Wyatt Cheng on EHP and Life Return Stats
D3 developer Wyatt Cheng stopped by Reddit to give a few answers to the community, concerning EHP, Life Steal, Life Regen, Life on Hit, Life per Spirit Spent and Health Globes. Check out the original post as well, submitted by fjdkf, as it's been written rather well.
This is a great discussion. Good OP post and good analysis.
I agree that if we cut incoming damage and healing rates in half, players would run with half as much EHP, like you predict. So we can't just go in there, make that change, and expect everything to be okay. I like the way you broke things into EHP and Sustain, it's a good way to look at the problem.
I think that framework also hints at a potential solution too. We need to look at ways to test both your character's Sustainability as well as your EHP. Melee classes already experience a good deal of this, which is why they were the first to adopt Life On Hit when the game first came out. As a melee class you are constantly taking melee damage from enemies, so your sustain is tested every second.
Right now we don't have many ways to stress the EHP of the ranged classes without it feeling bursty. I think this is because the game has an plethora of "Do the right thing and take 0 damage, do the wrong thing and take 100K damage". Since the outcomes are so binary, the correct way to build a ranged class is to try to do the right thing as often as you can, and maybe build enough survivability so you can make a mistake once in a while. What I think we need more of is "Do the right thing and take 20K damage, do the wrong thing and take 80K damage". You're still trying to do the right thing, but you're still taking damage either way. Besides, Sanctuary is a dangerous, violent and hazardous place - some amount of damage comes with the territory. When I play, I want to feel stress on my EHP and my Sustainability while still feeling satisfied for making smart plays. I also want to value incremental survivability choices I make on my gear and my skill build.
P.S. LpSS scales with attack speed too since you get more attacks per second which translates into more spirit generation.
I'll assume that you are talking about damage taken post-mitigation. Assuming the person has some life regen of some sort, wouldn't both situations be almost the same in terms of binary life-or-death circumstances? Under current circumstances, the 20k life would be healed within a second on a high-end DH with gloom up. 80k damage would usually kill the hero outright. Isn't this the spike damage that you are trying to avoid?
Good question regarding the 100K vs. 20K/80K scenario. I think I skipped a step in my answer which was assuming the player had about 100K EHP.
Let's take it into some more concrete numbers. Let's say a Demon Hunter has 30K health and regenerates 20K health per second via Gloom or what have you. In that world, a 30K hit kills you instantly.
Let's take a mechanic like Mortar. Suppose at the MP level you're playing at, Mortars hit for 30K. As a Demon Hunter you have 2 choices, you can either plan on avoiding all Mortars perfectly, and then you don't die. Or you can gear for either 35K basically you're allowing yourself some "slush". How you gear depends on your confidence in your ability to avoid those Mortars, and how sensitive you are to death.
If you gear for 30K or less, then you are going to feel like the game is really cheap and unfair when you do die, because when those mortars hit you that you had planned on avoiding, you go from full to dead instantly. (Side note, this contributes to Vortex feeling cheap. It's not that Vortex kills you directly, but Vortex causes you to take damage from something you had previously planned on avoiding completely, because you know it kills you)
On the other hand, if you gear for 35K health, then the game went from threatening to super-easy-mode, because the mortars are flying at you once every 2 seconds, and you heal to full in 1.5 seconds, now the mortars never actually kill you. In fact, you might not even bother dodging them anymore.
Now enter the "half-damage and half-healing". Instead of healing for 20K health per second, I heal for 10K health per second. But instead of taking 30K damage, I take 15K damage. In the absence of any other changes, I think we'd see exactly what you predict. The game didn't become interesting, instead, I'm just going to gear with 15K health.
However, what if the way in which we cut Mortar damage wasn't just a strict 50% cut. What if we took the 30K damage and made it "0 damage if you avoid the Mortar by 20 yards, 5K damage if there is a Nearby Hit of about 8 yards, and 25K damage if you are hit directly".
Now as a player I can't reliably plan on avoiding all the Mortar hits. If I want to play super-duper safe I can avoid all the mortars, but I'm probably running a ton and doing almost no damage. Realistically I'm probably going to take 5K "incidental damage" constantly. This is going to eat into my 10K healing per second. If I take a 25K hit that leaves me at 5K health, I'm scared but not dead. Over the next 5 seconds I can't afford to take another direct hit, and can either play super conservatively to avoid even 5K hits, or I can play more aggressively and accept that my 10K regen per second is being partially counteracted by the 5K near-hits and just make a concerted effort to avoid any more direct hits until my health is back up to a comfortable range.
