Categories
Sports

Zwiller: ZeLO’s CFP rankings and Week 10 picks

If you are a diehard college football fan like I am, then the CFP Rankings announcement show is appointment television. So, like many of you, I tuned into ESPN at 7 p.m. and saw the top 25 announced.

And the results were somewhat controversial. Having Georgia in third place while Tennessee took first was undoubtedly a strong take. And TCU sitting in seventh despite being undefeated, behind one-loss Alabama, while not surprising, certainly feels disrespectful.

Prestige rankings devalue Vols

So I turned to ZeLO to look at the prestige rankings, a metric I devised to take the ZeLO rankings along with some external factors to project the CFP.

Last week, when I wrote about the prestige rankings, I gave out ZeLO’s ten teams most likely to make the CFP (Ohio State, Georgia, Clemson, TCU, Oregon, Tennessee, Michigan, Alabama, Oklahoma State, Penn State and USC).

This week when I looked at the Prestige Top-10, the only significant change was that Penn State fell to 19th, with Illinois moving into ninth. 

While I like that ZeLO is a consistent and predictable model, ZeLO had the No. 1 team in the country (Tennessee) behind TCU and Oregon.

That is a problem, as the Volunteers now possess one of the best routes to the CFP. They need only beat the No. 3 Georgia Bulldogs this weekend and then either beat the Tide or keep it close, ensuring they are a one-loss team whose loss is a split to the SEC Champion.

What? I said it was the best route, not the easiest.

The problem was genuinely concerning. I wanted the Prestige rankings to predict how the CFP would rank teams accurately. Not having the No. 1 team in the top 5 is a significant miss.

The solution was fairly straightforward, however. Part of the Prestige rankings was considering the future projections ZeLO has made, so projected win-loss records, remaining SOS and conference champions.

ZeLO has Tennessee losing to Georgia, making them a one-loss team that does not make their conference championship, which explains why they are on the outside looking in. Everyone ahead of the Volunteers is projected to win their conference championship, by ZeLO, at least.

So, I removed the future projections component of the prestige rankings, and here is what ZeLO gave me.

ZeLO’s CFP Rankings

In a not-at-all-shocking twist, had Ohio State (95.46) in first, and nothing I did would change that.

Tennessee did move up to second, while Georgia fell into third, so the committee got one of its hottest takes correct (despite my vehement disagreement).

Clemson’s ranking fourth did not at all surprise me. What did surprise me was Clemson (87.45) having a nearly identical rating compared to Georgia (87.85), which I suppose says a lot about the two teams’ dominance over some (respectfully) meh teams.

That will change for both this weekend as Georgia plays Tennessee and Clemson faces an Irish team hellbent on repeating their 2020 upset victory.  

In fifth and sixth are Michigan (86.88) and Alabama (86.51). The difference between these two schools is their record, though Michigan’s SOS prevents it from making the top 4 over Clemson.

The rest of the Top 10 is where ZeLO starts to disagree with the CFP rankings. In seventh lies USC (85.22). Oregon (84.93) sits eighth, TCU (84.03) ranks ninth and UCLA (82.81) rounds out the top ten.

Considering ZeLO had the right teams, just jumbling the order makes me feel pretty good about the model’s job predicting the rankings.

That ZeLO swapped out LSU (10th by the CFP) for UCLA (12th) seems logical. At best, putting LSU in 10th feels like giving the Alabama-LSU game some additional stakes.

It also has the bonus of giving Alabama a resume booster, so when the committee tries to sneak a 2-loss Crimson Tide team in at four at the end of the season, they can point to this matchup.

I kid (sort of).

Last week ZeLO had one of its best weeks of the season, going 37-11 (77.1%) and beating FPI, who went 34-14 (70.8%). FPI still has a 14-game lead, but ZeLO has a 3-game lead since Week 4. 

I will get to another analysis piece on why I think ZeLO has outperformed FPI, but because of the CFP, I had to delay it another week.

Now, onto the picks.

No. 6 Alabama @ No. 15 LSU

This is the most critical Alabama-LSU game since the 2019 meeting between the two squads when No. 2 LSU traveled to Tuscaloosa to take on the No. 3 Tide. That game acted as an elimination game, with both squads being undefeated and tied for first place in the division.

LSU would go on to win the game, denying Alabama a chance to win the SEC championship, ultimately preventing them from making the CFP.

This time, the stakes are different. LSU is most likely out of the CFP as they have two losses (Florida State and Tennessee), but the Tigers can play spoiler to the Tide. The Tide does have a chance to make it to the Playoff, although they would need to win out, winning the SEC. LSU could deny them that chance and make it to the SEC title game themselves.

ZeLO thinks the Tide will win the game and keep their playoff ambitions alive, giving Alamaba a 59.3% chance of winning. 

No. 2 Tennessee @ No. 1 Georgia

It is hard to think that a top-3 matchup between two divisional foes is less important than Alabama and LSU but hear me out. The loser of this game will, in all likelihood, finish the season as a one-loss team to the potential conference champion. A team with that resume could still easily make the CFP.

ZeLO currently favors Georgia to win the game and the conference, but Tennessee fans should still be optimistic about the game and the season. As we learned on Tuesday, a win against Alabama is something this Committee highly values. A one-loss Volunteers’ squad should still make the CFP.

No. 5 Clemson @ Notre Dame

This result shocked me when I first saw it, but ZeLO is giving the Irish a 39.1% chance of knocking off Clemson. That might not sound like a lot, but those odds have dramatically improved since Notre Dame dropped their game against Stanford.

Two reasons drive the change. One, Notre Dame has looked competent again, handling their business against UNLV and then making Syracuse look like UNLV the following week. Those numbers have helped the Irish move up ZeLO’s power rankings.

The second reason is that Clemson has looked, well, mortal. They have narrow wins against Syracuse, Florida State and Wake Forest. Those results have not helped Clemson’s stats, and in the eyes of ZeLO, Clemson is not the same team they were projected to be. The Tigers may look great on paper, but they may be a paper-tiger.

This might still be a Clemson win, but it should be close.

Leave a Reply