphasis on potential). A win for UCLA would help give them some actual legitimacy.