Back in the late 70s we used to make fun of the simplest use of the transitive property (just the final score) to compare teams and predict results. Context, even with lots of relational data points, still matters and that’s why I listen to you guys and respect what you have to say.
The main issue is teams like Georgia or Ohio State arent necessarily putting a ton of effort into or opening up the playbook in games against group of 5 or bad power 5 teams so taking away a bunch of stats and info from those games is pointless. The only thing that matters is how they do against teams that have somewhat equal talent. If Alabama lost to say Vanderbilt, no one would think Vanderbilt is actually the better team, it means obviously bama did something wrong. The model is broken if it takes a single thing from games played against Charlotte or eastern Washington and then uses that to predict how two ranked teams would do against each other.
Most college football fans, and sports fans for that matter, don’t have a clue. They argue in anecdote and personal belief, and for fun conversations amongst friends that’s fine. But when it comes to data and different types of data, it’s not on par with it. You two are saints for even bothering to try, but I suppose that patience is also required to develop this channel, your model, and obtained the educations you’ve got to get to this point anyway. So kudos all around.
The general goal post move in Colorado was when many many people not believing they'd win at all, then people who know nothing start chipping at a natty when that is not a realistic goal. Now the not win people are using the no nothings goal
College football fandom is firmly placed at the confluence of emotional reasoning and Dunning Krueger Effect. I appreciate guys like yourselves and Josh Pate for trying to bring a little pragmatism and analytical thinking to the space.
Could we say that what Daniel and Josh do is 100% using the transitive property but in a much larger sense? They just don't look at final scores, they look at what contributed to those scores, what teams did against other opponents, and how teams performed in different aspects of the game against all opponents, and then form an estimate of what will happen. It's transitive but not in the simple A beat B beat C, so A will beat C sense, but rather A has done this vs B and others, and B has done this vs C and others (and C has done this vs others), so A will likely do this vs C. They are adding a lot more variables, or factors rather, to the equation. It's all still predicting future outcomes based on past performances, just not being simple minded about it..
Transitive properties for the most part don’t work in college football because these are kids not professionals . I think that can’t be more stressed when doing comparisons like this . I know it’s one of the only metrics but it usually doesn’t end up working
@@CollegeFootballNerds ya no I get the concept you’re going with to an extent and it works to an extent but college football is played by amateurs and for the most part you don’t always get consistent play from players week in week out
On Shedour Sanders winning the Heisman: there are still currently 30 other players who have better odds to win the Heisman than he does. Not 3 players. Not 5. Not 10. Not 20. Not 25, but THIRTY. Shedour Sanders is currently at 200/1 to win the Heisman on FanDuel. To put those odds in perspective; here are the top five players with the best odds to win the Heisman: Michael Pennix and Caleb Williams are a little less than 4/1. Quinn Ewers is 5/1. And Bo Nix is 10/1. Again, Shedour Sanders is 200/1
It was just silly, uninformed, and unnecessary. I said at the time that he was likely the 5th best QB in his own conference, which people took as hating. But Penix, Williams, Nix, and DJU were also tearing things up at an even higher level, statistically. Colorado has honestly brought out a lot of bandwagon fans that have very little understanding of just how tiered CFB really is (including how low on that tier list TCU and Nebraska were expected to be this year).
How about we get back to what brought a lot of us to your channel. Talking football and comparing teams and discussing what each of you and the model thought about games. As far as I can tell y’all have done two game reviews for this week. What about LSU vs Ole Miss, Oregon St. vs Utah, Texas vs Kansas. Florida vs Kentucky. I’d like to hear about that as opposed to videos like this or no video at all. It’s just my opinion, but it seems like this channel is getting away from what made it great to begin with.
Through the history of our channel, we have consistently done 2-3 game previews per week. That's what we have time to do, and it's nothing new. We spent 15 extra minutes this week recording this rant and a model explainer, and our choice was that-or-nothing. With all due respect, this isn't our full-time job, and we have not (nor never will) have time to cover 6 games in a week so long as that's true. Just not something we can make happen.
