Bonus Stability makes the "zoomy" problem worse! Here's why, and how EHG could potentially fix this

[queermathsgirl]

Introduction

In the 0.8.5C patch notes, EHG said the following:

@Hackaloken

Echoes now grant bonus stability based on the number of enemies you kill, up to around double the normal amount of stability.

EHG have said on many occasions that they really do not like the “zoomy ignore all enemies” playstyle players tend to use in Monoliths, to either speed through Quests to get to Empowered, Quests to grind Boss Items/Blessings, or to get to a far away Orobyss for Corruption.

This change seems to be somewhat of an answer, perhaps for EHG to test the waters with what does and doesn’t work? At the surface, I suppose it makes sense; players who interact with the content get more stability to achieve the goals mentioned above, thus make faster progress.

The problem is that this “around double” bonus stability is that it doesn’t fix this. In fact, it makes this problem worse.

Data

Data

To investigate this, I hopped into several Last Epoch Twitch streams and watched them complete echo after echo. I recorded the stability the echo granted them, how much bonus stability they could get and how much bonus stability they did get. I ignored Arena echoes as these always give the full bonus, and would thus skew the results.

These streamers all played in the same way: take a direct route to the echo objective and kill most enemies on the way.

These are the findings I got after 100 echoes of results.

Findings

Simply by killing most enemies on the main route, they managed to get on average 50.54% of the bonus stability available.

This ended up resulting in an overall 154% stability revenue per echo. So, by doing nothing different, players who farm enemies while taking the most direct path to the objective get a 54% increase in stability.

There are some other things to look at now:

Not farming

There are a lot of players who just zoom to the objective as fast as possible, ignoring monsters unless absolutely necessary. This is because the monster rewards and xp are not great compared to the rewards and xp gained for completing echoes. See PerryThePig’s video on this for more explanation.

These players would get a slight bonus average due to the existence of Gate echoes, so this a slight buff for them. These figures ranged from 0-33% of the bonus stability. There seem to be an even mix of Boss, Patrol, Spire and Gate echoes (ignoring Arena) so overall an ~8% buff to these players for changing nothing.

Now you may think “Hey these players get 8% bonus but farming along the way yields a 54% bonus, isn’t that worse? It’ll take them more echoes to unlock stuff”.

And you’re right, it would. However there is another factor in all this: time.

I’m a fairly rubbish player, and even I manage to average about 1m30s per echo when zooming through 100+ Corruption. Sometimes it’s longer, but sometimes it’s shorter (like a 10s echo because the spire spawned right next to me… bug?). When I just farm along the way without too much of a focus on farming, it usually takes me 2-3 times as long.

Let’s assume, to be on EHG’s side as much as possible, that whilst farming along the way it takes 3 minutes (double). Then

Stability = 1.54 * s

where s is the base stability of the echo.

Now since the just zooming playstyle would complete 2 echoes in this time, this style would be Stability = 2 * 1.08 * s = 2.16 * s, right? Wrong!

This is because the further you travel from the origin, the more stability each node gives. I have seen +1 and +2. For the sake of keeping on EHG’s side, let’s take +1. This method would then yield the stability:

Stability = 2 * ( 1.08 * s + 1) = 2.16 * s + 2

This isn’t a per-node stability, but the amount of stability you’d get in the same time that a “farm along the way” playstyle would get that 1.54 * s stability.

An easier way to show this comparison is to consider the amount of stability gained over, say, 20 echoes. In the table below, the first row refers to the amount of echoes run in 0.8.5C and the second row is how many echoes worth of stability this is compared to pre-0.8.5C echoes.

Farming Along The Way	Zooming
20	44.375
30.8	47.575

The gap between these two playstyles has gotten smaller. Running 20 echoes while farming along the way would now yield you the same as 30.8 pre-0.8.5C echoes, which is a significant increase. Meanwhile in the same amount of time, just zooming would see you complete 44.375 echoes, yielding the same amount of stability as 47.575 pre-0.8.5C echoes, a slight increase.

But wait, it gets worse!

Zoomy Farmers

There are some builds in the game that I’d dub “zoomy farmers”, such as Hail of Arrows. Since HoA is a duration ability, you can just zoom through the echo and aggro every enemy along the way. Since aggro doesn’t drop in LE (from what I can tell), you will have a massive hoard of enemies chasing you. Simply cast HoA and all these enemies will run through the skill and die.

This method would give the player the same 154% stability as the “farming along the way” method, with the benefit of taking as long as the zooming method (with a 1s timeloss each time you cast HoA, so if this is kept to a minimum or if you just lure the entire hoard to the objective, then you will effectively have the same time as zooming).