You're right this would be a significant shift in the environment, which is why you haven't seen such a change.
Note: I'm not actually saying we're going to change anything with Mortar. I'm just using it for illustrative purposes because I think it's a good example for people to wrap their minds around right now.
So basically, you're trying to do three things at once(correct me if I'm wrong). 1. Cut healing mechanisms by a large margin 2. Keep incoming damage high if you ignore mechanics, and thus more punishing due to point #1. 3. Make incoming damage far lower, yet consistent, for heros if they navigate the mechanics correctly.
Lifesteal is another issue, and is currently too good.
The broad statement of "cut incoming damage in half, and cut healing in half" is breaking down at the level of detail we're examining. Your 3 summary points are more accurate. The exact amounts would be much more finely calibrated.
I'll reiterate, there's no concrete path to get to a goal, but I think it's worth discussing whether that goal is a worthy one. Specifically, what advantages does it have for the fun-factor of the game if healing is less, ignored mechanics are still punishing, but there is far lower yet consistent damage for heroes who navigate the mechanics correctly (the three points you mentioned). Rather then asking whether 50% is the correct tuning number, what are the goals that correct tuning would achieve? I tend to think there would be numerous advantages but I'd like to hear what other's think.
I think buffing the numbers on things like LpSS and Life Regen is reasonable. I lean more towards LpSS.
Passive Life regen is good for allowing you to withstand some amount of damage over time, but when things go wrong and I'm close to death then my response is to back off and run in circles or run away until my regen has time to kick in. If I become overly reliant on this, then I may find myself running away a lot. This is not universally a bad thing, but it's bad if it happens too much or it becomes the dominant way in which I play my Monk or Barbarian. (Ranged classes don't suffer from this as much because they have tools to continue doing damage while minimizing how much damage they take)
If we were to buff LpSS instead, then when I'm low on life I still want to play carefully and conservatively but I still want to engage monsters, generate spirit, and then spend it to heal myself. I think LpSS has the potential to be a great stat, but the current tuning numbers are too low to be competitive in today's environment. I think Life for Fury Spent could be tuned to be attractive as well. If I were low on Life (and healing were harder to come by then it is now), it would encourage me to do things that generated Fury and then spend that fury to heal myself.
One of the cases I think health globes break down the most right now is when I am fighting a large monster. Let's say I'm normally killing a monster every second, and getting a health globe every 5 seconds (every 5 monsters). The health globes are very comfortably healing me so I'm fine.
Then I come up to a boss, champion, rare, etc. and suddenly I'm health globe starved for 30 seconds. This is enough to kill me. Now I need other sources of healing from my gear or skills.
I don't think this is necessarily a problem, but it happens to a more extreme state right now then I think it needs to. Monsters currently have the ability to drop health globes incrementally as they take damage (ie. drop a globe at 50% health), but we could probably use that more, to try and remove those long periods against very high health targets where you are health globe starved.
Well if you can't depend on health globes for healing when you need them most, wouldn't it make sense, as a player, to instead build around one of the life regen affixes like LS or LoH instead? If LoH or lifesteal are strong enough to cover a 30-second gap in health globes, they'll be way more than enough to sustain yourself through the trash. It just seems that the inevitable outcome is that health globes become the secondary source of healing instead of the primary source.
Yes. I think your analysis is correct. I think your assessment is why healing from health globes feel marginalized in the current live game at high levels of play.
One possible solution would be if you weren't starved for 30 seconds on the elite. Like if the elite dropped a health globe every 20% health, then you'd get a health globe every 6 seconds while fighting this high health elite. (Source)
Well that's great. Life steal. Aka the only way you can make builds anyway you like is taking a hit to the face with a nerfbat? Really? How is it too good? I see DH's still die on MP10 even with shadow power up. even my pro geared 80k HP barb gets a few deaths in soft-core from time to time with my 6% Life steal.
Saying life steal is the problem is a little blown out. If healing get nerfed everyone is going to go back to the same 1 build just like back when the game came out to do inferno. Life steal is powerful if you are able to stack it. But look at what you pay to get them. Unless you are lucky and find a 3% life steal weapon with perfect stats your going to be paying 1.5B to get something decent with AP CRIT DMG SCOK and 1100+ DPS with Life steal. If you can pay for the power? or work for it? What's the problem??
Isn't getting powerful and making hard stuff easy the full point for MP levels? I don't get what you guys think is the problem. Stop trying to fix something that's not broken.
