IO Purchasing Question

For me...
I just like knowing that I get hit due to the 5% and not by the .4%




ahhhhhh earth quick!!!!

Thank you for the crystal clear explanation Werner. I didn't fully appreciate the difference between being almost to the soft cap and all-the-way to the soft cap.

And now I feel like I have to explain mitigation and survivability.

A while ago, the forum came to a consensus on at least the meaning of "mitigation". That's not the same as saying everyone agrees, more like saying "the handful of people participating in a thread about mitigation a long time ago agreed".

We decided that mitigation was the percentage of damage prevented that would have hit you if you'd had 0% defense and 0% resistance. So if enemies are putting out 500 DPS, no mitigation STILL has you taking only 250 DPS, because the enemies only have 50% to hit. Now, you might take a different amount from something other than even level minions. It doesn't actually end up mattering, because the percentage mitigation calculated this way works out the same regardless.

In a formula, mitigation is this (well, keeping in mind that 45% defense is the "cap"):

mitigation = 100% - (100%-2*defense)*(100%-resistance)

So if you have 45% defense and no resistance, you have 100% - (100%-2*45%)*(100%-0%) = 100%-10% = 90% mitigation
If you have 60% resistance and no defense, you have 100% - (100%-2*0%)*(100%-60%) = 100%-40% = 60% mitigation
If you have 45% defense AND 60% resistance, you have 100% - (100%-2*45%)*(100%-60%) = 100%-10%*40% = 100%-4% = 96% mitigation

Given the definition, we can revisit the first 0.4% vs. last 0.4% discussion, and we'll see that BunnyAnomaly is completely correct, that the mitigation of both is equal. I did not state otherwise, and did not disagree with this point.

0% defense = 100% - (100%-2*0%)*(100%-0%) = 0% mitigation
0.4% defense = 100% - (100%-2*0.4%)*(100%-0%) = 0.8% mitigation
difference = 0.8% mitigation

44.6% defense = 100% - (100%-2*44.6%)*(100%-0%) = 89.2% mitigation
45% defense = 100% - (100%-2*45%)*(100%-0%) = 90% mitigation
difference = 0.8% mitigation

So the same, right? Yes. The mitigation is the same.

But we have to introduce ANOTHER concept - survivability. There is less consensus on the definition of survivability, as well there should be. But a standard approach (and the one I use despite its weaknesses) is to treat survivability as the amount of incoming damage that you can on average survive indefinitely, ignoring such details as the fact that a random walk can still kill you. Survivability is then defined like this:

survivability = net healing and regeneration / (100% - mitigation)

So if you heal 25 hit points per second, and you have 90% mitigation (such as from 45% defense and no resistance), then you have 25 / (100%-90%) = 25 / 10% = 250 survivability. This corresponds in a fairly loose sense to how much incoming damage you can survive. As an aside, the "Wernerscore" from my survivability spreadsheet is essentially two times the survivability under this definition. Being used for relative comparisons only, as it should, the scale hardly matters and we could just as easily multiply by five or divide by PI.

Now we can reexamine the difference that 0.4% defense makes to our survivability. Again, scale doesn't matter, so for comparison, we can assume any amount of healing and regeneration. Let's stick with 25 hit points per second.

0% defense = 0% mitigation = 25 / (100% - 0%) = 25 survivability
0.4% defense = 0.8% mitigation = 25 / (100% - 0.8%) = 25.2 survivability
A 0.2 difference in survivability, very roughly speaking another 0.2 DPS that you can survive.

44.6% defense = 89.2% mitigation = 25 / (100% - 89.2%) = 231.5 survivability
45% defense = 90% mitigation = 25 / (100% - 90%) = 250 survivability
An 18.5 difference in survivability, very roughly speaking another 18.5 DPS that you can survive. Not a big deal, but maybe you can add another minion. (Edit: And of course it would be a much higher number if we also had resistance and better regeneration and healing. I'd put the RWZ challenge at about 500 survivability. I'd put no temp no insp AV soloing at about 1000 survivability. I believe that Scrappers max out around 2500 survivability outside of special cases, like a Dark Melee/Invuln facing purely smashing/lethal enemies.)

