How to Fix PDM

[Abaco 21]

There’s a lot of talk in my guild about PDM and how strong it is, and I’ve heard devs comment that their previous attempt at lowering PDM values wasn’t well received. I think this is because the current implementation of PDM skews to make higher values much more effective than a casual glance would expect. In this post, I will give an example of this dynamic and suggest a different approach that ONLY nerfs the problematic extremes.

First, let’s consider the addition of 5% more PDM when going from 0->5% pdm and going from 45->50% pdm.

For this example, the defender will have 10000 hp and be getting hit with 100 damage attacks.
With 0% damage reduction, the attacks hit for a full 100, so it would take (10000/100) = 100 attacks to kill the creature.
With 5% damage reduction, the attacks only hit for 95, so it takes (10000/95) = 105.26 attacks.
So 5% PDM makes you last 5% longer, that makes intuitive sense. Now, let’s look at 45%.
With 45% damage reduction, the attacks only hit for 55, so it takes 10000/55 = 181.18 attacks.
With 50% damage reduction, the attacks only hit for 50, so it takes 10000/50 = 200 attacks.

At the high levels, that extra 5% makes you last an extra 19 attacks! It’s almost 4 times as good as the first 5% were!

Here’s a chart measuring the Effective Health of a character (basically, how much damage would need to be dealt to kill that character at various levels of PDM).

0   | 100%
5   | 105%
10 | 111%
15 | 117%
20 | 125%
25 | 133%
30 | 143%
35 | 154%
40 | 167%
45 | 182%
50 | 200%

Any amount of healing makes this more extreme. If you heal even just 10% of the incoming damage, it’s like having an extra 10% pdm, which makes the chart look like:

0   + 10% healing | 111%
5   + 10% healing | 117%
10 + 10% healing | 125%
15 + 10% healing | 133%
20 + 10% healing | 143%
25 + 10% healing | 154%
30 + 10% healing | 167%
35 + 10% healing | 182%
40 + 10% healing | 200%
45 + 10% healing | 222%
50 + 10% healing | 250%

Here going from 45->50% PDM increases your effective health by 28% compared to the 6% when going from 0->5% PDM.

I wasn’t on the test server for the previous PDM nerf, but from what I heard, it consisted mainly of halving PDM values on different skills. However, the nonlinear relationship between the PDM value and its effectiveness means using this strategy to bring the high values in line would probably render the lower values useless. I’m guessing that contributed to the negative player feedback.

There’s a lot of ways of approaching this, but I think the best one is what League of Legends does, by making PDM give a consistent amount of effective hit points regardless of how much PDM you have so far.

Mathematically, you can accomplish this with the following equation:

modified_damage = damage / ( 1 + ( pdm/100 ) )

Using this formula, let’s consider the 0%, 5%, 45%, and 50% cases with my 10000 hp 100 damage example.

0 PDM
modified_damage = 100/(1 + 0/100) = 100
10000/100 = 100 attacks

5 PDM
Modified_damage = 100/(1 + 5/100) = 95.24
10000/95.24 = 105 attacks

45 PDM
Modified_damage = 100/(1+45/100) = 68.97
10000/68.97 = 144.99 attacks

50 PDM
Modified_damage = 100/(1+50/100) = 66.67
10000/66.67 = 149.99

There we have it! PDM gives a consistent return per percentage point! Now you can balance whether a skill should give 5 or 10% PDM without worrying that the bonus could be 6X as strong depending on other abilities!

For one final chart, let’s say you just took my formula and patched it in. Let’s look at how it would affect different PDM levels.

PDM | Current Effective Health | Proposed Effective Health | difference

0   | 100% | 100% | 0%
5   | 105% | 105% | 0%
10 | 111% | 110%  |-1%
15 | 117% | 115%  | -2%
20 | 125% | 120% | -5%
25 | 133% | 125% | -8%
30 | 143% | 130% | -13%
35 | 154% | 135% | -19%
40 | 167% | 140% | -27%
45 | 182% | 145% | -37%
50 | 200% | 150% | -50%

It brings the high end of PDM in line while barely touching individual ability amounts! It also makes it much easier for both designers and players to understand how good different amounts of PDM are.

==Why this matters==

The current system of PDM heavily incentivizes groups to hug Fortified Protection Stakes and wear down enemies with attrition over 10-minute long fights, and more players and guilds are waking up to this meta. Forcing players into a narrow meta kills a lot of the great variety that’s possible in Crowfall, and the fact that this meta is such an inert style of play is even worse. I found it so boring that instead of playing, I spent my evening reviewing chain rule so I could derive a new formula to fix it.

[BarriaKarl 2,393]

PDM is a crutch. And fixing it will require a complete rework.

Damage and Healing in crowfall is completely out of wack.

