Aroddo

Those in my dreams.

I thought there were...

I was absolutely not interested in calculating "time to live". I'm likewise not interested in playing Demigod in Excel.

I linked the chart from somewhere else because it already clearly showed the mitigation-armor relationship.

The sheets are a bit truncated so they only show the bonus damage gained by using a selected choice of skills. More would be overkill and not part of the numbers I wished to present: Bonus melee damage due to armor reduction skills.

If you consider other derived values more important than meaningless bonus damage values - ok. That's your god-given right. That doesn't automatically make me wrong, though.

On a side note: I consider "time to live" a value of limited usefulness because the value has to be recalculated ever time anyone uses a skill or  an item. Even walking changes ttl while the %-damage bonus stays largely constant inside an exchange of blows.

That's interesting... but quite disappointing. Basically, this simply adds to the proof that the QoT doesn't scale very well as the game goes on.

A percentile effect would eliminate the ability to send towers (and players) into the negative, but it would allow for better scaling as the game continues. Of course, the downside to such a change is that the ability would be less effective early on.

I think another important thing to consider is actual % damage reduction, which is in the realm where transitive is.  I ran some numbers on this recently, so let me post them:

First of all, since it makes teh maths easy, let us pretend we are taking 100 damage, when at 0 mitigation. Consider what then is the value of 500 armor?

0 to 500 = 17% reduction (83.3 damage)

Mitigation goes up to a little less than 17%, and that is all we care about here.  But what happens when you get another 500 armor on top of that?

Well, at 500 armor we were taking 83.3 damage.
At 1000 armor we'll take 71.4. At a glance you may look at this as a change of 83.3 - 71.4 =  11.9%, but that isn't actually the right way of thinking about it. Instead you should look at what percentage of your initial damage you are taking after the increase in armor, that is , what is 71.4/83.3 = 85.7% of your original damage, so the reduction is 14.3% of your previous damage.

500 to 1000 = 14.3% (71.4 damage)

So yeah, this isn't horribly impressive right? It's still less isn't it?  This is true, but the math that goes into this means that the decrease in value of armor is not NEARLY as sharp as it looks.  For further proof, let us take a more bombastic example:

With 4000 armor, you would be taking 38.4 damage out of the original hundred.  Let us once again check to see what happens when we increase that by 500.

4500 armor = 35.7 damage.  Looks pretty small right?  Using the 'bad math' you would see your mitigation change from 61.6 to 64.3, and think 'who cares? 3% less damage isn't much'.

But once again you need to look at the actual percent of the initial damage your were taking, aka what is 35.7/38.4. It equals 93%, so the mitigation is still 7% more.  Over the course from change of 500 to 1000, all the way to 4000 to 4500 area, the value of 500 armor has been decreased only by about half. Not nearly as bad as the perceived change of decreasing to about one fifth.

For those who care, in order to maintain a constant value for armor, you actually have to start with an exponential function.  For example:

DF = DI*(.9^A)

DF = final damage

DI = initial Damage

A = Armor value

In the above scenario, each point of armor would decrease damage by exactly 10% compared to the value previous to it. DG uses a much finer grain on armor, so the .9 would be more like .009, but whatever.