You completely missed the point. It's not about the difficulty, it's about gameplay design. They were talking about making things less bursty, so there are grey areas between "I'm at full health and staying there indefinitely" or "oops I just died in half a second".
the whole problem stemmed from the type of playstyle implemented and trying to make game-speed/killing-speed match the fast pace of previos generations of Diablo.
they just implemented FAR too much incoming damage to the game in order to keep that speed up. its too late now to make any decent combat style work. imo leave LS alone and FIX THE $#@%#&^% LOOT BEFORE ANYTHING ELSE
imo leave LS alone and FIX THE $#@%#&^% LOOT BEFORE ANYTHING ELSE
They are, calm your farm!
Rollback Post to RevisionRollBack
Bashiok - Blizzard Representative - 08/01/2011 -"So how many skill combinations are there now? Well taking into account 6 active skills, all the rune combinations, and 3 passives we currently expect each class to have roughly 2,285,814,795,264 different build combinations."
"Hey, I thought you'd like the witty irony of grub-on-glowie violence!"
You completely missed the point. It's not about the difficulty, it's about gameplay design. They were talking about making things less bursty, so there are grey areas between "I'm at full health and staying there indefinitely" or "oops I just died in half a second".
What point? He's talking a bunch of game design philosophical garbage that actually means nothing in the context of the current game. If you are taking too much damage then you have the option to lower the MP level, change skills, or gear for more EHP. We're always going to be balancing between DPS, EHP, and regen, so if they just nerf regen then we will gear for more EHP to make up for the reduced regen. In order to make it impossible to outgear smart gameplay they would have to remove defensive stats or nerf them to the ground to the point where they didn't matter at all and remove MP levels since nobody will be able to survive them.
How do you balance fighting 1 mob might vs fighting 10? When 1 phase beast hits you it hurts, when a pack of 4 (not sure if they can be horde affix for >4, don't think I've ever seen it in over 1k hours) all hit you better have a shit ton of EHP and if one of the faerie things hit you with theire armor debuff then god help you.
There are also so many affix combos where it is easy to do the right thing for each individual affix but put them together and they prevent you from doing what you want. How do you balance being frozen/jailed/walled in a tight space filled with lasers?
Would they completely kill CM spec too? If not then CM wizards would be even more mandatory to freeze everything so that you can live easily and dps more efficiently in groups.
They might retune regen numbers to bring them more in line with each other on current gear but they can't do anything to drastically change gameplay without making many other significant changes. ie. maybe in the expansion...
Isn't getting powerful and making hard stuff easy the full point for MP levels? I don't get what you guys think is the problem. Stop trying to fix something that's not broken.
Dude, next thing I'm expecting is the news that Jay Wilson is BACK to Diablo 3, since he ruined Project Titan.
That BS design sound SO MUCH like him.
this discussion depends upon what you mean by "Bursty." If it means your character dies like being squashed like a grape to solve this you could add a pre-death timer which could start out at one second for a non-buffed player. The pre-death timer would show that your character is mortally wounded and will die shortly but can still do certain is special pre-death skills which can only be done during the pre-death timer. Those skills would have to be added to the game in a patch. Some ideas might be Death Explosion, Healing Explosion for your party, Disease Cloud, or Soul Fire. In this case the character knows that they are going to die however they have a chance to do something to improve the chances of their party to succeed. Also if you mean by "Bursty" the rapid fluctuation of health during melee combat there might be a way to create a delay factor in both the damage and the health regeneration from potions and health quotes and anything that gives health like life steel. This delay factor could start out at 0.1 seconds forward on buffed character and could possibly be buffed reduce delay for increasing health and increased delay for decreasing health.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
D3 developer Wyatt Cheng stopped by Reddit to give a few answers to the community, concerning EHP, Life Steal, Life Regen, Life on Hit, Life per Spirit Spent and Health Globes. Check out the original post as well, submitted by fjdkf, as it's been written rather well.
Originally Posted by Blizzard (Blue Tracker / Official Forums)
I agree that if we cut incoming damage and healing rates in half, players would run with half as much EHP, like you predict. So we can't just go in there, make that change, and expect everything to be okay. I like the way you broke things into EHP and Sustain, it's a good way to look at the problem.
I think that framework also hints at a potential solution too. We need to look at ways to test both your character's Sustainability as well as your EHP. Melee classes already experience a good deal of this, which is why they were the first to adopt Life On Hit when the game first came out. As a melee class you are constantly taking melee damage from enemies, so your sustain is tested every second.
