Xypherous on Health vs Resistances and more.

Continue reading for some verbose and brainy posts, from Xypherous, regarding the popularity of health items vs resistance items in the post S3 rework, how this has affected AP Burst mages, and more!

( Warning: We all love Xypherous but BOY can he talk! I've tried to sort through his tidal wave of 40+ posts this evening and pluck the most interesting or relative out of it. Sorry if things are messy! )

Health Vs Resistances
Xypherous has been prowling the forums posting heavily on a thread initially inquiring if it was a poor design choice to have resistances be a less effective stat than health.

In his first reply, Xypherous shared his thoughts on health as the most prominent defensive stat and why he prefers it to resistances - armor or magic resistance.

"I'm actually a fan of health being the primary defensive stat - mostly because health has severely intrinsic features that resistances don't, that lets it be more easily controlled.

For example:

1. Health doesn't reduce the effectiveness of lifesteal type effects - or the secondary effects of an opponent's abilities.

2. Raw health isn't multiplicative with other healing or regeneration effects - allowing it to be effective without drastically multiplying in value with other effects (self heals, shields and the like.)

3. Health is a more obvious indicator of high durability than resistances - just due to how basic UIs work.

4. Because of high health per level gains, the relative effectiveness of health is far more temporary than resistances - This is just a natural fallout of our stat per level gains, not an intrinsic advantage of health.

For example, let's take the following case here:

You have a Bruiser with 1000 health, 300 armor fighting an AD carry, versus a Bruiser with 2000 Health and 100 armor.

1. % Armor Penetration is more effective in the first case than the second as % armor penetration values go up - values start to skew drastically as % armor penetration goes beyond ~20% or so, relatively neutral for low values of % armor penetration.
2. However, Life Steal - is far more effective in the second case - whereas it is completely destroyed in the first case.
3. The Bruiser's life steal / shield statistics are also neutral in the second case - whereas they are multiplied in the first case.

Thus, the AD carry or any damage dealer actually has multiple paths to deal with a high health target than they did - because the other defensive aspects of the bruisers *aren't* being multiplied or their own secondary defensive statistic (lifesteal) isn't being diminished as well.

The flip side of this is that Flat Armor Penetration, the primary statistic of assassins, can be weaker and still be as effective versus Bruisers because of their ability to negate a far larger portion of armor.

Naturally, this has the flipside of making certain resistance focused tanks weaker - but that's something I feel like we should solve on resistance based tanks. So please, if there's a particular resistance based tank that feels weak, let us know.

Let's take another example, in lane:

Let's say I'm fighting a primary physical damage dealer, whose probably itemized some sort of flat penetration. I have something like 800 health and 50 armor. How much gold do I need to spend to double my life-span?

1. Well, I could buy 800 health and have another 1200 effective health! Success!
2. Or I build 150 armor and have another 1200 effective health!

However, let's take a look at how things change when the game continues - as the game continues, the relative contribution of that armor to my health levels doesn't decay mostly because our game awards much more health per level than armor per level. In fact, by the time I'm level 8 or so, I'll probably have 1200 base health - and thus that 800 health is only increasing my effective at a lower ratio.

However, that 150 armor advantage is "still there".

Note this isn't an intrinsic advantage of health versus armor - just a side effect of how we award stats per level - but it does mean that the relative advantage of health in this case is temporary - rather than one that persists through the game."

Returning to the resistances side of things, he pointed out that in some ways that the value of resistances are "hidden".

"The strength of resistance is hidden in a lot of myriad ways. While they only apply to "half" the damage you're taking, that's not the whole picture. What resistances actually do is magnify the effect of your other self-healing, resistance and regeneration effects, while also mitigating the drain effects and % health effects of your opponents.

I agree that on the surface, it feels like resistances should feel stronger against a single type of damage but the power of the combined effect of resistances with other effects means that resistances can't be as strong innately on the 1:1 comparison."

When asked what sort of counter play is available for high hp and low resist builds, in comparison to just building armor or magic pen to counter resistances based builds, Xypherous responded:

"High Health / Low Resist has tons of counterplay - in fact, quite a good deal more than resistances in most cases. 

The most basic form of counterplay versus health is life steal. Innately, the higher health and lower resistances someone has, the more effective my life steal statistic becomes. It's why having 4000 health doesn't really help you against Warwick - but having 100 MR does. Your health pool is becoming *their* health pool.

Furthermore, as I've explained earlier, resistances bolster your own natural defenses - whereas health doesn't. Shields / Self-heals / Regeneration effects become far out of line when resistances are strong - because resistance magnifies these - these effects typically have very little counterplay of their own without further stacked effects."

Of course the mention of health brought up Warmog's Armor, the heaviest HP item in the game. Xypherous, however, had a bit of a surprising response to the claims that Warmog's Armor is problematic.

