He won't be. Just saying.Yeah, at some point he’s going to have to practice if he intends on playing against TCU.
He won't be. Just saying.Yeah, at some point he’s going to have to practice if he intends on playing against TCU.
Yeah, at some point he’s going to have to practice if he intends on playing against TCU.
And the more variables one’s model has, the less accurate they become and the more variations one has to run.All models are wrong, some models are useful.
It's not models itself but rather the inputs and assumptions and data controls used in them.If models are so worthless/controversial, why do Vegas and pro gamblers use them
so much?
It’s a good thing that computers are fast and can simulate hundreds of thousands of outcomes within a few minutes.And the more variables one’s model has, the less accurate they become and the more variations one has to run.
That's accounted for by the teacup of marbles.It's not models itself but rather the inputs and assumptions and data controls used in them.
How is Manhattan's model accounting for schematic changes, especially for a new coach? What assumptions are being made there? How about changes in competition? Is it the same playing in the SEC v Pac 12?
MB needs to provide data to support his model being able to actually model CU here.
Why the f would MB need to provide that?It's not models itself but rather the inputs and assumptions and data controls used in them.
How is Manhattan's model accounting for schematic changes, especially for a new coach? What assumptions are being made there? How about changes in competition? Is it the same playing in the SEC v Pac 12?
MB needs to provide data to support his model being able to actually model CU here.
1. Schematc changes on their own are not as important to offensive player predictions since my model rates usefulness on a per pay basis. The model does account for plays per game, which is the critical statistic. There’s ample evidence in the model from Kent State and similar offenses to help predict per game play counts.It's not models itself but rather the inputs and assumptions and data controls used in them.
How is Manhattan's model accounting for schematic changes, especially for a new coach? What assumptions are being made there? How about changes in competition? Is it the same playing in the SEC v Pac 12?
MB needs to provide data to support his model being able to actually model CU here.
Are you implying your model beats the battle-tested system of:1. Schematc changes on their own are not as important to offensive player predictions since my model rates usefulness on a per pay basis. The model does account for plays per game, which is the critical statistic. There’s ample evidence in the model from Kent State and similar offenses to help predict per game play counts.
2. Defensive players are modeled on a usefulness per play basis in all schemes. The key variable we’ve found is the number of offensive plays, not the defensive scheme.
3. Competition is assessed by opponent inputs. My model rates every player. A productive player against other productive players has a higher rating than a productive player against less productive players.
4. My model does not apply only to Colorado. It applies to every team in FBS. We rate FCS players after their season concludes for transfer purposes only.
My inputs and analtytical choices are carefully designed to be as accurate as possible since the predictions generated are attached to actual wagered dollars. My models in American College Football and European Football have beaten sportsbooks every year for over a decade.
hell yeah love a good k hole on game dayAre you implying your model beats the battle-tested system of:
>wake up at 9am
>hungover
>take a piss
>”oh damn there’s CFB today”
>pours cereal bowl
>sits on couch in boxers
>eat cereal and open draftkings
>bet on G5 traps you know nothing about
My friends have used it to great success. I think they’re only a few K in the hole
I wish that I got to bet against those people. My opponents are corporations who have business relationships with sports media/leagues.Are you implying your model beats the battle-tested system of:
>wake up at 9am
>hungover
>take a piss
>”oh damn there’s CFB today”
>pours cereal bowl
>sits on couch in boxers
>eat cereal and open draftkings
>bet on G5 traps you know nothing about
My friends have used it to great success. I think they’re only a few K in the hole
Lots of dudes leaning heavily into the downers these days, never got it as an upper connoisseur myself. Call me LilStimmyhell yeah love a good k hole on game day
I wish that I got to bet against those people. My opponents are corporations who have business relationships with sports media/leagues.
All of this and no top 25 or pac predictions. Come on @manhattanbuff, i gotta see what you got and if it's the goods.1. Schematc changes on their own are not as important to offensive player predictions since my model rates usefulness on a per pay basis. The model does account for plays per game, which is the critical statistic. There’s ample evidence in the model from Kent State and similar offenses to help predict per game play counts.
2. Defensive players are modeled on a usefulness per play basis in all schemes. The key variable we’ve found is the number of offensive plays, not the defensive scheme.
3. Competition is assessed by opponent inputs. My model rates every player. A productive player against other productive players has a higher rating than a productive player against less productive players.
4. My model does not apply only to Colorado. It applies to every team in FBS. We rate FCS players after their season concludes for transfer purposes only.
My inputs and analtytical choices are carefully designed to be as accurate as possible since the predictions generated are attached to actual wagered dollars. My models in American College Football and European Football have beaten sportsbooks every year for over a decade.
You too good for some spray paint in a paper bag, Mr Fancypants?Lots of dudes leaning heavily into the downers these days, never got it as an upper connoisseur myself. Call me LilStimmy
MB needs to provide data to support his model being able to actually model CU here.
Teabagged lolololololololMy models in American College Football and European Football have beaten sportsbooks every year for over a decade.
You have to call his 1-900 number. $4.99 per minute.All of this and no top 25 or pac predictions. Come on @manhattanbuff, i gotta see what you got and if it's the goods.
Let's see who got more of the goods, me or you. I just watch a lot of football, listen to coaches, and know tendencies extremely well. I don't have any cpu algorithms or anything like that. I actually think that makes you a lot better for teams I don't watch but for teams I watch and heavily watch, idk, I think I can compete. I was one of the top Draftkings guys for years and I never create a cpu algorithm once or paid anyone for anything.
Because the model says something differently than I want it to say after mainlining koolaid for months.Why is the model of someone who could quit his day job and be a professional gambler being questioned when the models he uses, not only for CFB but for European soccer as well, are sought after by extremely wealthy clientele lol
Thats a good question.Why is the model of someone who could quit his day job and be a professional gambler being questioned when the models he uses, not only for CFB but for European soccer as well, are sought after by extremely wealthy clientele lol
Don't be that guy. I haven't disrespected no one here. Don't be that guy. I've been respectful the entire time even when I wasn't aware of his history.Because the model says something differently than I want it to say after mainlining koolaid for months.
I think Denver_sc was only talking about himself, not anybody else.Don't be that guy. I haven't disrespected no one here. Don't be that guy. I've been respectful the entire time even when I wasn't aware of his history.
In Barzil, I have a thread that details the answer to your request.All of this and no top 25 or pac predictions. Come on @manhattanbuff, i gotta see what you got and if it's the goods.
Let's see who got more of the goods, me or you. I just watch a lot of football, listen to coaches, and know tendencies extremely well. I don't have any cpu algorithms or anything like that. I actually think that makes you a lot better for teams I don't watch but for teams I watch and heavily watch, idk, I think I can compete. I was one of the top Draftkings guys for years and I never create an cpu algorithm once or paid anyone for anything.
As Ken said, I was only speaking for myself.Don't be that guy. I haven't disrespected no one here. Don't be that guy. I've been respectful the entire time even when I wasn't aware of his history.
I listened to his podcast this morning. I felt like he really wanted to say they will start 3-0, but just couldn't.Klatt was very complimentary and bullish on CUs chances this season and said he thinks CU wins at least one of the first two. He’s also not very high on TCU.