questions

I have been checking the exp formula with my buff. I have found that even with the 200% buff, my exp is twice of what it "should" be.

Like, I put in all my stats for the battle. It comes up with exactly half according to the exp formula * my 200% buff. Then I times my exp by two and I get what I truly get in my battle.

Ender you being helpful and giving out 2x exp?

Beware that these formulas are kinda screwy. They are sometimes off by a factor of 2, 4, etc.

-Experience - Against Training Bots, Botsunauthorized

It's always times 2 though.

So you're saying you're getting 400%?

can i have the xp-formula .. i want to test too

I guess I am getting 400%.

Let me check with out the buff.

Without Buff 200%
Exp Formula   x2      Actual
222889      445779  401201
316983      633967  570570
236583      473165  425849

With Buff 200%
Exp Formula   x2     xBuff  Actual
219938      439875  879751  791776
270972      541944  1083888 975500
245856      491712  983423  885082

Everything seems to be 11% higher after the x2

Perhaps this:

<[trainbot level] * [damage] * [INT]/[level&win-loss related variable]/ 2 * [trainbot level variable] * [block related variable]> * <[buff modifier]>

is incorrect.

Maybe the buff modifier only modifies part of that, and not the entirety of that function.

I tried messing with it again and can't seem to figure out the whole half as much exp thing. I don't think Ender put in the /2 after the win/loss variable.

Whoops. Thought I had 72 int this whole time and only had 63. That makes up for the 11% difference.

Still the problem of x2 though.

All I know is that my average EXP with 200% EXP buff is much lower than double my average unbuffed EXP. If anything, it's closer to only 150% of my unbuffed XP.

Could you run some tests?

Fuck no and fuck you!

Here's the exp code if you want to take a look:

public static function getExperienceTrain($isWin, $trainerLevel, $level, $damage, $intel, $blocks, $attempts) {
    switch (true) {
        case $trainerLevel < 300:
            $expMod = 1.0;
            break;
        case $trainerLevel < 450:
            $expMod = 1.3;
            break;
        case $trainerLevel < 600:
            $expMod = 1.6;
            break;
        case $trainerLevel < 750:
            $expMod = 1.9;
            break;
        case $trainerLevel < 900:
            $expMod = 2.2;
            break;
        case $trainerLevel >= 900:
            $expMod = 2.5;
            break;
    }

    $blockMod = 1 + $blocks / ($attempts - $blocks);

    switch (true) {
        case !$isWin && $trainerLevel < $level:
            $variable = 300;
            break;
        case !$isWin && $trainerLevel >= $level:
            $variable = 200;
            break;
        case $isWin && $trainerLevel < $level:
            $variable = 150;
            break;
        case $isWin && $trainerLevel >= $level:
            $variable = 100;
            break;
    }

    return round($trainerLevel * $damage * $intel / $variable / 1 * $expMod * $blockMod);
}

Buff exp is applied to the result of this.

return round($trainerLevel * $damage * $intel / $variable / 1 * $expMod * $blockMod);

should be

return round($trainerLevel * $damage * $intel / $variable / 2 * $expMod * $blockMod);

sorry everyone, but why would you divide by 1?

That likely explains the comment on BU about the formula being off by a factor of 2. What's in the game now is correct because early on I'd verified using builds with known exp/hour values with the help of Ms. Rose.

Then why divide by one? :P

Probably as a reminder to myself that leaving off the divide by 2 was an intentional choice and not an omission.

okie dokie.