re: The Big 10 and ACC are setting the stage for locking up 3/4 playoff spots

quote:

---

Ideally, but we're dealing in a reality where you're more than happy with an arbitrary number.

---

and what objective metric are you putting forward here? I’d struggle to think of a metric I enjoy or find particularly fair that would grant merit towards a team for playing significantly fewer football games relative to their peers.


I’ll put it like this, in any attempt to find objectivity; like in trying to rank resumes in a sport, sample size is paramount. With a larger sample size, like with playing more games, you allow for the elimination of outliers and a better gauge as to the relative quality of teams. It’s why the computer rankings are generally regarded as terrible until a few weeks into the season, because they begin to normalize as more data pours in. This is also why it isn’t some linear map of relative importance (i.e. the difference in our ability to discern relative quality of teams between playing 11 and 12 games each is not the same as playing 0 and 1 games each)

Now if two teams have a one game disparity between them, I think it’s more than fair to say that you can compare the weakest opponent of the first team to the open date on the second team’s schedule and say playing the actual football team is significantly more impressive. I’m really not sure how that’s even in dispute. You have to go out there, risk injury and an upset, and actually play a football game.

quote:

---

Teams like UCF, Boise State, and Utah are laughing at you.

---

did they ever get screwed over by a team they had a better resume than?

quote:

---

and what objective metric are you putting forward here?

---

It's always interesting to see objectivity tossed around so flippantly when it suits one's position.

If you want objectivity, the rules must be changed. Because they aren't we're left with subjectivity.

Make an objective argument for anyone else.

quote:

---

did they ever get screwed over by a team they had a better resume than?

---

Subjectively, sure. And that's what we have.

You'll make some claims about objective metrics, and you'll fail, because no matter what we do those metrics are subjectively weighted depending on what we need for our desired outcome.

Fans do it. Coaches do it. Voters do it. And the committee does it.

quote:

---

I have no skin in the game regarding OSU, I just think you’ve failed to convince me that our desire for larger sample sizes is arbitrary enough to discard entirely. Obviously all computer formulas and criteria are inherently subjective in regards to the importance we ascribe to them or their input variables, but algorithms like ELO are probably the best we can do to minimize subjectivity and I think a commonly agreed upon composite of such algorithms (akin to Massey) could suitably be used to provide enough objectivity to not rail against the arbitrary nature of this process. It’s what I’d do, personally. And I think minimizing subjectivity in these types of processes is a goal with a lot of merit, and sufficient sample size is going to be imperative with such a goal.

---

Ok.

Where does Massey rank Ohio State, A&M, etc?

quote:

---

This says nothing to the fact that I reject completely the point I believe you were making earlier that the jump from an open date to an FCS school is equivalent for analysis purposes to the jump between an FCS school and a conference foe.

---

What do you think Alabama's annual 60+ point win over teams like Western Carolina tells us? Anything we wouldn't know of they sat home that week?

Looks like 6th. I can send you the link if you need it.

So the closest thing we have to objectivity seems to slightly undervalue Ohio State and Cincinnati, and overvalue A&M?

quote:

---

I would probably argue the formulas being used haven’t appropriately normalized for number of games played, but as I sense we aren’t going to converge on this issue and it’s time for food, I’ll leave on that note. But I understand you can attack me for wanting to tinker with an objective metric and if we have the freedom to do so, it introduces subjectivity. But it comes down to I think the objective metric needs to be commonly agreed upon, and for my money I think the algorithm needs some adjustment.

---

That's inherently subjective. To have any amount of objectivity in data analysis, we analyze the data we have. We don't manipulate it to our liking or modify our processes because the data we get isn't the data we want.

quote:

---

I think you at least see the data driven reasons and the pragmatic (the injury point) reasons why I think we need to take a serious look at significant discrepancies in number of games played and not cast it off so easily.

---

Of course, just the same as I'd think we could agree that the Massey rankings show us more than an arbitrary number of games played.

Ohio State needs to play. If NW beats them, they are out and not deserving of a top 4 spot anyway. I wish we would’ve played Texas A&M this past weekend.

I don't disagree, but I'm not the one who brought up Massey or Sagarin, both of which favor Ohio State more than the committee.

I don't know what games against Michigan and Rutgers would tell us that we don't already know.

And even if we accept the premise that 6 games versus 8 games creates a huge gap in the analysis (I don't accept that), who should be the 4th team?

As a Florida fan, it isn't us. It wasn't going to be us without winning out, no matter what Ohio State did next week. Who else?