Changing player skill ratings

[rcleslie]

jd, Your random height generator worked like a charm. I was then able to run your editor with no sweat. I will have to alter some of the heights -- they can't all be NBA centers -- but that is a minor issue compared to the tedious labor your generator has rendered unecessary.

Now here is a question you might be able to answer: does your random height generator produce a normal or a rectangular distribution of heights? The reason I ask is that it might be applied to generating ratings distributions for all skills, and perhaps used to set distribution parameters (probably just the means) so as to create teams of varying character and strength. At the very least it would be interesting, with, e.g., an entirely fictitious college conference to create two or three "top" teams, one or two "doormats" and the rest "good" to "not too bad", etc. without having to edit each entire roster. Just a thought. I have no idea how involved that might be.

Apart from such pipe dreams I again thank you for your help! You should get a promotion!

[jdhalfrack]

jd, Your random height generator worked like a charm. I was then able to run your editor with no sweat. I will have to alter some of the heights -- they can't all be NBA centers -- but that is a minor issue compared to the tedious labor your generator has rendered unecessary.

That's what my aim was. I figured that some (or many...) of the heights would be weird, but I just wanted you to be able to use the editor. [:)]

Now here is a question you might be able to answer: does your random height generator produce a normal or a rectangular distribution of heights?

[&:] I'm not sure what that means. [&:] Basically, it just took a random number between 65 (5'5") and 84 (7'0") for each player and applied it to the player. It then saved the league database.

The reason I ask is that it might be applied to generating ratings distributions for all skills, and perhaps used to set distribution parameters (probably just the means) so as to create teams of varying character and strength. At the very least it would be interesting, with, e.g., an entirely fictitious college conference to create two or three "top" teams, one or two "doormats" and the rest "good" to "not too bad", etc. without having to edit each entire roster. Just a thought. I have no idea how involved that might be.

While, I am not sure 100% you are saying, I do think you can do some of what you are asking already with the editor I have made. If not, write it in a little bit more specific terms of what you want (I'm kind of stupid... so explain it REAL easy to me [;)] ), and I'll see what I can do with future versions of my editor, or something else.

You should get a promotion!

[:)] Thanks for the vote of confidence. I do aim to please. [:)]

JD

[Marauders]

Basically, it just took a random number between 65 (5'5") and 84 (7'0") for each player and applied it to the player. It then saved the league database.

 
He was asking if it was just random, on a bell curve, or had another variable.
 
Perhaps 68 to 80 inches would give better overall results.  There are not many 5'5" or 7'0" players in football.  Well, not for adult teams anyway.

[jdhalfrack]

He was asking if it was just random, on a bell curve, or had another variable.

Perhaps 68 to 80 inches would give better overall results.  There are not many 5'5" or 7'0" players in football.  Well, not for adult teams anyway.

OOOOOOH, duh.... You know, I didn't even think about that. Of course I should have made it peak in the middle (or around 6'3"ish) and spread it out in a bell curve.... god I AM an idiot...

Well, I added a random option to my global edits on the next version of my editor... now I will have to go back and add the option to make it a bell curve random with the user allowed to pick the spread and the center. That makes so much more sense now! Thanks guys! [:)]

JD

[Marauders]

JD, you are always so enthusiastic.
 
With so many positive waves, maybe we can't lose, baby!
 
Woof woof woof!
 
 
 

[rcleslie]

JD, I was going to say that Marauders was right about the two curves and try to write something explaining that a bit more. Fortunately you caught on without my having to do that. My family tells me that my explanatory style is "disertational" so you have saved yourself and anyone else reading this part of the forum from a probable bore.

To answer the more important question re what I want to do:

1. I like simulating college and/or high school seasons, thus
2. I don't have a whole lot of use for the pro draft structure for my own team (although it will do).
3. What I would like to be able to do is begin a season with a first practice/try-outs group of 75-85 players whose ratings in all categories are normally distributed (i.e., fit a bell curve)and are not already keyed to positions. Then I can study the rating profile for each player and decide what position or positions that player will play. (I would love it if some could go both ways, D and O.
4. I don't much care what the other teams do - they can use the draft, etc. and will be coached by the AI. I am doing this solitaire.
5. I suppose the easiest way to do what I want within the current game structure is to run the draft for all teams (even mine) and then generate a normal distribution of skill ratings for my own roster using something like JD's program. A normal distribution generator would be a real benefit.
6. I have never finished a season and then played another without resetting the league for one reason or another - usually due to installing a new patch or update. The point being that I haven't experienced what happens re roster changes from one season to the next. I am assuming I can work around that as long as I can print and edit my rosters which I can now do thanks to everyone's help! SO...if this explanation is clear enough...any ideas?

Thanks again JD, Marauders, DW, and all.

[jdhalfrack]

Okay, not to bring up an old topic, but I completed the "normal distribution" option of my editor when it comes to randomly assigning skill attributes. It's pretty sweet. I'm not sure why I didn't do this earlier. Anyway, with the Max FB update coming out "soon," I may just release what I have done now and finish the stats editing portion for the first release after the new update.

Anyway, thank's rcleslie for the idea!

JD

[rcleslie]

jd,

I have finally made use of your distribution utility and it does everything I want and then some. Thanks again!

[David Winter]

Re: Player Height.

According to DW this parameter does not figure in the algorithms that determine play outcomes, so I suspect it may not be a high priority item.

I'm not sure when I said this, but if I did, I misspoke. Player Height can impact play results. In the 3D game, Height can impact the success of passing plays. Taller receivers can catch balls that might be slightly overthrown, whereas a shorter receiver may not be able to catch it. A shorter QB may not be able to throw over linemen as well as a taller one.

[Marauders]

No wonder all of Shaggy's receivers are 7' tall!  That bastard!
 
[;)]

[DreamTeams]

Re: Player Height.

I'm not sure when I said this, but if I did, I misspoke. Player Height can impact play results. In the 3D game, Height can impact the success of passing plays. Taller receivers can catch balls that might be slightly overthrown, whereas a shorter receiver may not be able to catch it. A shorter QB may not be able to throw over linemen as well as a taller one.

I love this.

[quixian]

Player Height can impact play results. In the 3D game, Height can impact the success of passing plays. Taller receivers can catch balls that might be slightly overthrown, whereas a shorter receiver may not be able to catch it. A shorter QB may not be able to throw over linemen as well as a taller one.

Great, now I've got to fix all my player heights and weights for my NFL roster....[;)]

Was hoping I could let that go for a little bit. But, it proved I wasn't crazy when I thought the 6'2 330 version of Randy Moss I came across in playtesting was a bit less than dominating... [X(]