So in the time it takes to complete 1 echo as a “farm along the way” player, we have:

Stability = 2 * ( 1.54 * s + 1) = 3.08 * s + 2

Putting this into the table from before gives us:

Farming Along The Way	Zooming	Zoomy Farming
20	44.375	44.375
30.8	47.575	65.975

And suddenly we can see that while the 0.8.5C patch closes the gap between “farming along the way” and zooming, it creates a much bigger problem in another playstyle, one significantly worse than the zooming vs casual style pre-0.8.5C.

Some explanation on the mathematics

Mathematical Explanation

So how did I get the 44.375, 47.575 and 65.975? Probably if you tried to recreate this, you didn’t calculate these.

The 44.375

Running the echoes twice as fast means that when the “farming along the way” playstyle completes 20, the zooming playstyle will have completed 40.

Assuming 20 these 20 nodes each get further away from the origin, the 21st node would give +1 relative to the 20th node, the 22nd +2 relative to the 20th node, etc. This gives the series:

1 + 2 + … + 19 + 20

The sum of this is given by:

Sn = n * ( a1 + a2) / 2 = 20 * ( 1 + 20) / 2 = 210

Now this is the sum of stability, which we want to convert to into number of echoes this represents. Generally you’d divide this by the average base stability granted by echoes, however these additional points of stability come from the 21st to 40th nodes, or the upper half of nodes completed.

So I went into the data I collected and instead of finding the median of the stability of nodes run, I took the upper quartile (which was 48). Dividing 210 by 48 gives:

210 / 48 = 4.375

And

40 + 4.375 = 44.375

The 47.575

Similarly for this case, we get +4.375. However from the Not Farming subsection, we get an extra ~8% stability from bonus stability. So we have:

20 * 2.16 = 43.2 pre-0.8.5C echoes worth of stability

And thus

43.2 + 4.375 = 47.575

The 65.975

Again we have +4.375. This time we have a multiplier of 3.08, giving

20 * 3.08 = 61.6

and

61.6 + 4.375 = 65.975

Note that since I am taking values that are on EHG’s side as much as possible, that the actual figures would show a greater difference.

---

How Could EHG Fix This?

Method 1 - Alter the equation

Method 1 - Alter the equation

We could alter it by dividing base stability by 1.54 (or perhaps slightly larger?) and then give the bonus non-linear or exponential scaling, such that ~50% completion would result in a similar total statibility. For example:

Farming Along The Way	Zooming
20	44.375
30.8	30.89

We can see this strategy still buffs casual play, but nukes the zooming playstyle to more or less the same.

Unfortunately this does nothing for the zoomy farming playstyle which remains at 65.975, but it has the benefit of being a very quick fix.

Method 2 - Scale bonus stability increase away from shortest path

Method 2 - Scale bonus stability increase away from shortest path

Within each echo map, add nodes/vertices at the objectives, the player spawn locations and the intersections where players can choose which way to run. Connect these with edges based on which way players can get to one from the other. Use Dijkstra’s algorithm to find the shortest path. Represent this as a spline. Create vectors normal to this spline. Scale the amount of % added to the bonus stability meter from killing enemies by the lengths of these vectors.

This method is even easier to implement and automate if you generate randomised echo maps as each tile would be its own vertex.

Based on having coded this myself (albeit in C++ and not C#/Java), it shouldn’t take too long to implement this (the long part is manually adding vertices onto the echo maps), nor is it too CPU intensive to impact player loading times, although tbf my code didn’t have to run the entirety of a video game at the same time.

Method 3 - Revert

Method 3 - Revert

Simply revert this bonus stability change and take it back to the drawing room and think of a new approach.

---

Conclusion

I don’t think the fix to try and close the gap between casuals and zoomers worked. Might have seemed good in theory, but didn’t work out in practice. It also create a much bigger problem that makes the original problem mute.

I collected data from 100 non-Arena echoes at 100+ corruption run by streamers, the statistics shown by this data, and how EHG could fix this if they want. The data and statistics are likely to be reliable and representative of echo runs in general given the sample size of 100 (compared to the standard minimum of 30). As mentioned, I took values benefitting EHG’s side throughout these calculations. In reality, these values are much bigger, with the zoomy farming playstyle likely being worth over a hundred echoes instead of 66.975.

---

EDIT 1: I am a weak-willed little bitch.

EDIT 2: Turns out EHG didn’t make this change to fix the zoomy problem, so my bad.

@Trasochi

The purpose of these changes was specifically to narrow the gap between rushing to the objective without killing enemies and taking a possibly less direct route while killing enemies on the way. Most importantly it’s meant to avoid a situation where killing enemies feeling like a waste of time in echoes.

