League Forums

• «
• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• »

Re: If There Is Any Hope of Surviving Version 0.4.3........

By Infinity on Trial

8/31/2018 7:19 am

jdavidbakr wrote:

Here is a rough explanation of the math with the attribute movement. The movement of the future value is based on a normal distribution random number. That means that it should trend toward its mean. The standard deviation of this distribution is based on the distance between the current and future values. Volatility is a value that moves the mean. So 0 volatility means the mean is zero - which means that the movement is just as likely to go up as it is to go down. Volatility of 100 means that the mean of the distribution is at the max either up or down. The only thing you don't know is whether that 100 is positive or negative.

So yes, someone with 0 volatility can and always has been able to recover what they lost in their first training camp. Training camp, though, moves a distance of about equal to the movement during the entire season, so to move back you would need an inverse of roughly the same random number generated at each game stage all season long. If a 0 volatility player moves by a value outside of the third standard deviation (0.2% chance) you're right, he probably won't recover from that. But if the more likely scenario happens where he's at the outside of one standard deviation, he could recover most of that during the season and in the next training camp. But he's also just as likely to continue to drop.

I appreciate this explanation, but in my experience, I've never seen a player lose potential then gain even part of it back. Or vice versa.

Liked by hchoudhry17, CrazyRazor, setherick

Re: If There Is Any Hope of Surviving Version 0.4.3........

By setherick - League Admin

8/31/2018 7:52 am

Infinity on Trial wrote:

jdavidbakr wrote:

Here is a rough explanation of the math with the attribute movement. The movement of the future value is based on a normal distribution random number. That means that it should trend toward its mean. The standard deviation of this distribution is based on the distance between the current and future values. Volatility is a value that moves the mean. So 0 volatility means the mean is zero - which means that the movement is just as likely to go up as it is to go down. Volatility of 100 means that the mean of the distribution is at the max either up or down. The only thing you don't know is whether that 100 is positive or negative.

So yes, someone with 0 volatility can and always has been able to recover what they lost in their first training camp. Training camp, though, moves a distance of about equal to the movement during the entire season, so to move back you would need an inverse of roughly the same random number generated at each game stage all season long. If a 0 volatility player moves by a value outside of the third standard deviation (0.2% chance) you're right, he probably won't recover from that. But if the more likely scenario happens where he's at the outside of one standard deviation, he could recover most of that during the season and in the next training camp. But he's also just as likely to continue to drop.

I appreciate this explanation, but in my experience, I've never seen a player lose potential then gain even part of it back. Or vice versa.

That's because it does not happen with players, and it only rarely happens with coaches. And the coaches that go up and down basically stay within a standard deviation of their starting score, and therefore, never actually improve. The whole mechanism is deterministic as opposed to relative based on coaches, owners, playbooks, attributes like DI and IN, etc. I think Vol just needs to go away for players and relative gains and losses be based on a combination on coach skill and a players IN and DI. It would certainly make those attributes more interesting.

But, right now, I'd settle for a passing game that was not nerfed and coverage that worked.

Re: If There Is Any Hope of Surviving Version 0.4.3........

By Infinity on Trial

8/31/2018 8:01 am

Given the wonky results we are seeing and evidence that contradicts JDB's explanation, I have to wonder if there are bugs in the calculations. As it is, player development is working about as well as power rankings.

Re: If There Is Any Hope of Surviving Version 0.4.3........

By setherick - League Admin

9/25/2018 10:35 am

lellow2011 wrote:

...also putting a fast, light guy at TE and throwing to them deep is highly exploitable. They usually get LB coverage and end up wide open down the middle.

This was always the case. It's also really easy to stop by adjusting coverages.

• «
• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• »