The fact that this video was necessary is a solid reminder that there are a lot of morons out there. A mathematical anomaly that has always confounded me is this: The average person is at a below average level of intelligence. I’m convinced this is fact, but I’m also aware that it isn’t possible. These are the thoughts that keep me up at night.
That is actually a logical fallacy. It is why so many don't understand how a poor version of Ohio State (TCU) ran all over Michigan. A bigger part is they don't know where Ohio State's weakness even is. Some are close in saying short run game or similar, but they don't know why.
This is not what people are looking for from you two. Maybe during the doff season but you are barely releasing college football videos. Sorry this was a miss for me
Dude, we did 3 preview videos last week, we did 2 previews this week, which is our normal cadence. We also have a model explainer in the works, and we launched a whole website. That's about the limit of what we have time to do. This isn't our day job.
Daniel making the biggest mistake of all time, "telling his wife that she's wrong" 😅😂
Facts
We must chose the hill we want to die on and it’s almost always not the hill we find ourselves on, in that moment 😂
I thought it was odd how Daniel suddenly stopped responding to texts after releasing this...
I was spiking my danger signals as soon as I heard it! 😂
@@CollegeFootballNerds may Daniel's memory be eternal.
Back in the late 70s we used to make fun of the simplest use of the transitive property (just the final score) to compare teams and predict results. Context, even with lots of relational data points, still matters and that’s why I listen to you guys and respect what you have to say.
The main issue is teams like Georgia or Ohio State arent necessarily putting a ton of effort into or opening up the playbook in games against group of 5 or bad power 5 teams so taking away a bunch of stats and info from those games is pointless. The only thing that matters is how they do against teams that have somewhat equal talent. If Alabama lost to say Vanderbilt, no one would think Vanderbilt is actually the better team, it means obviously bama did something wrong.
The model is broken if it takes a single thing from games played against Charlotte or eastern Washington and then uses that to predict how two ranked teams would do against each other.
On top of the fact that’s it’s really hard to get these talented rosters up to play these cupcake games.
Should have titled this episode: “Tearing Down The Transitive Goalposts”
Most college football fans, and sports fans for that matter, don’t have a clue. They argue in anecdote and personal belief, and for fun conversations amongst friends that’s fine. But when it comes to data and different types of data, it’s not on par with it. You two are saints for even bothering to try, but I suppose that patience is also required to develop this channel, your model, and obtained the educations you’ve got to get to this point anyway. So kudos all around.
Y'all fight with folks on Twitter and I fight with folks on TH-cam 😂
My pet peeve is from my own fanbase.
When i don't predict my team to win a NC, then people say "you must just be a fan of [rival school]"
Bama fan?
@@DrSlipstream Auburn
@@BradenINSANE Oh weird. I run into mostly doomer Auburn fans saying 6-6 is the ceiling. War damn btw
The general goal post move in Colorado was when many many people not believing they'd win at all, then people who know nothing start chipping at a natty when that is not a realistic goal. Now the not win people are using the no nothings goal
That's why CFB polls are infuriating.
It's okay, Charlotte deserves it
College football fandom is firmly placed at the confluence of emotional reasoning and Dunning Krueger Effect. I appreciate guys like yourselves and Josh Pate for trying to bring a little pragmatism and analytical thinking to the space.
Now do “simulated pressures.” I don’t think there’s an announcer in college football who has ever used that term right.
Could we say that what Daniel and Josh do is 100% using the transitive property but in a much larger sense? They just don't look at final scores, they look at what contributed to those scores, what teams did against other opponents, and how teams performed in different aspects of the game against all opponents, and then form an estimate of what will happen. It's transitive but not in the simple A beat B beat C, so A will beat C sense, but rather A has done this vs B and others, and B has done this vs C and others (and C has done this vs others), so A will likely do this vs C. They are adding a lot more variables, or factors rather, to the equation. It's all still predicting future outcomes based on past performances, just not being simple minded about it..
What does relative performance mean on the website?
Why did you guys stop doing Saturday night live streams?