EDIT 3: Updated an incorrect equation.

EDIT 4: Did I have a secret copy of this to paste back in? Nope, I freaking rewrote it by hand. And added some more clarity along the way., including more explanation of the mathematics in the data section.

You can thank ttnnyy for this rewrite.

EDIT 5: These people had very great ideas that build on this. Go check out their comments.

@XLVI_carpo

I want to add more ways to incentivise not zooming (numbers just as examples)

• additional bonus objectives

@Llama8

make the bonus non-linear so the first X% of mobs in the map give a relatively small bonus & the last X% of the mobs in the map give a much larger bonus.

@ClownCombat

What if, the extra stability reward only activates at 70%++ (of the bar filled).

Or: They would introduce the bonus reward for full clear or 90%++.

@vapourfire

Without penalising zooming play, how about simply rewarding a more casual slower approach with a time based solution.

[XLVI_carpo]

Very nicely written. I’m with you on all points. This has not helped with the zoomy playstyle at all. And as I said in a comment in the patchnotes: Clearing maps for bonus isn’t a linear time investment. As you said you took all numbers in favour of EHG and even then you come to your conclusion. In reality it is even worse as you say. Clearing just the way to the objective is a lot less time for around 50% of bonus than clearing full map for the last 50%. I’d guess the last 50% will take 3x-4x as long as the first 50%.

Nobody will stop zooming.

I want to add more ways to incentivise not zooming (numbers just as examples)

• additional bonus objectives

  • open 10 chests for 50% bonus
  • activate 5 shrines for 40% bonus
  • kill 80% of enemies to get 1 additional unique/set/exalted item at the end
  • kill 20 rare enemies to get an item with a random tear sealed affix at the end
  • clear the map to get a dungeon key fragment for a random dungeon (10 fragments make a key)

• prevent the echo objective from spawning/being accessible untill certain requirements are met:

  • kill X amount of enemies
  • loot X amount of special quest item (random drop from enemies)
  • activate 3 random on the map spawned devices to get access to the objective

• add/hide fancy stuff in the map

  • make glyphs and runes only availible through chests - so people will farm the maps in search for the chests
  • add caves with mini encounters that have special rewards (i.e. runes/glyphes, dungeon keys, free skill respec, cosmetics (or special currency to earn cosmetics)
  • remove the echo reward from end of time and put it into a chest somewhere on the map

and:
get rid of the objective marker on the minimap!

Thanks for that great post @queermathsgirl !

Can you please do me a favour and not delete your posts after one day. Please?

[Enkidu]

I have a few questions and I do hope you will indulge me. First I have nothing but respect for math wizards so good on you.

Did you engage with the system before starting your math hunt?
Did you have any idea as to the likely outcome prior to starting your numbers hunt?
How long did it take you to realize your conclusion was likely during the process?

Now with the stated goal of shortening the divide between the farmers and zoomers, and looking at the results, how much faith do you have in the developers? Or do you think their goals are somewhat different? Personally I find when someones actions don’t match up with the goals its not that they are crazy but their goals are different.

I ask because it did not take me near as long to reach the same conclusion, albeit without the math, so I have to wonder if putting in all of the legwork gave you any insight into the development process and goals.

[queermathsgirl]

@Enkidu

Did you engage with the system before starting your math hunt?

I have not played in 0.8.5C yet (due to emotional issues resulting in me wiping my characters, and now not having time to re-grind 20 hours for Empowered). I have however run hundred of echoes pre-0.8.5C and watched several streamers post-0.8.5C.

Hopefully that answers the question? It’s almost midnight so I may not be interpreting it correctly.

@Enkidu

Did you have any idea as to the likely outcome prior to starting your numbers hunt?

Oh, the moment I read the patch notes I knew it wouldn’t fix the problem and would just make it worse. It just didn’t know to what extent as I wasn’t too sure how the bonus stability system worked.

@Enkidu

How long did it take you to realize your conclusion was likely during the process?

After watching the first echo EpochBuilds did, where he got 46% of the bonus stability by barely doing anything.

@Enkidu

how much faith do you have in the developers?

I have faith. EHG seem like good devs, and even listened to my comment months ago complaining about Acolyte’s “boob string”, even though it was minor in the scheme of things.

I also have 1st-hand experience at how much of an utter PITA coding and fixing stuff without breaking everything can be. Too much experience

@Enkidu

Or do you think their goals are somewhat different? Personally I find when someones actions don’t match up with the goals its not that they are crazy but their goals are different.

