Jump to content
Dante Unbound: Share Bug Reports and Feedback Here! ×

Hey DE, care to explain this? (Undesirable effects of a flawed DPS cap.)


Traumtulpe

Recommended Posts

2 minutes ago, Traumtulpe said:

As already mentioned, I was expecting diminishing returns instead of a hard cap in the first place. Also, those goals are not even met. I can kill an Acolyte in one second or less, using a melee weapon.

At the very least, the case with Vex Armor doing nothing, and people complaining about the effectivity of guns, should make it obvious that the current cap is in the wrong spot.

But how else will diminishing returns work, mathematically?  If you consider DPS, that is basically already what we are getting -- currently DPS gains asymptotically approach 0 as DPS approaches (Health+Shields)/10.  The problem is that, with higher fire rate, your Damage per Shot has to go down for DPS to have diminishing returns.  I'm trying to think of a way to guarantee Acolyte screen time without reducing Damage per Shot, but given the damage-to-time conversion, the best I can come up with are:

1. Forget Acolyte screen time.  As you said, melee currently kills them in less than 1 second.  So make the game even easier by making Acolytes no stronger than regular enemies?
2. Give Acolytes an insane 9-second Damage Reduction buff (maybe every hit goes down to 1) after they show up, or after they take the first hit of damage, ensuring their screen time.  After the buff expires, all of our hits deal "regular" damage.

Or is there something else other than screen time that DE can balance by?

Link to comment
Share on other sites

2 minutes ago, MqToasty said:

But how else will diminishing returns work, mathematically? [...] Give Acolytes an insane 9-second Damage Reduction buff (maybe every hit goes down to 1) after they show up, or after they take the first hit of damage, ensuring their screen time.

Diminishing returns don't have a hard cap. Maybe Vex Armor would only increase my damage by 25%, relative, but that would be a lot better than literally no difference.

Your example is a bit forced, you could give them destructible armor on body parts, for example.

Link to comment
Share on other sites

2 minutes ago, Traumtulpe said:

Diminishing returns don't have a hard cap. Maybe Vex Armor would only increase my damage by 25%, relative, but that would be a lot better than literally no difference.

Your example is a bit forced, you could give them destructible armor on body parts, for example.

I do not know if this is how they coded it, but if you are deep enough in the diminishing returns, it may seem like a hard cap.  Here's an example:

Suppose the DPS formula is:
Adjusted DPS = 53,000 * DPS / (DPS + 10,000)

Just like DR from Armor, this is diminishing returns, right?  But since you've already hit 51,000+ Adjusted DPS before Vex Armor, you may not see any difference with Vex Armor because the added damage gains you so little.  Basically, diminishing returns will not help the Vex Armor situation, but adjusting by something other than DPS might make the Fire Rate differences more convincing.

Link to comment
Share on other sites

8 minutes ago, MqToasty said:

you may not see any difference with Vex Armor because the added damage gains you so little

There is literally not a single point of damage added by Vex Armor. We have an easy to reach, hard DPS cap. This is not how diminishing returns work at all.

Link to comment
Share on other sites

1 hour ago, MqToasty said:

But how else will diminishing returns work, mathematically?  If you consider DPS, that is basically already what we are getting -- currently DPS gains asymptotically approach 0 as DPS approaches (Health+Shields)/10.  The problem is that, with higher fire rate, your Damage per Shot has to go down for DPS to have diminishing returns. [...]

The problem is that gains are not approaching 0, but a negative number.
If you have to reload your weapon, your DPS is higher with worse gear. It should be

better gear -> better overall performance (though not necessarily linear)

but it is

better gear -> better performance up to a point, until:
worse gear -> better performance (taking of a fire rate arcane improves kill speed)

Link to comment
Share on other sites

To summarize what I have demonstrated in the first post:

  1. An Arcane that says "Increases fire rate by 120%" is lying, in truth the effect is "Reduces ammo efficiency by 75%".
  2. An ability thas says "Increases weapon damage by 759%" is lying, in truth the effect is "Nothing".

I'd still really like someone from DE to show up and say something about that. Fat chance, huh?

