Top 25 teams since 1960 (power ratings)

quote:

---

MilkJug

---

Is every thread you start now just a veiled advertisement for your website?

quote:

---

Is every thread you start now just a veiled advertisement for your website?

---

No. I'm sparking conversation... of which you just started. So, thanks.

The graphic is auto-generated, easier (and prettier) than typing it out. Do you suggest I not do a graphic anymore?

These metrics are... weird, to say the least.

Alabama's national title teams of 2015 and 2017 don't make the list, but 2016 and 2018 do?

Two different teams from 2019 (including the #3 team) are higher than the actual national champions, LSU? It's not like LSU had an easy schedule.

1989 Houston? That lost 2 games? They only faced 3 teams that finished ranked and lost to 2 of them.

I get that you're saying "records" don't count, but apparently neither does head-to-head. How else do you have 2019 LSU behind a team they beat pretty severely and who played a softer schedule (2019 Clemson), who in turn beat your #3 team (2019 Ohio State)?

Look, I'm all for making sure historical teams don't get buried under the conceits of modern perspectives; that said, if your system is giving results this flawed then you should probably double check how you are weighing certain factors.

And, to be frank, 2011 Alabama and 2011 LSU, if missing from any list like this (but does contain 1989 Houston) is definitely flawed.

quote:

---

I get that you're saying "records" don't count, but apparently neither does head-to-head. How else do you have 2019 LSU behind a team they beat pretty severely and who played a softer schedule (2019 Clemson), who in turn beat your #3 team (2019 Ohio State)?

---

Head-to-head does matter, but it's not "special". It uses the team's entire body of work. Again, where a team "deserves" to be placed isn't considered. If 2019 LSU and 2019 OSU were to play tomorrow, this model would favor OSU. If 2019 LSU and 2019 Clemson were to play tomorrow, this model would favor Clemson (very slightly). Just because a team beats another team once, doesn't mean they will beat them 100% of the time. We see this every year. Every game is different. This model is a measurement of how dominant a team was against their schedule and then adjusted to a common league-average team across eras. It works very similar to SP+, Sagarin, ESPN FPI. Go look at ESPN's FPI rating for 2019. It's OSU, Clemson, then LSU. I know you probably think "well ESPN FPI is shite", well these models are the best we can do at predicting the outcomes of games. They're not "great", but they are the best we can do with the limited data we have.

quote:

---

And, to be frank, 2011 Alabama and 2011 LSU, if missing from any list like this (but does contain 1989 Houston) is definitely flawed.

---

2011 LSU and Bama had all-time great defenses. They did not have great offenses, relative to other eras. They were just out of the top 25. 2011 LSU - 31.4, 2011 Bama - 30.4. And actually, the 2016 Bama defense had one of the best defenses on this list (and much better than 2011 Bama), can you guess why?

quote:

---

Look, I'm all for making sure historical teams don't get buried under the conceits of modern perspectives; that said, if your system is giving results this flawed then you should probably double check how you are weighing certain factors.

---

I'm not "weighting certain factors". It's machine-learned. The model optimizes itself to be as accurate as it can (with its given data) at determining outcomes.

My apologies, Grok had the wrong numbers, they lost by 6... in the semis... to the team who lost by 17... to the team you have ranked behind both.

quote:

---

y apologies, Grok had the wrong numbers, they lost by 6... in the semis... to the team who lost by 17... to the team you have ranked behind both.

---

Look at the rest of OSU's schedule that year. You're putting too much weight into one game.

I'm assuming you watched that OSU v Clemson game. Tell me, if 2019 OSU and 2019 Clemson played tomorrow, who do you think would win? I would have OSU...

quote:

---

They didnt lose on a fluke play or questionable call... they lost by 17 head to head on a neutral field to end the season.

You can try to rationalize to yourself why your model is as it is, but it doesnt stand even basic logical checks regardless of what pretzels you have yourself tied up in to explain it away and as such isnt worth further exploring or discussing.

Its like bad click bait from a podcast crew drumming up interest during the offseason.

---

Ok. Well Alabama clobbered Georgia in the regular season in 2021. Then Georgia kicked their arse in the NC game. I guess we should go rewrite the history book, because those outcomes make no sense...

I would say 95 Nebraska or 01 Miami. Curious how how good that Miami defense would have been at containing that Nebraska running game. We weren't ready for that rushing attack at all in that championship game. Nebraska's O linemen were all juiced. I am not even sure if those O linemen did much in the NFL as the juice was carrying them in college. We were pretty bad that game though and couldn't even wrap up their 200 pound QB on all those runs.

Auburn should have won that 2013 game. They had the lead in the final minutes