"Not actually certain it's as big of a problem as the current environment thinks it is. Warmog's isn't really efficient in terms of gaining raw health per gold if you do the math on it. Buying any combination of ruby crystal items, for example, will net you more health than gold than buying Warmog's. 

At the moment, it looks like it's more an environmental thing than an actual cost to gold effectiveness ratio. 

Although, there's a couple of things tuned around old armor values that we'll probably need to look at in the future:

1. The tower damage / armor penetration formulae were based off old health / armor values. We'll probably have to revisit that soonish so that towers have more damage and less armor penetration.

2. We'll have to do a pass on character that used to scale primarily off health because their tuning is probably off - Probably not a huge surprise that Shen or Olaf are pretty decent now."

Addressing a summoner lamenting about the "lack of health options" besides Warmog's, he replied:

"While Warmog's may be the defining health item - Bulwark / Randuin / Locket of the Iron Solari / Sunfire Cape all over above 400 health in a single slot. Roughly half the armor items are as much health items as armor items at their core. Randuin's is the only item in there that provides more than 50+ armor in a single slot - so most of those are items are actually core health items."

However, he did admit that Warmog's is a bit cheap right now and mentioned it's price is going up a little.

"Yeah, on PBE the cost of Warmog's is inching up slightly by 180 gold or so.

We'll keep adjusting the cost until we feel like it's in a decent spot. Like I said, I'm not saying the current tuning is perfect and that we've definitely overshot this patch - but we'll keep tuning it until it is.'

Xypherous also spoke a bit about the problematic versatility Giant's Belt after a summoner pointed out that it's perhaps currently TOO safe of a purchase for top lane.

"The strength of the various health components are actually on our radar - I agree with you that it's probably too safe of an item when unfinished right now. It's something we're going to address next patch and then take another look at afterwards to see if it needs further work.

So don't get me wrong, I'm not saying health tuning is perfect right now or that we aren't making gold : stat efficiency so let us know of stuff like this!"

The effect of High HP on Burst Mages 

I'll admit it, I stuck this picture of LeBlanc in here not only to signify the topic change to mages but also to give your poor eyes a break from that massive wall of text.

Switching gears to high health has affected some of the types of champions, Xypherous spoke to how AP burst mages have ended up now that health is a more viable itemization option.

"Yes, burst mages do get hit when health becomes the dominant statistic - mostly because, it turns out that health is a pretty good substitute for magic resistance when the going gets rough.

There's various implications of that - but that's why you see things like Liandry or DFG existing. We're doing an AP pass next patch to help mages out - and after that we can start looking at specific characters if they still feel weak."

To continue on that point, he also responded to a summoner who pointed out that Liandry's Torment isn't really the best item to build on burst caster.

"It is a waste on a burst caster, I agree - but it gives burst casters a way to itemize into more of a sustained damage caster, depending on what they are fighting.
Brand is an example of this, for example. While Brand is typically a burst mage, Liandry + Omnomnomnomicon can turn him into an effective sustained mage.

DFG attacks it by adding to more burst - while Liandry tries to attack it by offering an different playstyle out. We'll see how it goes."

He also tossed out a quip about Liandry's Torment and why, when designing the item, he chose to go with current health % rather than maximum health %.

"I chose % current because its a marginally safer effect that lets us be a little more aggressive with the numbers on the item. It tends to also create a better playstyle dynamic because it's not entirely self-contained - you need % current health plus something else in order to murder someone - whereas % max health only needs more of itself, for example."

Change Runes to help balance HP vs Resist itemization?
Within the Health vs Resistances thread, one summoner pointed out that nearly everyone runs armor and MR runes, which, he claims, is "skewing the cost effectiveness of health by making every character begin with higher resistances." In response, Xypherous spoke briefly about rune sets and why Riot is hesitant to mess with them as an alternative to messing with items so much.

"Yes, this is true - however, runes are something we're much much more careful about touching because players have significant investment into their rune sets as is. If we have to do rune adjustments we will - but it's something that we want to traditionally be very cautious about as there's a lot more potential for disruption here than anywhere else."

Misc Tidbits about Specific Items
Moving on to items, Xypherous had a few things to say about specifics items.

Regarding Ohmwrecker, he mentioned it's current state as being week but something that needs to be handled carefully.

"Ohmwrecker was an item that was probably most polarizing in community testing / feedback - so we decided to play it a little more cautiously. I don't disagree that it's weak right now - but we want don't really want to see the item become "core" on anything, it should sit in a niche spot if possible.

It's definitely somewhat weak - we'll address this when we get the chance but community feedback was very polarized around the item when we first introduced it on PTR."

He also shared his thoughts on the Negatron Cloak and it's build paths.

"... I do suspect that Negatron is in an odd spot - it feels like it should be cheaper and give less MR so that its upgrades could also be slightly cheaper or give more MR. The build path for the MR items providing most of the MR you need seems to be strangling how good the final MR item can be sometimes. Not sure.'

PHEW! That was quite the wall of text. Congratulations on making it through.