Link to comment
Share on other sites

1 minute ago, Her_Lovely_Tentacles said:

The problem is that gains are not approaching 0, but a negative number.
If you have to reload your weapon, your DPS is higher with worse gear. It should be

better gear -> better overall performance (though not necessarily linear)

but it is

better gear -> better performance up to a point, until:
worse gear -> better performance (taking of a fire rate arcane improves kill speed)

Well, the DPS (discounting reload time) does improve slightly with arcane velocity.  But yes, I agree that the reload does make it perform worse with arcane velocity.  Yet that is where the multiplayer part comes in, right?  If you got 4 guys beating up Violence (ironically?), then you may not need to reload and thus will deal more damage with arcane velocity.

40 minutes ago, Traumtulpe said:

There is literally not a single point of damage added by Vex Armor. We have an easy to reach, hard DPS cap. This is not how diminishing returns work at all.

Okay, I think we need a more concrete example to show you what I mean.  Again, I do not know if this is the equation currently in use (the constants are almost definitely wrong).  However, the structure of the diminishing returns equation is pretty straightforward and this example can demonstrate how there will always be hard limits.  And when you are near the limit your gains could easily be so small that they get truncated in the UI and/or rounded down:

Suppose the formula was actually:
Adjusted DPS = 51,750 * RDPS / (RDPS + 100)

So your Raw DPS would have been:
(53.76 * 962) = 51,750 * RDPS / (RDPS + 100)
RDPS + 100 = 1.000636 * RDPS
100 = RDPS * (1.000636 - 1)
RDPS = 156,723

With Vex Armor your Raw DPS would be approximately (assuming you only have a +220% Hornet Strike and no other damage boosting mod):
RDPS = (7.59 + 3.2) * 156,723 / 3.2 = 528,450
Adjusted DPS = 51,750 * 528,450 / (528,450 + 100) = 51,740
@ 53.76 shots per second, each shot will deal 51,740/53.76 = 962.4 damage, still rounded down and displayed as 962.

Link to comment
Share on other sites

2 minutes ago, MqToasty said:

Well, the DPS (discounting reload time) does improve slightly with arcane velocity.

No, the DPS (discounting reload time) stays exactly the same, which is why the very short reloads have any impact at all.

We know from testing that DPS cannot be improved beyond a certain point, far below what is achievable. Your posts are semantics, the DPS is capped, that much is obvious. Common implementations of diminishing returns do not work like this at all, there may be a theoretical ceiling, but it is always unreachable.

Link to comment
Share on other sites

5 minutes ago, Traumtulpe said:

No, the DPS (discounting reload time) stays exactly the same, which is why the very short reloads have any impact at all.

Show me your work.  I showed you my math in my first post on page 3, and your DPS (discounting reload time) increased from 51,717 to 51,744 with arcane velocity.

6 minutes ago, Traumtulpe said:

We know from testing that DPS cannot be improved beyond a certain point, far below what is achievable. Your posts are semantics, the DPS is capped, that much is obvious. Common implementations of diminishing returns do not work like this at all, there may be a theoretical ceiling, but it is always unreachable.

Oh for crying out loud, why do I even bother?  I did not write this post from 6 years ago, but read it and learn how diminishing return formulas work for yourself:

https://gamedev.stackexchange.com/questions/109985/simple-diminishing-return-with-cap

Link to comment
Share on other sites

6 minutes ago, Traumtulpe said:

To summarize what I have demonstrated in the first post:

  1. An Arcane that says "Increases fire rate by 120%" is lying, in truth the effect is "Reduces ammo efficiency by 75%".
  2. An ability thas says "Increases weapon damage by 759%" is lying, in truth the effect is "Nothing".

I'd still really like someone from DE to show up and say something about that. Fat chance, huh?

That isnt at all the truth though. You talk as if this applies to the whole game, in which case you'd be correct that the tooltips are lying. But this applies only to very specific mobs that have a special DR mechanics that accounts for everything damage related in your kit in order to not go above the damage hardcap. 

Clip one shows faster fire rate, this means the game adjusts the damage output to compensate for more ticks occuring per second in order to treat damage increasing stats nearly the same.

