IOU Pet Upgrade Analyzer v0.1d by Indigo Ferret

[Tezexlo]

I checked your formulas with the dmg shown ingame...
Your table is a bit hard to use for this. It would be better, if it calculate the normal dmg, crit, etc in different cells.



First, it's not your idle autodmg, but your non idle autodmg, with this, the normal pet dmg fits with the shown value.
Killing Blow is actually halve the shown value (so misleading ingame).
Critdmg is bugged, you actually make less(!) dmg atm. And its dmg*(1+critmulti+0,2). So standard critmulti without upgrades is 20%.
What are you doing with your double attack? It's just 50% of the normal attack...

So the formulas would be:
dmg (so it acually shows the standard dmg the pet does):
=B24*(1+B17)*(1+B18)*(1+B19)*(1+B11+B12-(B13*0,4))

dps:
=(B25*(1-B7)+B25*(B7*(B8+0,2)))*B9*(1+B10/2)*(1+B14/5/2)


I didn't check the speed and if you have a crit chance without an upgrade, since it isn't as trivial as this.

And on the discussion about how we value the killing blow multiplier, I'm here more with IndigoFerret's 20% since the overkill only happens here. So the monster with ~4% hp left will be killed even without the killingblow dmg (depends on monster hp, for me its the 4% atm).

@IndigoFerret
Thx for the table, I'm looking forward to use it.

[Hipoo]

(06-30-2016, 01:33 PM)Aghannor Wrote: Where did you get the formulas for determining how each of the various upgrades affects DPS? Specifically, whether they're additive or multiplicative with each other?
He probably copied the dps formula from Blackie.

In fact, you'd be better off ignoring it entirely and spending those PP on strengthening instead, which gives you 3 times the damage boost for far less PP.
If you try to use the analyzer, it shows strengthening is the best upgrade to start with (damage/P.P. spent wise), and it is.

Likewise, your sheet has final blow act as a 20% multiplier (multiplying by 1+B14/5).  But why divide by 5, when it's supposed to activate at under 25% hp? If we're assuming that all enemy HPs are equally likely, shouldn't it be 1/4 rather than 1/5?  And if we're not assuming that all enemy HPs are equally likely, how do we end up at 1/5 expected activation?
I didn't understand the final blow multiplier being divided by 5 either, but that keeps my pet dps way closer to "/cmd comparedps" than if it was divided by 4.

Comparing the pet dps% from sheets vs "/cmd comparedps" still gives me a higher % in sheets (not by much, though). (48.88% vs 49.54%)

(06-30-2016, 05:37 PM)Tezexlo Wrote: Critdmg is bugged, you actually make less(!) dmg atm. And its dmg*(1+critmulti+0,2). So standard critmulti without upgrades is 20%.
The critical bug is weird. It does less damage but if you check your pet dps% (in game) before and after you upgrade a critical chance node, you will see it actually increases dps. So it could be a visual bug, who knows.

To use this sheet in a google doc, download the file and then import->upload. Ifna formula doesn't work though, so you'll need to fix. Another fix needed is log -> ln in the dps formula (or just change the multiplier to *(1+B10/2) as Tezexlo mentioned, it doesn't change much).

[IndigoFerret]

IFNA - It looks like this is an issue for many people, so I will be updating the spreadsheet with IFERROR which should solve this problem.

The formulas came from someone who posted a very simple set of calculations in Discord.  I don't remember who it was, but Hipoo suggests it was Blackie, and that is probably right.  I state in the Information tab that I found these formulas and that I am not sure they are 100% accurate.  I appreciate and help in correcting these formulas.

Aghannor said:
Specifically, the description of Auto Damage says it increases damage on pets gained from your auto damage.  Which suggests the upgrade value should add some percentage of your non-pet damage to your bet damage.  But in your sheet, it's treated as a basic multiplier of the pet base damage, which is additive with (and for all relevant purposes identical to) strengthening.  I'm specifically referring to Pet Upgrade Sum cell 26, where B25 (base pet damage with orb/ascension modifiers) is multiplied by 1+B11+B12-0.4*B13.  If that's correct, then the auto damage upgrade has nothing to do with your auto damage at all.  In fact, you'd be better off ignoring it entirely and spending those PP on strengthening instead, which gives you 3 times the damage boost for far less PP.

So, it seems that the wording for Auto Damage is not clear.  However, if you suggest that the Auto Damage should be a modifier to Base Damage % that shows on the Pet screen, that would substantially increase the Auto Damage value and make it better than any other upgrade.  This is not what most people are experiencing.  Absent clear evidence that the Auto Damage stat should affect the formula in some different way, I will leave it where it is.

Aghannor said:
Likewise, your sheet has final blow act as a 20% multiplier (multiplying by 1+B14/5).  But why divide by 5, when it's supposed to activate at under 25% hp? If we're assuming that all enemy HPs are equally likely, shouldn't it be 1/4 rather than 1/5?  And if we're not assuming that all enemy HPs are equally likely, how do we end up at 1/5 expected activation?