Right now we don't have many ways to stress the EHP of the ranged classes without it feeling bursty. I think this is because the game has an plethora of "Do the right thing and take 0 damage, do the wrong thing and take 100K damage". Since the outcomes are so binary, the correct way to build a ranged class is to try to do the right thing as often as you can, and maybe build enough survivability so you can make a mistake once in a while. What I think we need more of is "Do the right thing and take 20K damage, do the wrong thing and take 80K damage". You're still trying to do the right thing, but you're still taking damage either way. Besides, Sanctuary is a dangerous, violent and hazardous place - some amount of damage comes with the territory. When I play, I want to feel stress on my EHP and my Sustainability while still feeling satisfied for making smart plays. I also want to value incremental survivability choices I make on my gear and my skill build.
P.S. LpSS scales with attack speed too since you get more attacks per second which translates into more spirit generation.
I'll assume that you are talking about damage taken post-mitigation. Assuming the person has some life regen of some sort, wouldn't both situations be almost the same in terms of binary life-or-death circumstances? Under current circumstances, the 20k life would be healed within a second on a high-end DH with gloom up. 80k damage would usually kill the hero outright. Isn't this the spike damage that you are trying to avoid?
Good question regarding the 100K vs. 20K/80K scenario. I think I skipped a step in my answer which was assuming the player had about 100K EHP.
Let's take it into some more concrete numbers. Let's say a Demon Hunter has 30K health and regenerates 20K health per second via Gloom or what have you. In that world, a 30K hit kills you instantly.
Let's take a mechanic like Mortar. Suppose at the MP level you're playing at, Mortars hit for 30K. As a Demon Hunter you have 2 choices, you can either plan on avoiding all Mortars perfectly, and then you don't die. Or you can gear for either 35K basically you're allowing yourself some "slush". How you gear depends on your confidence in your ability to avoid those Mortars, and how sensitive you are to death.
If you gear for 30K or less, then you are going to feel like the game is really cheap and unfair when you do die, because when those mortars hit you that you had planned on avoiding, you go from full to dead instantly. (Side note, this contributes to Vortex feeling cheap. It's not that Vortex kills you directly, but Vortex causes you to take damage from something you had previously planned on avoiding completely, because you know it kills you)
On the other hand, if you gear for 35K health, then the game went from threatening to super-easy-mode, because the mortars are flying at you once every 2 seconds, and you heal to full in 1.5 seconds, now the mortars never actually kill you. In fact, you might not even bother dodging them anymore.
Now enter the "half-damage and half-healing". Instead of healing for 20K health per second, I heal for 10K health per second. But instead of taking 30K damage, I take 15K damage. In the absence of any other changes, I think we'd see exactly what you predict. The game didn't become interesting, instead, I'm just going to gear with 15K health.
However, what if the way in which we cut Mortar damage wasn't just a strict 50% cut. What if we took the 30K damage and made it "0 damage if you avoid the Mortar by 20 yards, 5K damage if there is a Nearby Hit of about 8 yards, and 25K damage if you are hit directly".
Now as a player I can't reliably plan on avoiding all the Mortar hits. If I want to play super-duper safe I can avoid all the mortars, but I'm probably running a ton and doing almost no damage. Realistically I'm probably going to take 5K "incidental damage" constantly. This is going to eat into my 10K healing per second. If I take a 25K hit that leaves me at 5K health, I'm scared but not dead. Over the next 5 seconds I can't afford to take another direct hit, and can either play super conservatively to avoid even 5K hits, or I can play more aggressively and accept that my 10K regen per second is being partially counteracted by the 5K near-hits and just make a concerted effort to avoid any more direct hits until my health is back up to a comfortable range.
You're right this would be a significant shift in the environment, which is why you haven't seen such a change.
Note: I'm not actually saying we're going to change anything with Mortar. I'm just using it for illustrative purposes because I think it's a good example for people to wrap their minds around right now.
So basically, you're trying to do three things at once(correct me if I'm wrong).
1. Cut healing mechanisms by a large margin
2. Keep incoming damage high if you ignore mechanics, and thus more punishing due to point #1.
3. Make incoming damage far lower, yet consistent, for heros if they navigate the mechanics correctly.
Lifesteal is another issue, and is currently too good.
The broad statement of "cut incoming damage in half, and cut healing in half" is breaking down at the level of detail we're examining. Your 3 summary points are more accurate. The exact amounts would be much more finely calibrated.
