Torchlight 2 Dodge Dexterity Diminishing Returns Guide
Torchlight 2 Dodge Dexterity Diminishing Returns Guide by Rokiyo
I’ve been doing a bit more hasty excel number crunching… This time on the theory that Dodge from Dexterity has diminishing returns. Turns out it does… But not in the way I expected! If any stats buffs would like to check my logic, it would be greatly appreciated.
First up is the dexterity formula itself, which seems to fit rather cleanly to a quadratic curve. I’ve managed to get pretty close to the ingame dodge numbers with the following:
Dodge = (-0.000002 * Dexterity)^2 + (0.002 * Dexterity) + 0.0002
Secondly, it should be noted that I’m treating damage mitigation from dodge as if it were a direct reduction in incoming damage. In theory, over an extremely large number of hits, 10% dodge and 10% physical damage reduction should end up providing the same benefit to your survivability. In reality, dodge causes incoming damage to become spikey, and an unlucky streak of hits could more easily overwhelm your health buffer.
EDIT: A couple of important things have been pointed out by others in this thread: Regardless of where you’re getting your dodge from, it caps out at 75%. For an outlander with the appropriate skills, this apparently means you can hit the 75% dodge cap at around 100 dexterity. It has also been noted that dexterity provides other benefits, and that meeting gear requirements might push you over this mark anyway.
Graphing the output of this formula gives us this:
Just as general consensus tells us, diminishing returns all the way. By this graph alone, it would be easy to believe that every point sunk into Dexterity would be progressively worse than the point spent before it.
The key thing to note however, is that a statistic like Dodge% has inherent escalating returns:
- At 1% dodge, 99 blows in 100 will land. At 2% dodge, 98 blows in 100 will land. Going from 1% to 2% results means you get hit 99% as often.
- At 98% dodge, 2 blows in 100 will land. At 99% dodge, 1 blow in 100 will land. Going from 98% to 99% means you get hit half as often. If you can survive for an average of 60 seconds without heals at 98% dodge, then you’ll survive for an average of 120 seconds without heals at 99% dodge.
EDIT: There is no way to go above 75% dodge. This is an example only.
To truly see the value of each point of dodge, we need to look at how it impacts your time to live. To determine this, I used a series of Excel formulas that should equate to:
TTL Extension = (1 / (1- (-0.000002 * Dexterity)^2 + (0.002 * Dexterity) + 0.0002)) – (1 / (1- (-0.000002 * (Dexterity – 1)) ^2 + (0.002 * (Dexterity – 1)) + 0.0002))
In general, each additional point of Dexterity granted a 0.2%-0.25% increase in Time-To-Live, as displayed below:
What this graph is showing is that Dodge from Dexterity has escalating returns right up until 212 Dexterity. Each additional point of Dexterity up to and including 212 increases your time-to-live by more than the previous point you spent. After this however, it begins to drop off pretty sharply. At your 358th point of Dexterity, you receive as much benefit as the 1st point granted, and beyond that you start to see less and less relative survivability granted by Dexterity.
As a matter of interest, applying these TTL bonuses to a base TLL of 60 seconds produces the follow TLL graph:
For someone who would survive for 60 seconds without heals at zero Dexterity, adding Dexterity continues to yield escalating returns right up until they hit the 99 seconds at 212 Dexterity. The returns remain above the baseline of 0.2% right up until 358 Dexterity, which yields a TTL of 140 seconds.