Also, you say in your OP that clip #2 is 20% shorter... how does that add up? Clip 1 takes 16 seconds from start to finish, clip 2 takes up the rest of the 31 seconds of that whole video. That is a difference of 1 second, that is not 20% lower dps. Did I miss something in the video showing the 2 clips you talk about or are you refering to some other clip being the second one that you havent posted? All the video shows is that the system works since it takes all damage increasing sources into account.

Link to comment
Share on other sites

1 minute ago, MqToasty said:

read it and learn how diminishing return formulas work for yourself

If you are able to reach a point where you experience no further increases for your investment, you are talking about a cap. There may be diminishing returns leading up to that cap, but it is a cap nonetheless.

The entire point of diminishing returns on player stats is to enable improvement while limiting it.

While I feel sorry for your precious math, a 0.00001% increase that cannot even be displayed in game does not count.

Link to comment
Share on other sites

7 minutes ago, SneakyErvin said:

That isnt at all the truth though. You talk as if this applies to the whole game, in which case you'd be correct that the tooltips are lying. But this applies only to very specific mobs that have a special DR mechanics [...] Also, you say in your OP that clip #2 is 20% shorter... how does that add up? Clip 1 takes 16 seconds from start to finish, clip 2 takes up the rest of the 31 seconds of that whole video.

Everything else can be oneshot. Everything where your damage matters has this DR.

I have the files in a video editor, the first clip is slightly longer than 17 seconds, the second one is almost precisely 14 seconds. And I made sure to cut them at exactly the same moments (right before the first ammo is used, and right before the Steel Essence pickup get's displayed).

Link to comment
Share on other sites

5 minutes ago, Traumtulpe said:

If you are able to reach a point where you experience no further increases for your investment, you are talking about a cap. There may be diminishing returns leading up to that cap, but it is a cap nonetheless.

The entire point of diminishing returns on player stats is to enable improvement while limiting it.

While I feel sorry for your precious math, a 0.00001% increase that cannot even be displayed in game does not count.

Um, okay...  Since you believe you have the "correct" understanding of diminishing returns (and everyone else is wrong), why don't you post your answer to the Game Development Stack Exchange question?  I'm sure the rest of the world can benefit from your insight and future games will become better for it!

Link to comment
Share on other sites

22 minutes ago, Traumtulpe said:

Everything else can be oneshot. Everything where your damage matters has this DR.

I have the files in a video editor, the first clip is slightly longer than 17 seconds, the second one is almost precisely 14 seconds. And I made sure to cut them at exactly the same moments (right before the first ammo is used, and right before the Steel Essence pickup get's displayed).

Damage matters everywhere just the same. It isnt like you have any use for more damage versus acolytes either. Would you prefer inflated health pools just in order to see massive numbers? Heck, the acolytes are probably some of the most fragile mobs in the game even with their damage cap.

I hope you dont claim the first clip provided here is slightly longer than 17 seconds, since it ends pretty exactly on 16 seconds with clip 2 landing the first shot nearly exactly at the 17 second mark, which results in the ttk being 16 vs 15 seconds in total i.e a 1 second in TTK difference since the full clip is 31 seconds long.

Link to comment
Share on other sites

6 hours ago, MqToasty said:

have to ask though, what kind of change are you hoping for?  The problem, as I see it is this.  DE is attempting to balance the game while maintaining (at least) these aspects of the game:

1. Provide a heavily moddable weapon system with more than a dozen types of damage, variable number of shots per trigger pull and variable fire rate.
2. Provide some semblance of a challenge where field bosses do not die too quickly (10 seconds seems to be the target for Acolytes here).
3. Multiplayer, allowing different players to tackle the same foes with different weapon setups.

Unrelated to the thread but I personally believe some of the dozen damage types can go.

Same with the dozen+ of enemy health types, the "complexity" of damage types has long been power crept unless flat out enforced by things like The Wolf of Saturn Six due to the fact that anything capable of blasting through super high level Grineer makes mulch of the other main factions just as easily.

But that's just my opinion on that matter.

Link to comment
Share on other sites

6 minutes ago, Aldain said:

Unrelated to the thread but I personally believe some of the dozen damage types can go.