Contrary to what I was saying earlier, the last 0.4% makes very roughly 100x more difference than the first 0.4%. This at first took me by surprise, and had me double-checking math. But in my earlier argument, I was assuming the difficulty level stayed constant for each character. Here, each character essentially cranks up the difficulty to match their new survivability, and the high defense toon can crank it up MUCH more for a given actually-hitting-him DPS difference, because all of the additional damage is being heavily mitigated.

Or maybe I've made a horrible mistake, since it does sound excessive. If so, it'll have to wait, because I'm late to dinner at a friend's place. Or maybe someone else can point out where I'm wrong.

Edit: Ah, dinner's fallen apart. Still not sure about what I'm saying, but here's the chart I'm generating assuming 0% resistance and 10 hit point per second regeneration. It does show how the last 1% of defense gives a couple orders of magnitude more survivability than the first. Interesting.

OK, does this correspond to our normal understanding? We normally say that someone at 40% defense is taking twice as much damage as someone at 45% defense. This is reflected in having half the survivability score. Check. 30% should be another double, another half. Check. 10% should be another double, another half. Check. Yeah, looks right. So yeah, a couple orders of magnitude difference, not just one.

Edit2: No, I'm back to saying just one order of magnitude. I think what's important isn't the raw survivability increase, but the PERCENT survivability increase. I've added that as another column below, which gives the single order of magnitude difference between the first percent and the last percent that I think is most meaningful in terms of how it will actually feel in game.

Defense Mitigation Survivability Increment % Increase
0% 0% 10.0
1% 2% 10.2 0.2 2.0%
2% 4% 10.4 0.2 2.1%
3% 6% 10.6 0.2 2.1%
4% 8% 10.9 0.2 2.2%
5% 10% 11.1 0.2 2.2%
6% 12% 11.4 0.3 2.3%
7% 14% 11.6 0.3 2.3%
8% 16% 11.9 0.3 2.4%
9% 18% 12.2 0.3 2.4%
10% 20% 12.5 0.3 2.5%
11% 22% 12.8 0.3 2.6%
12% 24% 13.2 0.3 2.6%
13% 26% 13.5 0.4 2.7%
14% 28% 13.9 0.4 2.8%
15% 30% 14.3 0.4 2.9%
16% 32% 14.7 0.4 2.9%
17% 34% 15.2 0.4 3.0%
18% 36% 15.6 0.5 3.1%
19% 38% 16.1 0.5 3.2%
20% 40% 16.7 0.5 3.3%
21% 42% 17.2 0.6 3.4%
22% 44% 17.9 0.6 3.6%
23% 46% 18.5 0.7 3.7%
24% 48% 19.2 0.7 3.8%
25% 50% 20.0 0.8 4.0%
26% 52% 20.8 0.8 4.2%
27% 54% 21.7 0.9 4.3%
28% 56% 22.7 1.0 4.5%
29% 58% 23.8 1.1 4.8%
30% 60% 25.0 1.2 5.0%
31% 62% 26.3 1.3 5.3%
32% 64% 27.8 1.5 5.6%
33% 66% 29.4 1.6 5.9%
34% 68% 31.3 1.8 6.3%
35% 70% 33.3 2.1 6.7%
36% 72% 35.7 2.4 7.1%
37% 74% 38.5 2.7 7.7%
38% 76% 41.7 3.2 8.3%
39% 78% 45.5 3.8 9.1%
40% 80% 50.0 4.5 10.0%
41% 82% 55.6 5.6 11.1%
42% 84% 62.5 6.9 12.5%
43% 86% 71.4 8.9 14.3%
44% 88% 83.3 11.9 16.7%
45% 90% 100.0 16.7 20.0%

Wow, this thread took a direction I didn't expect. But hmm, what I've actually taken from this is something else. I've tried using Mid's to get my DM/SD up to snuff, and they are clawing at 43-44. Sure, losing 0.1 percent to melee defense doesn't mean much, but if 5 powers lose 0.1, then you're losing 0.5, which is half a percent when you are already scraping to cap yourself.

