Testing the Toy

[Misguided Rhetoric]

I know it’s just rankings (take with a grain of salt) but there is no way to explain Irion County being ranked below May and Jonesboro (having beaten both). I know strength of schedule is the dominant factor but head-to-head should be factored in somehow.

 

[wildcatfan4]

I know it’s just rankings (take with a grain of salt) but there is no way to explain Irion County being ranked below May and Jonesboro (having beaten both). I know strength of schedule is the dominant factor but head-to-head should be factored in somehow.

I agree on IC should be above May and Jonesboro. I’m still not convinced strength of schedule is in the calculation.

 

[FenceLineGuy]

I agree on IC should be above May and Jonesboro. I’m still not convinced strength of schedule is in the calculation.

It's not. Had this talk with @Mike on it and strength of schedule has nothing to do with rankings.

 

[ModeratorMike]

I agree on IC should be above May and Jonesboro. I’m still not convinced strength of schedule is in the calculation.

It's not. Had this talk with @Mike on it and strength of schedule has nothing to do with rankings.

Not directly, no, but beating teams that are rated higher will get you a lot more points than beating lower rates teams.

Rest assured, I share all of your confusion in a lot of cases and I'll spend hours...many many hours...looking for ways to make it better this off-season. I was unsuccessful last year, finding any way I could've improved it, but we'll see how it goes this off-season.

 

[DOWNBOX]

I know it’s just rankings (take with a grain of salt) but there is no way to explain Irion County being ranked below May and Jonesboro (having beaten both). I know strength of schedule is the dominant factor but head-to-head should be factored in somehow.

The Toy only crunches numbers that a human enters into it.
Ie.
Mike can overload or underload the numbers the Toy crunches.

Myself, I think too much emphasis is put on strength of schedule, at least early in the season.

 

[wildcatfan4]

The Toy only crunches numbers that a human enters into it.
Ie.
Mike can overload or underload the numbers the Toy crunches.

Myself, I think too much emphasis is put on strength of schedule, at least early in the season.

I think teams that stack their schedule will top ranked teams want to go off strength of schedule

Teams that schedule less quality teams want to go of W/L.

At the end of the day it’s a number and doesn’t mean much. It would be different if this was D1 college and everyone is fighting for a playoff spot.

 

[kdstew3]

I think teams that stack their schedule will top ranked teams want to go off strength of schedule

Teams that schedule less quality teams want to go of W/L.

At the end of the day it’s a number and doesn’t mean much. It would be different if this was D1 college and everyone is fighting for a playoff spot.

So teams that might've played only one good team close, but didn't even win 5 games, are ranked too high, IMO.

I think strength of schedule, in addition to strength of win/loss, would be good. Teams play tough schedules, get blown out in most of the games against good teams, but then get rewarded in the rankings lol.

 

[DOWNBOX]

So how do we determine strength of schedule?
It is an opinion at the start of the season, before any games are played.
Look at Strawn this year started season ranked (like)5th, ended season 25th or 26th.
W/L is how teams get into the playoffs.

 

[ModeratorMike]

So how do we determine strength of schedule?
It is an opinion at the start of the season, before any games are played.
Look at Strawn this year started season ranked (like)5th, ended season 25th or 26th.
W/L is how teams get into the playoffs.

Strength of schedule, if it were a thing, would be fluid. It'd change each week as every teams ratings changes. If your schedule consisted of the top 10 in state week 1, you'd have a very high strength of schedule. If those 10 teams moved down throughout the season, so would your strength of schedule.

As I've said before, though, strength of schedule isn't something that is factored in. There is no strength of schedule rating. Beating higher ranked teams gets you more points than beating lower ranked teams. That's the full extent of SoS.

Maybe there is too much emphasis on beating high ranked teams, especially if they turn around and shut you out the next meeting or if you've lost to multiple other lower ranked teams? Maybe in those instances, where teams play twice, we should throw the first game out of the calculations? I'm not sure, but testing will tell me.

 

[T-boneisGood]

The Toy only crunches numbers that a human enters into it.
Ie.
Mike can overload or underload the numbers the Toy crunches.

Myself, I think too much emphasis is put on strength of schedule, at least early in the season

Would you explain what you mean by “overload or underload the numbers the Toy crunches?”

As I understand, the toy only works on point differential with a range of 1-45 in whole numbers.
I’m just nerdy and like to understand more if I’m missing something.

 

[DOWNBOX]

Would you explain what you mean by “overload or underload the numbers the Toy crunches?”