Same with the dozen+ of enemy health types, the "complexity" of damage types has long been power crept unless flat out enforced by things like The Wolf of Saturn Six due to the fact that anything capable of blasting through super high level Grineer makes mulch of the other main factions just as easily.

But that's just my opinion on that matter.

I agree with you.  I do not know how best to approach such a feature reduction, but it does seem like most of the types have become rather...  Pointless?

Link to comment
Share on other sites

8 hours ago, Traumtulpe said:

If you are able to reach a point where you experience no further increases for your investment, you are talking about a cap. There may be diminishing returns leading up to that cap, but it is a cap nonetheless.

 

Maybe DE’s diminishing return’s math is a bit harsh? Supposed your DPS is 500.000. DE’s DR reduced it to 500. Now you are gonna add 50.000 more DPS to make it 550.000. DE’s math then reduced that extra DPS to 0.0001, because this number is so small it rounds down to 0. Wiki’s math also proved this, if you deal 63.6 damage it will round it down to 63.

This shows that the damage system is broken beyond repair and needs a complete overhaul from the ground up. DE has to use convoluted DR system to keep players from killing enemies too fast. Maybe it’s time to trim the damage system fat and reduce the available damage types.

Link to comment
Share on other sites

42 minutes ago, DrivaMain said:

Maybe DE’s diminishing return’s math is a bit harsh? Supposed your DPS is 500.000. DE’s DR reduced it to 500. Now you are gonna add 50.000 more DPS to make it 550.000. DE’s math then reduced that extra DPS to 0.0001, because this number is so small it rounds down to 0. Wiki’s math also proved this, if you deal 63.6 damage it will round it down to 63.

Maybe. But the difference is semantics. The damage is hard capped, as far as gameplay is concerned. I'm not going to get caught up in weird math jokes, like .9999... =1. This is a game after all, what matters is functionality, not theory.

Link to comment
Share on other sites

8 hours ago, MqToasty said:

Okay, I think we need a more concrete example to show you what I mean.  Again, I do not know if this is the equation currently in use (the constants are almost definitely wrong).  However, the structure of the diminishing returns equation is pretty straightforward and this example can demonstrate how there will always be hard limits.  And when you are near the limit your gains could easily be so small that they get truncated in the UI and/or rounded down:

Assuming you are correct (since i'm not in the mood of that much math), that disminishing return would inevitably lead to a hard cap, my view however, is that players should never have to discover the cap in this way, it would just leave bad taste in our mouth.

Games are already smoke and mirror, the devs have the all power to decide if we could reach the cap, of if they should keep the cap, or make another cap just for that cap, ... Not all options are available obviously (like how DE can no longer control player's dps ceiling), but they can make that math whatever they want.to be.

I don't know how to best handle a situation like this since i'm not a developer, but even just a sign post saying that "congratulation, you reached the cap" from the enemy would have satisfied me, even if they trying kill me afterward with the excessive DPS i have, it would still better than figuring out they messed with the number behind the scene.

Keeping the enemy in screen for x second isn't a virtue, but being consistent in player experience should be.

 

Link to comment
Share on other sites

53 minutes ago, FireSegment said:

Assuming you are correct (since i'm not in the mood of that much math), that disminishing return would inevitably lead to a hard cap, my view however, is that players should never have to discover the cap in this way, it would just leave bad taste in our mouth.

Games are already smoke and mirror, the devs have the all power to decide if we could reach the cap, of if they should keep the cap, or make another cap just for that cap, ... Not all options are available obviously (like how DE can no longer control player's dps ceiling), but they can make that math whatever they want.to be.

I don't know how to best handle a situation like this since i'm not a developer, but even just a sign post saying that "congratulation, you reached the cap" from the enemy would have satisfied me, even if they trying kill me afterward with the excessive DPS i have, it would still better than figuring out they messed with the number behind the scene.

Keeping the enemy in screen for x second isn't a virtue, but being consistent in player experience should be.

 

🤔 Is such a signpost necessary, though? I don't know the individual situations, but it feels like at some point people would have been deleting the toughest enemies, and there may have still been upgrades to get by that point. I'd expect that not everyone would have reached the same point in the same way, so I figure the remaining upgrades would be for those who were going down an alternate route or had a different RNG experience.

