crootin

[shrugginatlas]

Last week, I posted on the other board a stupid thought I've been kicking around on college football recruiting. It is so riveting that it received precisely no feedback. That is sad, because it would be good to get comments before trying to move forward with the model. I'd rather not focus on some of the other parts of the model without getting the core parts right.

Anyway, the model can be found here.

 

[HOF Coach]

Last week, I posted on the other board a stupid thought I've been kicking around on college football recruiting. It is so riveting that it received precisely no feedback. That is sad, because it would be good to get comments before trying to move forward with the model. I'd rather not focus on some of the other parts of the model without getting the core parts right.

Anyway, the model can be found here.

I could not have said it any better!

 

[RaleighPittFan]

Is Chewie the one who used to pose with signs admitting that he ate his own poop? Or, was that a prior dog?

PS, the paper is good.

 

[shrugginatlas]

Is Chewie the one who used to pose with signs admitting that he ate his own poop? Or, was that a prior dog?

I am happy to report that this is *not* Chewie. He just has every other bad habit.

Glad you enjoyed. Criticism welcome.

 

[pittdan77]

Can you sum this up in simple language for those of us that fell asleep in statistics?

 

[shrugginatlas]

1. Crootin takes effort. Coaches have only so many resources, so crootin is, in many ways, just a question about how to devote resources among many risky prospects.

2. Players differ in how much they respond to crootin effort. Some croots are very responsive, while others are not. And, some croots have higher baseline affinity for a school. These differences affect how coaches ought to deploy their crootin effort. One of the punchlines from the analysis as it sits is that the latter is better than the former.

3. Crootin decisions aren't made in a vacuum, because other coaches are crootin, too. Each croot is a contest where a bunch of crootin efforts are the inputs in a function that outputs a set of probabilities, one for each coach.

So, in other words, you have to take what might be called a "general equilibrium" approach, where all coaches and all croots are modeled at once. No coach in a vacuum, because crootin is a competition among many coaches. And no croots in a vacuum, because coaches have only limited resources to deploy across all the croots.

The next analysis (assuming we feel comfortable with the substantive features of this model) will be to work scouting in.

 

[RaleighPittFan]

1. Crootin takes effort. Coaches have only so many resources, so crootin is, in many ways, just a question about how to devote resources among many risky prospects.

2. Players differ in how much they respond to crootin effort. Some croots are very responsive, while others are not. And, some croots have higher baseline affinity for a school. These differences affect how coaches ought to deploy their crootin effort. One of the punchlines from the analysis as it sits is that the latter is better than the former.

3. Crootin decisions aren't made in a vacuum, because other coaches are crootin, too. Each croot is a contest where a bunch of crootin efforts are the inputs in a function that outputs a set of probabilities, one for each coach.

So, in other words, you have to take what might be called a "general equilibrium" approach, where all coaches and all croots are modeled at once. No coach in a vacuum, because crootin is a competition among many coaches. And no croots in a vacuum, because coaches have only limited resources to deploy across all the croots.

The next analysis (assuming we feel comfortable with the substantive features of this model) will be to work scouting in.

Build a wall around the local kids, and be a good enough team (record wise; NFL pipeline; Bowl Games; academics; or, whatever the kid is interested in) that the ones from out of state want to come to Pitt to play.

 

[SempleFest06]

1. Crootin takes effort. Coaches have only so many resources, so crootin is, in many ways, just a question about how to devote resources among many risky prospects.

2. Players differ in how much they respond to crootin effort. Some croots are very responsive, while others are not. And, some croots have higher baseline affinity for a school. These differences affect how coaches ought to deploy their crootin effort. One of the punchlines from the analysis as it sits is that the latter is better than the former.

3. Crootin decisions aren't made in a vacuum, because other coaches are crootin, too. Each croot is a contest where a bunch of crootin efforts are the inputs in a function that outputs a set of probabilities, one for each coach.

So, in other words, you have to take what might be called a "general equilibrium" approach, where all coaches and all croots are modeled at once. No coach in a vacuum, because crootin is a competition among many coaches. And no croots in a vacuum, because coaches have only limited resources to deploy across all the croots.

The next analysis (assuming we feel comfortable with the substantive features of this model) will be to work scouting in.

Sounds about right to me. Many on this site dont understand #1 and #2. They don't understand how much extra resources/money it takes to effectively recruit the states of FLA and TX. That is why it is so important to have success in the WPIAL.

 

[pittdan77]

1. Crootin takes effort. Coaches have only so many resources, so crootin is, in many ways, just a question about how to devote resources among many risky prospects.

2. Players differ in how much they respond to crootin effort. Some croots are very responsive, while others are not. And, some croots have higher baseline affinity for a school. These differences affect how coaches ought to deploy their crootin effort. One of the punchlines from the analysis as it sits is that the latter is better than the former.

3. Crootin decisions aren't made in a vacuum, because other coaches are crootin, too. Each croot is a contest where a bunch of crootin efforts are the inputs in a function that outputs a set of probabilities, one for each coach.

So, in other words, you have to take what might be called a "general equilibrium" approach, where all coaches and all croots are modeled at once. No coach in a vacuum, because crootin is a competition among many coaches. And no croots in a vacuum, because coaches have only limited resources to deploy across all the croots.

The next analysis (assuming we feel comfortable with the substantive features of this model) will be to work scouting in.

Thank you.

My question is, mathematically, how does the situation change once the number of coaches/schools is truly narrowed down? Does effort increase the probability of a win or is it more about timing and knowing when to apply the effort?

 

[shrugginatlas]

That's a good question. I'll look into it.

 

[joeydavid]

When did illiteracy become a thing?

Last week, I posted on the other board a stupid thought I've been kicking around on college football recruiting. It is so riveting that it received precisely no feedback. That is sad, because it would be good to get comments before trying to move forward with the model. I'd rather not focus on some of the other parts of the model without getting the core parts right.

Anyway, the model can be found here.

 

[recruitsreadtheseboards]

1. Crootin takes effort. Coaches have only so many resources, so crootin is, in many ways, just a question about how to devote resources among many risky prospects.

2. Players differ in how much they respond to crootin effort. Some croots are very responsive, while others are not. And, some croots have higher baseline affinity for a school. These differences affect how coaches ought to deploy their crootin effort. One of the punchlines from the analysis as it sits is that the latter is better than the former.

3. Crootin decisions aren't made in a vacuum, because other coaches are crootin, too. Each croot is a contest where a bunch of crootin efforts are the inputs in a function that outputs a set of probabilities, one for each coach.

So, in other words, you have to take what might be called a "general equilibrium" approach, where all coaches and all croots are modeled at once. No coach in a vacuum, because crootin is a competition among many coaches. And no croots in a vacuum, because coaches have only limited resources to deploy across all the croots.

The next analysis (assuming we feel comfortable with the substantive features of this model) will be to work scouting in.

Did you write the code for the old NCAAFootball franchise for EA Sports where you had to recruit "prospects" and they all responded to different inputs??

 

[cashisking884]

Sounds about right to me. Many on this site dont understand #1 and #2. They don't understand how much extra resources/money it takes to effectively recruit the states of FLA and TX. That is why it is so important to have success in the WPIAL.

Yep. Just diminishing returns. A few 3* players here and there, with every fan base thinking, "those states produce a high rate of underrated players. So of course this player is one of those underrated players." And yet, there really is no team you can point to that has success with this strategy. At least at any meaningful level, for a sustained period of time.
Far better to take that money and go out and buy the best recruiters that are plugged into the PA-OH and mid-Atlantic area. Try to convince the 4* to stay close to home, rather than the 3* to travel across country.