As I understand, the toy only works on point differential with a range of 1-45 in whole numbers.
I’m just nerdy and like to understand more if I’m missing something.

Mike says strength of schedule is not factored in, so he then could not overload or underload.

 

[ModeratorMike]

Would you explain what you mean by “overload or underload the numbers the Toy crunches?”

As I understand, the toy only works on point differential with a range of 1-45 in whole numbers.
I’m just nerdy and like to understand more if I’m missing something.

That one confused me too.

The only numbers I plug in are scores. The only numbers that are ever considered are the scores I put in, and the team's ratings. What it does with those numbers, however, is pretty complex. There are quite a few things that I can tweak, though, which I'll do plenty of this off-season.

 

[T-boneisGood]

Mike says strength of schedule is not factored in, so he then could not overload or underload.

Cool, thank you.

 

[wildcatfan4]

So teams that might've played only one good team close, but didn't even win 5 games, are ranked too high, IMO.

I think strength of schedule, in addition to strength of win/loss, would be good. Teams play tough schedules, get blown out in most of the games against good teams, but then get rewarded in the rankings lol.

That one confused me too.

The only numbers I plug in are scores. The only numbers that are ever considered are the scores I put in, and the team's ratings. What it does with those numbers, however, is pretty complex. There are quite a few things that I can tweak, though, which I'll do plenty of this off-season.

Thank Mike for explain.

I do have question. How does one rating stay high if they are eliminated in the playoffs or district? If you put a number in what number would you put for them?

 

[DOWNBOX]

That one confused me too.

The only numbers I plug in are scores. The only numbers that are ever considered are the scores I put in, and the team's ratings. What it does with those numbers, however, is pretty complex. There are quite a few things that I can tweak, though, which I'll do plenty of this off-season.

Teams ratings.
At the start of a new season how do you get a Teams rating?

 

[jamwash]

You got it...starting next week...

if team == 'May' { newRating = oldRating - 100 }

The code is writing in JS?

 

[jamwash]

Dam you and your logical explanations

And yes I spelled that incorrectly on purpose. Don't wanna be throwing out 3 cent cuss words to the guy who can ban me for life

3 cent? Use to be 1.75 cents. Must be inflation

 

[ModeratorMike]

Thank Mike for explain.

I do have question. How does one rating stay high if they are eliminated in the playoffs or district? If you put a number in what number would you put for them?

No scores mean I don't put anything in for that team. It would be no different than if they had a bye week. They can still move up or down, based on what their opponents do. For example, a team with a 100 rating, who didn't make the playoffs could move up to a 120 rating if one of the teams they beat during the season moves up significantly as well. If they originally got 15 points for beating that team, and that team continues to improve, then by mid-playoffs they could be getting 20 points for beating that team rather than just the 15.

Teams ratings.
At the start of a new season how do you get a Teams rating?

That's what the pre-season ratings are. Those are put in by me, currently. Last season I didn't change anything from the season before. This season, I kept 90% of the teams in the same order they were already in, but I moved a few up and down based on the information I had about those teams. For example, I moved Benjamin up to #1 because I thought with the kids they had, plus their move-ins, they would be a very dominant force. Once I had all the teams where I wanted them, in order, I multiplied them all by like 0.6, or something (I really don't remember), just to bring the numbers down and make the teams closer together because Granger told me that's what needed to be done. Finally, I made sure every team was above 60, again because Granger said it needed to be that way. As I pointed out in a few tests last year, though, the starting point means less and less as the season goes on. By the time playoffs are here, it wouldn't matter if a team started in last place, they'd still be pretty close to where they are regardless.

The code is writing in JS?

No, it's written in Python and PHP. I just wrote it like that so it might be easier for folks to understand the point I was making. That might not have been the case if I'd written out the actual code necessary to do take points away from a specific team. Ha!

 

[Tigerfan85]

Or just put school names in a hat draw them out and it is what it is. Just leave May out. Seriously you do a lot for this site for us to come on here and have fun. The work you do is appreciated for sure. Well by most of us those WB folks are just hard to please.

 

[jamwash]

No scores mean I don't put anything in for that team. It would be no different than if they had a bye week. They can still move up or down, based on what their opponents do.

Wouldn’t it make sense to have the final ratings after the season then have a “playoffs rating” list

No, it's written in Python and PHP. I just wrote it like that so it might be easier for folks to understand the point I was making. That might not have been the case if I'd written out the actual code necessary to do take points away from a specific team. Ha!

Good ole Python (and whatever PHP is). Honestly I bet the entire algorithm could be written in Python