I don’t know what goes on behind the scenes, but to me it feels like they had an idea to meet this goal that seemed alright in theory but not in practice.

[Enkidu]

Thank you for the reply, and I do genuinely wish you a proper night and a better tomorrow.

It was a bit of a tongue in cheek way to draw out that the stated goal was not going to be fixed by the stability changes. So either the goal was something else entirely, or something went very very wrong in the development process.

[Llama8]

@queermathsgirl

This could be done by adding nodes/vertices onto the maps of the echoes at the objective, player spawn and intersection, and building an adjacency matrix from this (ie. a way of representing all the possible edges the player can travel along). Calculate the path of shortest length. Create vectors normal to this path. Scale bonus stability gain by distance.

I must admit, I like that idea. Either that or make the bonus non-linear so the first X% of mobs in the map give a relatively small bonus & the last X% of the mobs in the map give a much larger bonus.

@XLVI_carpo

Can you please do me a favour and not delete your posts after one day. Please?

Yeah, that too.

[ClownCombat]

Thanks for the great analysis, @queermathsgirl !

I would like to extend on the proposed solution - exponential scaling:

What if, the extra stability reward only activates at 70%++ (of the bar filled).

Or: They would introduce the bonus reward for full clear or 90%++.
That way, once could see if it makes sense for them to go for it, depending on how much they filled the bar already, after completing the main objective.

[Gilavar]

@XLVI_carpo

additional bonus objectives

This is the solution I like most

[queermathsgirl]

@XLVI_carpo

additional bonus objectives

I really like this too.

[Shtrak]

Very interesting. Now, let’s see how EHG interprete the new situation and how they adjust/react.

[queermathsgirl]

@ClownCombat

extra stability reward only activates at 70%++ (of the bar filled)

This would seem viable and would shut down the zoomy-farm playstyle quite a bit. Those who farm their way to the objective won’t have too much trouble going out of their way to get 70%. Regular zoomy would still be superior, but the gap between farming and zooming would be much smaller.

In my data the only value above 60% was this one wacky 89%. So it may not be entirely fool-proof, but seems like it would work 99% of the time.

[ClownCombat]

As a reply to:

“get rid of the quest marker”:

Well, nice perspective!

What makes the zoomy Playstyle possible?
The quest marker.

Maybe instead of increasing the rewards for cleaning, it could be as simple, as introducing the quest marker at 50% of the clear bar progress.

If I go find the quest objective earlier, well congrats me, on to the next echo we go!

Otherwise I slay myself through the map and get guidance at 50%~

[vapourfire]

Another very interesting post… I love how you turn gut feelings about an update / mechanic into something concrete and factual with good old complicated math…

Definitely think this is a “test” on the part of the devs to see how it works… Personally, I like that they are making changes but I agree that this isnt going to work as it stands right now.

I like your Scaling bonuses idea and I thought of another counterintuitive one that might piss some players off…

Without penalising zooming play, how about simply rewarding a more casual slower approach with a time based solution.

e.g.

1. No change to current values for people who want to zoom

2. Add a bonus multiplication of additional stability based on the time taken to completed… Ie… similar to your scaling based on mobs away from the path - just based not on distance from the shortest route, but on time… i.e. a spider killed in the first minute gives 10 stability, but the same spider killed in the 3rd minute gives 50…

3. Obviously have a cap based on whatever metric meets the requirement of and average mono run duration (bell curve kinda shape)… so that people cannot load the map, go make coffee and then return to get the best stability boost… or even link it to time “moving” to prevent someone camping in the map to get the bonuses…

So this doesnt penalise Zooming… it just rewards taking your time exponentially (to a point).

ps. PLEASE pretty please… dont delete your post… I am literally begging here… picture it… big powerlifting jock actually begging on his knees…

[Shtrak]

@XLVI_carpo

get rid of the objective marker on the minimap!

You mean, we should wander on the map, hoping to find something that looks like an objective?
Please no! I agree to reward non-zoomy players. I’m happy to visit maps and kill stuff (hey, that’s why I play the game), but I don’t want to be entirely lost.

[tny]

This was a great post. You have some amazing suggestions for improvements and there are more discussed in the thread. My issue is with your conclusion of “EHG tried to fix the zooming problem”. I don’t believe that was their intention in the slightest. Bonus stability is such a huge quality of life feature that adds decision making later in the game. I cannot fathom how removing it is one of your potential fixes to a problem they were not directly addressing.

@queermathsgirl

EHG have said on many occasions that they really do not like the “zoomy ignore all enemies” playstyle

Their statements were then always followed up with “We’d like to add an incentive for players to kill monsters within an echo.” which I believe the patch has accomplished. The objective is clearly listed at the very top of the patch notes.