CF is probably the only game I have seen where you have powers that literally reduce damage taken by a fourth and have it no matter. Because the damage is so big it doesnt really make any difference.

A more correct example would be:

the defender will have 10000 hp and be getting hit with 4k damage attacks.

I legit remember stepping on a old trap on my brig and seeing my hp get to half. 1 trap. ONE. I once had people get mad at me for saying 6k hits on a Sin was stupid. Im pretty sure you can find a clip of Fessors one-shotting people currently. Im pretty sure Blazzen would be happy to show you the GIF he has of me getting obliterated.

CF is lucky it still is somehow fun on small scale. If Fessors and BGs were fixed and they added Hellgates, Id prolly play it again. Stupid PDM, Damage, Healing, useless armor and all.

 

[Chipmypants 2]

Well said Abaco! Thanks for doing the work and saying what we have all been feeling.

[Gremel 133]

@AbacoHow would the proposed change effect the meta? If the current meta is to create a tight formation and slowly roll forward in your "PDM Safety Tank" crushing your foes under the treads... would the new meta adjust to just everyone plays ranged/mobile damage because there is very little reason to be grouped up in a tight formation?

Is the "PDM Meatball" just a counter to the ridiculous damage folks put out? 

Which came first, the "PDM Blob" or the 6k+ hits?

I think its fair to say pdm could use a rework especially as you scale beyond 1-2 groups. But to Barria's point so could damage etc... I would hate to trade one meta for another.

 

[Prudenthyaena 17]

Abaco's formula has an interesting characteristic -- if you put a constant in front of the pdm, the formula retains all the properties he wants, while adjusting the break even point for PDM. For example, change it to  modified_damage = damage / ( 1 + ( 2*pdm/100 ) ) makes 50% PDM stay at 50% PDM, scales up PDM below 50, and scales down the extreme cases. It would be easy to adjust the constant to match whatever point of diminishing returns you think is best.

However, introducing diminishing returns into defense while damage scales up with modifiers, seems to be game breaking to me. Damage modifiers are additive within their own type but multiplicative outside them. For example, my assassin has +40% poison and +50% crit damage, but because damage is multiplicative, I get  1.4*1.5 +2.1 --> 110% damage boost instead of a 90% boost. Top DPS stack modifiers to maximize the properties of multiplicative damage. PDM is already additive, putting it at a disadvantage to DPS, so adding another reduction makes DPS even more unbalanced. 

And then there's the unintended consequence of an effective healing nerf. To take an extreme case, suppose you are taking 500 points of damage per second and being healed 250 points per second, Without PDM, the healer doubles your life expectancy. Add in 50% PDM and you're immortal. With your formula, the PDM just triples life expectancy -- going from immortality to triple life is a huge nerf. Already healers have trouble keeping up with damage. It seems that, as Gremel says, this would push the meta to pure DPS, eliminating most of the variety we now have. 

[Gremel 133]

Side note: healing becomes much easier when folks stay together.  Healing is generally location based and that location is usually around the healer (I think confessor healers are maybe the exception).  which means healing is another push towards the blob/meatball formations.

[nihilsupernum 686]

All of this PDM talk is wrongheaded. Crowfall pretends to be an action MMO. Therefore, action (dodging, aiming) should matter.

The fact that it doesn't matter is why we have to talk about these soaking mechanics as if it were a tab target game.

Players should be mostly surviving based on their evasive and defensive actions, not based on mitigation and healing formulas. Mitigation should be a slight assist to compensate slower characters that are unable to dodge as well, or to reward a skillful block. Healing should be an occasional top-up to keep players in the fight, not a DPS vs HPS tug-of-war.

On 10/18/2021 at 9:56 PM, Abaco said:

The current system of PDM heavily incentivizes groups to hug Fortified Protection Stakes and wear down enemies with attrition over 10-minute long fights, and more players and guilds are waking up to this meta. Forcing players into a narrow meta kills a lot of the great variety that’s possible in Crowfall, and the fact that this meta is such an inert style of play is even worse. I found it so boring that instead of playing, I spent my evening reviewing chain rule so I could derive a new formula to fix it.

You could tweak the formula, and it might indeed be better than the status quo, but it won't make the game more interesting. That needs to come from the engine.

 

Edit: Example, Planetside

See how most of the survivability comes not from defense or healing, but from movement and tactics? Obviously, Crowfall is not a FPS, but it is supposed to be an action game. It could stand to be a lot more like that, and a lot less like .. what it is.

Edited October 25, 2021 by nihilsupernum

[Yoink 2,142]

PDM is pretty nuts but so is damage. I'd rather keep PDM as is and introduce ways to punish PDM blobs that are ineffective vs non blobs, whatever that might be.