If true, why would they still chase the upgrades? The game would already be fairly obviously broken by then. Perhaps it's a mindset thing...? I have my theories and reckon I know, but it's fascinating talking to those who've reached this point

edit: Personally I'd keep chasing because it's there, but I'd have identified that I'm already at the top. I'd probably treat the additional DPS as different options; perhaps I don't want a fire-rate increasing mod, so I'd go for an Arcane instead and slot in Motus Signal to replace the fire-rate mod

Link to comment
Share on other sites

It's lame, it sucks, and did I mention it sucks and it's also lame?

Although what it really says to me is that this game's got way too many damage multipliers*, and too much spread between low end  and high end damage output in content that almost everybody is encouraged to play.  Resulting in DE shenanigans that will make you grind your teeth if you're aware of them. 

 

 

*And, as a pretty enthusiastic Banshee player, it almost physically pains me to type those words.

Link to comment
Share on other sites

3 minutes ago, Tiltskillet said:

It's lame, it sucks, and did I mention it sucks and it's also lame?

Although what it really says to me is that this game's got way too many damage multipliers*, and too much spread between low end  and high end damage output in content that almost everybody is encouraged to play.  Resulting in DE shenanigans that will make you grind your teeth if you're aware of them. 

*And, as a pretty enthusiastic Banshee player, it almost physically pains me to type those words.

+1

Come on DE, when even the Banshee players are saying it you know it has to be a problem.

Link to comment
Share on other sites

19 hours ago, Traumtulpe said:

Are you trying to waste my time? I have nothing more to say to you.

Nope, just trying to get you to prove something you claimed that cannot be seen by the human eye really. Watching the edited video makes it look like test 1 takes 16 seconds and test 2 takes 15. Since that is what the eye percieves, the gap between the two is highly pointless to argue since it will never be experienced by a player. No one will go "zomg! this was a slow kill!" because they wont notice.

If anything they'd notice that each hit hits for far less when attack speed is added.

Also, are you sure the difference actually comes from attack speed being penalized and not from DoT status being capped to 4 and constantly reapplying itself faster than it can get a chance to tick when the attack speed is added? Lots of games have those issues with DoTs, which is why tight rotations in games such as WoW and FFXIV are important for DoT casters, in order to not waste full ticks due to reapplying them too early. I wouldnt be surprised if it is the same here when stacks are capped. It usually isnt an issue since they can more or less stack infinitely here.

Link to comment
Share on other sites

12 hours ago, FireSegment said:

Assuming you are correct (since i'm not in the mood of that much math), that disminishing return would inevitably lead to a hard cap, my view however, is that players should never have to discover the cap in this way, it would just leave bad taste in our mouth.

Games are already smoke and mirror, the devs have the all power to decide if we could reach the cap, of if they should keep the cap, or make another cap just for that cap, ... Not all options are available obviously (like how DE can no longer control player's dps ceiling), but they can make that math whatever they want.to be.

I don't know how to best handle a situation like this since i'm not a developer, but even just a sign post saying that "congratulation, you reached the cap" from the enemy would have satisfied me, even if they trying kill me afterward with the excessive DPS i have, it would still better than figuring out they messed with the number behind the scene.

Keeping the enemy in screen for x second isn't a virtue, but being consistent in player experience should be.

 

Well, I do agree games are smoke and mirror.  I personally enjoy peeking under the curtain to figure out how the ingenieur designed all the tricks, and if I ruined the game for you because of that, then I apologize.  But in my defense, I did not create this thread and only hopped in around page 3 -- by then I believe there was already plenty of speculation and the veil had already been partially lifted.

As for your sign post idea, wouldn't that break the magic for even more gamers, though?  While I assume most gamers will neither be this min-maxed nor notice that their Vex Armor is having no appreciable effect on Acolytes (even OP did not until the way-too-obvious damage per shot reduction led him to comb through his other videos), wouldn't that signpost ruin the game for more gamers who do not want to see all of the gears, levers and pulleys behind the scenes?

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...