@Hackaloken

Our aim with these changes is to improve the rewards for killing enemies in echoes when farming timeline boss drops and/or corruption.

The patch has addressed numerous issues continuously brought up by the community.

Timeline completion requirements
As you mentioned, most players will see an increase of 50% or more in their stability gains without changing their playstyle. The removal of the first two quest echoes in empowered also streamlines the process.

Corruption
On a fresh timeline I was able to accumulate 2 Gazes of Orobyss by the time I reached a +10 corruption node. Providing 8 corruption each, they are worth an entire monolith run. Again, players are not required to alter their playstyle for this benefit. Completing the quest echo and not dying to the shade is a small ask for the amount of time potentially saved.

Stability gains and bonus stability
Compare the initial node of my web against a deep one.


It’s quite easy to determine whether or not clearing beyond an objective is worth the effort early on. However, once you begin farming at capped stability deep into a web the “zoomy-farm” playstyle can become suboptimal. The web thins out and it is often more efficient to take the scenic route to an objective. Naturally, map texture and objective spawns influence your play and decision making but that’s part of the beauty of bonus stability - it’s optional. The amount you choose to interact with it is totally up to you.

My point in all of this is to say that you must look at the features altogether. Bonus stability provides players access to a quest boss faster which feeds into higher corruption gains.

[vapourfire]

Without doubt this is a complex issue… but its also open to interpretation… and because everything is linked its hard to look at everything from every perspective.

I see the “incentive to kill monsters within an echo” to be exactly the same as “fixing the zooming problem”… The how they go about this, then impacts on the stability issue but to me its still killing monsters vs rushing through the objective.

If you really want to take the broader concept into account - I also see the stability issues as addressing the complaint by many players that it takes too long to do mono bosses… and indirectly how alts can access mono content faster too…

I dont link stability to corruption levels like you do… I see these as a separate update to motivate doing more Orobyss islands and potentially getting to higher levels of difficulty faster - entirely different thing to me than farming mono bosses. - and another thing addressing peoples complaints as to how long it can take to get to higher corruption levels on alts…

I see lots of different sides to this… and as I have said before, I think its good the devs are testing things out… but I dont think as you do, that this isnt addressing zooming - I think they are painting with a bigger brush to see how things go…

[Enkidu]

@tny

It’s quite easy to determine whether or not clearing beyond an objective is worth the effort early on. However, once you begin farming at capped stability deep into a web the “zoomy-farm” playstyle can become suboptimal. The web thins out and it is often more efficient to take the scenic route to an objective.

Given that previously the “method” of choice was to use nodes to reroll uncompleted for an additional shot at the desire echo bonus, coupled with the reroll all nodes, before moving onto a higher corruption/reset. I wonder how much if any this is an actual factor worth considering. I am not making a farming vs zooming comparison simply questioning if any zoomy is going to entertain such a question, if not it does seem like a moot point where as there is no functional change for that play style, regardless of the overall changes.

[Hashashiyyin]

A lot of game design is psychological and this feature is no exception. Just giving you any explicit benefit and a cute progress bar to fill up makes slower clears feel a lot more rewarding than they were before, even though relatively speaking they’re still less efficient than fast clears. For a lot of players this is enough, they never did the math and generally care more about how their preferred playstyle feels rather than how efficient it is. And even powergamers are somewhat influenced by this kind of mind trick stuff.

I don’t think the “zoomy farming” is at all equivalent to the “ignore all enemies” playstyle that EHG have said they want to avoid. Adding the requirement to actually kill enemies while rushing the shortest path makes it significantly more challenging as both a skill check and a test of your build, especially as Corruption increases (and with the changes I guess a lot of players will reach high corruption much faster). It’s not obvious to me that this is a “bad” pattern or playstyle.

IMO this patch is a major success that makes mono much more fun to play. That said I don’t think some numbers tuning is a bad idea. It would be fine/good to slightly up the proportion of extra rewards reserved for slow maps or fuller clears. But things should definitely not go all the way to strictly incentivising full clears - in a lot of respects objective rush is a more interesting pattern with a higher skill ceiling.

[ClownCombat]

Thanks for the insights on how much bonus corruption one can expect per Gaze of Orobyss!

Also, I agree with your point, that we should also look closer at what the goal of EHG was bzw. what was stated.

[ClownCombat]

Another Idea, I had in my mind, which might seem a little bit too powerful, but what if:

Every time you full clear an echo, the maximum bonus stability is raised by 1.
(This stacks obviously, but resets once you slay Orobyss)

This way it becomes more and more rewarding to full clear.