Or something. Maybe I'm reading it wrong. But I can attest that 40-42 is not enough for me in some occasions already (I have ~30 def, and hit a luck for 12.5, and sometimes need to hit a second to get unhittable enough not to die in larger packs, like accidental double pulls)

Oh, that's hardly a negative comment. This is exactly the kind of information I wanted to know. There may of been some debate, but it was on one of the issues I was curious about, so all ill didn't come of it. I was already surprised at the feedback, since all I really wanted was a yes or no while rubbing my hero merits together impatiently

BunnyAnomaly (if you're still reading), in response to the question "What IO bonuses do you shoot for w/ Willpower?" in another thread, you said:

Quote:

Originally Posted by BunnyAnomaly Regen bonuses amount to very, very little. On my WP/ tank she gets barely 1hp/s off a set bonus for regen, and when she has 100hp/s base with just a single enemy, why would you build for more?

I assume you're not saying that regeneration is always useless, but rather that an additional 1 HP/S amounts to very little because you already have so much?

So if you regenerate 99 HP/S, adding 1 HP/S to get 100 HP/S amounts to very little. While if you only regenerate 10 HP/S, adding 1 HP/S to get 11 HP/S is more valuable. The value of regenerating another 1 HP/S of damage depends on how much you already regenerate.

And if you are taking 100 HP/S of damage at 0% defense, adding 0.5% defense to lower that to 99 HP/S amounts to very little. If you are at 44.5% defense, you are taking 11 HP/S of damage, and adding 0.5% defense to lower that to 10 HP/S is more valuable. The value of preventing another 1 HP/S of damage depends on how much you already prevent.

Can you see the equivalence of these two points? The first is how I understood the one you made in the other thread. The second is the one I made in this thread, which you referred to as misleading and intellectually dishonest.

And as a similar line of reasoning that may be more clear and concise than any I have used so far: If you are taking 100 DPS at 0% defense, adding 0.5% defense lowers that to 99 DPS. If you are taking 100 DPS at 44.5% defense, adding 0.5% defense lowers that to 91 DPS. Preventing 9 DPS is more valuable than preventing 1 DPS.

I wish you would quote my entire post where I state the order of preferences I have on a particular powerset, because a single line taken out of context of what is an 'order of importance' isn't fair.

Adding 1 point of hp regen is practically useless on any character. I could have likewise said "why build for regen when you are only at 10 hp/s" but that was a meaningless statement to contribute to the other thread which had nothing to do with adding regen to low regen characters.

That above paragraph has the same reasoning as what I use here. 1 hp/s is to me quite meaningless. Just like a very tiny fraction of defence that will protect you from 1 hp/s of damage, which I consider likewise meaningless. I believe saying that increasing from 10 hps to 11 hp/s as a "10% increase and so is worth it" doesn't do any justice to what is really happening. Hence the term: lies, damn lies and statistics. Statistics aren't lies, but they are very easily written in such a way as to make something seem credible or meaningful.

The core of my quote is actually that you should build for defence and recharge on a Willpower character. Both give significantly better returns for investment.

Lastly, I do soft cap all my characters and lean towards a buffer zone for DDR.

Oh, absolutely, the +rech comes in handy with the primary. I just wanted to make sure I wasn't missing something on the WP side. As for the +HP, I'll have to take a hard look at that build in Mids; it's possible that I hit the HP cap via ancillary benefits from Def/Rech sets, but I'm always looking for oversights on my part.