I'll reiterate, there's no concrete path to get to a goal, but I think it's worth discussing whether that goal is a worthy one. Specifically, what advantages does it have for the fun-factor of the game if healing is less, ignored mechanics are still punishing, but there is far lower yet consistent damage for heroes who navigate the mechanics correctly (the three points you mentioned). Rather then asking whether 50% is the correct tuning number, what are the goals that correct tuning would achieve? I tend to think there would be numerous advantages but I'd like to hear what other's think.
I think buffing the numbers on things like LpSS and Life Regen is reasonable. I lean more towards LpSS.
Passive Life regen is good for allowing you to withstand some amount of damage over time, but when things go wrong and I'm close to death then my response is to back off and run in circles or run away until my regen has time to kick in. If I become overly reliant on this, then I may find myself running away a lot. This is not universally a bad thing, but it's bad if it happens too much or it becomes the dominant way in which I play my Monk or Barbarian. (Ranged classes don't suffer from this as much because they have tools to continue doing damage while minimizing how much damage they take)
If we were to buff LpSS instead, then when I'm low on life I still want to play carefully and conservatively but I still want to engage monsters, generate spirit, and then spend it to heal myself. I think LpSS has the potential to be a great stat, but the current tuning numbers are too low to be competitive in today's environment. I think Life for Fury Spent could be tuned to be attractive as well. If I were low on Life (and healing were harder to come by then it is now), it would encourage me to do things that generated Fury and then spend that fury to heal myself.
One of the cases I think health globes break down the most right now is when I am fighting a large monster. Let's say I'm normally killing a monster every second, and getting a health globe every 5 seconds (every 5 monsters). The health globes are very comfortably healing me so I'm fine.
Then I come up to a boss, champion, rare, etc. and suddenly I'm health globe starved for 30 seconds. This is enough to kill me. Now I need other sources of healing from my gear or skills.
I don't think this is necessarily a problem, but it happens to a more extreme state right now then I think it needs to. Monsters currently have the ability to drop health globes incrementally as they take damage (ie. drop a globe at 50% health), but we could probably use that more, to try and remove those long periods against very high health targets where you are health globe starved.
Well if you can't depend on health globes for healing when you need them most, wouldn't it make sense, as a player, to instead build around one of the life regen affixes like LS or LoH instead? If LoH or lifesteal are strong enough to cover a 30-second gap in health globes, they'll be way more than enough to sustain yourself through the trash.
It just seems that the inevitable outcome is that health globes become the secondary source of healing instead of the primary source.
Yes. I think your analysis is correct. I think your assessment is why healing from health globes feel marginalized in the current live game at high levels of play.
One possible solution would be if you weren't starved for 30 seconds on the elite. Like if the elite dropped a health globe every 20% health, then you'd get a health globe every 6 seconds while fighting this high health elite. (Source)
Ha. Bagstone.
Saying life steal is the problem is a little blown out. If healing get nerfed everyone is going to go back to the same 1 build just like back when the game came out to do inferno. Life steal is powerful if you are able to stack it. But look at what you pay to get them. Unless you are lucky and find a 3% life steal weapon with perfect stats your going to be paying 1.5B to get something decent with AP CRIT DMG SCOK and 1100+ DPS with Life steal. If you can pay for the power? or work for it? What's the problem??
Isn't getting powerful and making hard stuff easy the full point for MP levels? I don't get what you guys think is the problem. Stop trying to fix something that's not broken.
they just implemented FAR too much incoming damage to the game in order to keep that speed up. its too late now to make any decent combat style work. imo leave LS alone and FIX THE $#@%#&^% LOOT BEFORE ANYTHING ELSE
Bashiok - Blizzard Representative - 08/01/2011 -"So how many skill combinations are there now? Well taking into account 6 active skills, all the rune combinations, and 3 passives we currently expect each class to have roughly 2,285,814,795,264 different build combinations."
"Hey, I thought you'd like the witty irony of grub-on-glowie violence!"
How do you balance fighting 1 mob might vs fighting 10? When 1 phase beast hits you it hurts, when a pack of 4 (not sure if they can be horde affix for >4, don't think I've ever seen it in over 1k hours) all hit you better have a shit ton of EHP and if one of the faerie things hit you with theire armor debuff then god help you.
There are also so many affix combos where it is easy to do the right thing for each individual affix but put them together and they prevent you from doing what you want. How do you balance being frozen/jailed/walled in a tight space filled with lasers?
Would they completely kill CM spec too? If not then CM wizards would be even more mandatory to freeze everything so that you can live easily and dps more efficiently in groups.
They might retune regen numbers to bring them more in line with each other on current gear but they can't do anything to drastically change gameplay without making many other significant changes. ie. maybe in the expansion...
That BS design sound SO MUCH like him.