Please just try to get cfb fans to learn to use adverbs and the difference between "lose" and "loose"
Good video. But it’s like trying to bail out the ocean!
When does the model on y’all’s website update? Is it updated by say Monday after a weekend of football?
It updates automatically as scores come in from the games. So starting Thursday night, those games will impact the model downstream.
@@CollegeFootballNerds thank you!
Suggestion: leave Twitter. It sounds like you're surrounded my morons on Twitter... in other words, you're experiencing typical Twitter.
Transitive properties for the most part don’t work in college football because these are kids not professionals . I think that can’t be more stressed when doing comparisons like this . I know it’s one of the only metrics but it usually doesn’t end up working
It's the (rather large) distinction between judging a team off one game, and refusing to acknowledge overall schedule quality, that's the rub here.
@@CollegeFootballNerds ya no I get the concept you’re going with to an extent and it works to an extent but college football is played by amateurs and for the most part you don’t always get consistent play from players week in week out
On Shedour Sanders winning the Heisman: there are still currently 30 other players who have better odds to win the Heisman than he does. Not 3 players. Not 5. Not 10. Not 20. Not 25, but THIRTY. Shedour Sanders is currently at 200/1 to win the Heisman on FanDuel. To put those odds in perspective; here are the top five players with the best odds to win the Heisman: Michael Pennix and Caleb Williams are a little less than 4/1. Quinn Ewers is 5/1. And Bo Nix is 10/1. Again, Shedour Sanders is 200/1
It was just silly, uninformed, and unnecessary. I said at the time that he was likely the 5th best QB in his own conference, which people took as hating. But Penix, Williams, Nix, and DJU were also tearing things up at an even higher level, statistically. Colorado has honestly brought out a lot of bandwagon fans that have very little understanding of just how tiered CFB really is (including how low on that tier list TCU and Nebraska were expected to be this year).
Can y’all point out Georgia has just not been good at setting an edge all season
I’ve seen you guys use the transitive property plenty of times on here when it benefitted your stance
I was expecting a Keynesian quote: “when the facts change, I change my mind, what do you do, Sir?” @Josh
How about we get back to what brought a lot of us to your channel. Talking football and comparing teams and discussing what each of you and the model thought about games. As far as I can tell y’all have done two game reviews for this week. What about LSU vs Ole Miss, Oregon St. vs Utah, Texas vs Kansas. Florida vs Kentucky. I’d like to hear about that as opposed to videos like this or no video at all. It’s just my opinion, but it seems like this channel is getting away from what made it great to begin with.
Everyone said all these newcomers are good for football. This is what we have to deal with. Not feeling great for the fans.
Through the history of our channel, we have consistently done 2-3 game previews per week. That's what we have time to do, and it's nothing new. We spent 15 extra minutes this week recording this rant and a model explainer, and our choice was that-or-nothing. With all due respect, this isn't our full-time job, and we have not (nor never will) have time to cover 6 games in a week so long as that's true. Just not something we can make happen.
The fact that this video was necessary is a solid reminder that there are a lot of morons out there. A mathematical anomaly that has always confounded me is this: The average person is at a below average level of intelligence. I’m convinced this is fact, but I’m also aware that it isn’t possible. These are the thoughts that keep me up at night.
I assume if we were to find out what Daniel thinks he is right about with his wife will lead to the Collgee Football Nerd show.
Haha you said doodoo
Second.
....or whatever.
That is actually a logical fallacy. It is why so many don't understand how a poor version of Ohio State (TCU) ran all over Michigan. A bigger part is they don't know where Ohio State's weakness even is. Some are close in saying short run game or similar, but they don't know why.
This is not what people are looking for from you two. Maybe during the doff season but you are barely releasing college football videos. Sorry this was a miss for me
I, for one, welcome our new College Football Algebra Overlords.
I agree this is not what I’m looking for
Dude, we did 3 preview videos last week, we did 2 previews this week, which is our normal cadence. We also have a model explainer in the works, and we launched a whole website. That's about the limit of what we have time to do. This isn't our day job.
First
First