Tezexlo said:
And on the discussion about how we value the killing blow multiplier, I'm here more with IndigoFerret's 20% since the overkill only happens here. So the monster with ~4% hp left will be killed even without the killingblow dmg (depends on monster hp, for me its the 4% atm).

I felt that the 20% modifier made sense since it is not activated when monster HP is over 25%, but when it is activated, the monster is killed faster, thus the 25% of the time that is active ends up being less that 25% of the time.  20% seems reasonable and the difference between 20% and 25% wouldn't have a material affect on the value of this upgrade.

Tezexlo said:
First, it's not your idle autodmg, but your non idle autodmg, with this, the normal pet dmg fits with the shown value.
Killing Blow is actually halve the shown value (so misleading ingame).

Idle Auto Damage - It has been unclear to me which value to use coming from the Zilla IOU Complete MultiCalc.  Where can I find the value you expect for autodmg?
Final Blow - What is the justification for halving this percentage?

Tezexlo said:
So the formulas would be:dmg (so it acually shows the standard dmg the pet does):
=B24*(1+B17)*(1+B18)*(1+B19)*(1+B11+B12-(B13*0,4))
dps:
=(B25*(1-B7)+B25*(B7*(B8+0,2)))*B9*(1+B10/2)*(1+B14/5/2)

Making the changes to these formulas gives me much higher calculated Pet DPS values than before.  Maybe once I understand what you expect for autodmg, this might make more sense.

[Hipoo]

That autodamage = your regular auto hit (WITHOUT 10% idle bonus damage)

It's (base damage)*(event points+ascension)*(support)*(inferno level)*(orb)*(challenges)

[IndigoFerret]

(06-30-2016, 05:37 PM)Tezexlo Wrote: dps:
=(B25*(1-B7)+B25*(B7*(B8+0,2)))*B9*(1+B10/2)*(1+B14/5/2)

I am going to change that to:
dps:
=(B25*(1-B7)+B25*(B7*(1+B8+0,2)))*B9*(1+B10/2)*(1+B14/5/2)

But thanks for the help!

[Aghannor]

(07-01-2016, 05:20 AM)IndigoFerret Wrote: So, it seems that the wording for Auto Damage is not clear.  However, if you suggest that the Auto Damage should be a modifier to Base Damage % that shows on the Pet screen, that would substantially increase the Auto Damage value and make it better than any other upgrade.  This is not what most people are experiencing.  Absent clear evidence that the Auto Damage stat should affect the formula in some different way, I will leave it where it is.

I agree that that's not what Auto Damage does - but the description suggests that's what it should be doing.  I just meant to say I think the description in the game is extremely misleading if it's just a multiplier to the pet hit damage, and not a modifier to the base damage %.  

(07-01-2016, 05:20 AM)IndigoFerret Wrote: I felt that the 20% modifier made sense since it is not activated when monster HP is over 25%, but when it is activated, the monster is killed faster, thus the 25% of the time that is active ends up being less that 25% of the time.  20% seems reasonable and the difference between 20% and 25% wouldn't have a material affect on the value of this upgrade.

Ah, okay - so it's an approximation.  That's what I was getting at, really.  

Thanks for doing this - definitely interesting to poke around at what's under the hood, though I'm still curious as to how the formulas you started with were derived.

[IndigoFerret]

Version 0.1c is up now.

Fixed - Critical damage calculation corrected adding the base amount of 20%.
Fixed - IFNA function replaces with IFERROR
Fixed - Corrected the number of nodes for Attack Speed and Critical Chance
Fixed - Revised Pet Damage and Pet DPS calculations
Fixed - Pet Upgrade Input tab - rename Idle Auto Damage as Auto Damage
Added - Pet Upgrade Input tab - Active Idle DPS input item to allow for Total DPS calculation

[chaosguard]

Which numbers should we be using for the inputs with regards to the following:
Auto Damage
Active Idle DPS
Ascension Speed Upgrade <-- Is this Idle Speed or Pet Speed?

[IndigoFerret]

(07-02-2016, 08:42 AM)chaosguard Wrote: Which numbers should we be using for the inputs with regards to the following:
Auto Damage
Active Idle DPS
Ascension Speed Upgrade <-- Is this Idle Speed or Pet Speed?

I use Zilla's IOU MultiCalc (https://iourpg.com/forum/showthread.php?tid=1633) for the two damage numbers:

Auto Damage is from Main tab, Important Derived Stats, Active Hit.
Active Idle DPS is so named in the same section of the Main tab.

Ascension Speed Upgrade is from the game: Upgrades > Ascension > Pet Speed

[Kaerfameer]

since everyone was throwing out math i want to add something...

1 + 2 = 4