I see what you did there ESPN. Pretty gutsy College football rankings: ESPN reveals 2024 preseason poll
To be clear, the FPI isn’t a poll. It’s a computer algorithm. It’s been out for a couple of months. The preseason version is based on returning production and transfers, but after a few games the rating is solely determined by on field outcomes. You can read about how it works below. A short note mentions Florida: Some schools have risen up the ranks since the end of last season. Florida is up from 41st to 20th on the strength of 15 returning starters and a strong showing in the transfer portal. College football FPI release: Top teams, new playoff trends
It also projects us to finish with a 5.9-6.1 record this year. I don't ever recall seeing any end of year poll with a 6-6 team ranked in the top 25. But I could be wrong.
Good point. This is something that always perturbed me about polls. The #20 team would lose to the #5 team by 6 points and then drop in the rankings. But shouldn’t that be exactly what the #20 team does? This problem is addressed by the algorithms. If a computer predicts that a team should lose by 10, and it only loses by 5, the rating of the losing team increases. The #20 team should lose 6 games if those games are against the #1-19 teams.
That's our honest ranking that does NOT use our strength of schedule into consideration... and that SOS is the reason why people polls don't have The Gators in the top 25.
Wide receivers Ja’Quavion Fraziars and Andy Jean both have lower body injuries that will keep them out for an extended period of time. An abscessed tonsil in the case of CB Devin Moore.
Let’s get Moore a dose pack, toradol and an antibiotic. Get him in the field! Go GATORS! Man I’m so ready for 8/31!
One more nerd note on ESPN’s FPI: it turns out to be a relatively good predictor compared to its competitors. A website called prediction tracker looks at game outcomes vs predictions for many algorithms each year. I compared the results of the models that had data for the last five years (skipping 2020 when many models weren’t running) and found, unsurprisingly, that the Vegas line was the best predictor of game outcomes, predicting 72.9% of contests correctly (the midweek and updated lines basically tied for #1). The ESPN FPI ranked 9th, predicting 72.4% of contests correctly. The “worst” model on the list was Cleanup Hitter, which ranked 46th on the list, predicting 67.3% of contests correctly. Summary: In the long run, about best you can do is to predict the winner of a college football game ~73% of the time. This is the accuracy of the Vegas line, but many algorithms, including the FPI, are virtually as good.
IV decadron and he will be ready to go. Used to give it in the ER when they couldn’t even swallow. 2 hours later they were drinking like a champ. Also get the ENT to cut that open and he’s dancing.
To the OP preseason polls are dumb. Should not even try until after week 4. Way to many changes and portal has exacerbated this.
Interestingly, statistician Ken Pomeroy found that the preseason AP poll in basketball is pretty good, and actually a better predictor of postseason success than the final AP poll. You might be surprised to hear this, but I’m a big fan of the pre-season AP poll. There is no doubt poll participants have their biases in the pre-season. They may tend to over-estimate the importance of the previous postseason, especially when a team needed more than its fair share of luck to advance. But otherwise, whatever biases are present are uniquely individual, and in the collection of 70 or so ballots, those biases are cancelled out, leaving a useful signal. The end result is that it provides a better picture of the state of college hoops before the season begins than any single person or algorithm could produce. It’s informed groupthink at its finest. The preseason AP poll is great Now perhaps the story is different in football or has changed with the transfer portal, but I was always surprised by his finding in bball.