Warrior General Games

Yes I know this implemented several seasons ago, but here are some example just make me curious.......

Build method:
Speedster WR, 68 in., 175 lbs, speed to 25 and catching to 16.
enhance speed to 1 star, intense train 3 times, enhance speed to 2 star, intense train till Day 48.

1 player's speed end with 34 and 22%: http://aaasahi.myweb.hinet.net/GLB/breakthrough.JPG
1 player's speed end with 33 and 81%: http://aaasahi.myweb.hinet.net/GLB/nobreakthrough.JPG

The former use auto-train, the later use manual train. So I guess Bort forgot close breakthrough code under auto-training?

*bug ticket sent.

Its not a Bug per Bort,

"The only thing we disabled are breakthroughs for newly created players to stop an exploit. There's still the training variance."

Further info from previous Testing/Bug Files;
The precise values of the variance on training gains are not known, (initially suggested to be upto +/-2.5%, but thats not official and is known to be incorrect, as variance of +5% has been seen.)

The variance is completely random, and often it appears as though its not awarded, due to the variance value being close enough to 0 for display purposes.

This is something that has always happened, but is very rarely noticed, and is intended per Bort.
It also has no fixed relation to Manual/Auto training.

surely a little 1% here or there is likely, but the amount those guys are different surely means more than the amount youd expect with the same trains?

i know when I've build dots with the same value and trained them pre boosting in the few days of their lives, that the training gains were all pretty exact. to the single %.

Originally posted by Mike Rogers
Its not a Bug per Bort,

"The only thing we disabled are breakthroughs for newly created players to stop an exploit. There's still the training variance."

Further info from previous Testing/Bug Files;
The precise values of the variance on training gains are not known, (initially suggested to be upto +/-2.5%, but thats not official and is known to be incorrect, as variance of +5% has been seen.)

The variance is completely random, and often it appears as though its not awarded, due to the variance value being close enough to 0 for display purposes.

This is something that has always happened, but is very rarely noticed, and is intended per Bort.
It also has no fixed relation to Manual/Auto training.

I know there are training difference.
Other player I created with same method end with 82% 86% 87% 88%.
I don't think training difference can produce 40% gap.

Yeah. Good catch.

Originally posted by aaasahi
I know there are training difference.
Other player I created with same method end with 82% 86% 87% 88%.
I don't think training difference can produce 40% gap.

if gets random roll everytime at 5% more . 5%more X 8 = 40% more

The Law of Large Numbers applies here.

Originally posted by Mike Rogers
Its not a Bug per Bort,

"The only thing we disabled are breakthroughs for newly created players to stop an exploit. There's still the training variance."

Further info from previous Testing/Bug Files;
The precise values of the variance on training gains are not known, (initially suggested to be upto +/-2.5%, but thats not official and is known to be incorrect, as variance of +5% has been seen.)

The variance is completely random, and often it appears as though its not awarded, due to the variance value being close enough to 0 for display purposes.

This is something that has always happened, but is very rarely noticed, and is intended per Bort.
It also has no fixed relation to Manual/Auto training.

Which is why I

Originally posted by CyberNinja

Bort did deem it not a bug but I did ask for clarification and ask him to look at the possibility of break throughs not being turned off when on auto training. But for now, WAI. If he says otherwise, it will be corrected.