The Age of Winter 555 corrupted timline cleared to see all types of nodes there

Feedback and Suggestions
1

Hi,

I’ve cleared all nodes in The Age of Winter 555 corrupted just to see how many nodes would there be for rings and amulets, that I am target-farming.

Well unfortunately - not that many!

  • unique or set ring - 5 nodes
  • unique or set amulet - 7 nodes
  • unique item - 1 node

Other things that I really don’t want to farm in this timline, but they spawned here:

  • exalted amulets - 1 node
  • exalted shields - 1 node
  • exalted rings - 1 node
  • exalted bows & quivers - 2 nodes (WTF, I am melee!)
  • exalted maces - 1 node

All nodes in the timeline: ~220
Number of nodes that I am interested in farming: 12

Don’t you think that there should be much more nodes of targeted items in each timeline? With so little number of target-farming nodes, it is much faster and better just doing boss-prophecies and clearing nodes only for favor/stability.

Screenshots of the timline:

1 Like
2

“Target farming” lol

I never thought about it but now that I see that’s why there are not a lot of nodes I am interested in monos anyway. If I just focus on doing boss most rewards are trash. I wonder if you were very unlucky or this is how many of theese spawn.

And how many unique rings you need to do to get red ring? About 200 on average?

3

There are 19 unique/set rings that are random drops. Red ring has a reroll chance of 98%. That gives us a base chance of 0.001, or 1 in 1000.

4

Thanks for the effort. And man, that looks really depressing.
I’d say though that for CoF prophecies have replaced timeline target farming entirely and for MG you just go to the Bazaar.
The entire timeline concept seems to be pretty pointless right now. I also see little need to put effort into exploring a timeline and to make use of the special nodes unless one is specifically looking for Orobyss’ items.
Which is really sad, because an entire design slice and interactable mechanic is down the drain.

1 Like
5

That’s because the strategy for target farming in Monos relies on juggling Vessels of Chaos and Vessels of Memory. You need to only clear the good rewards (as far as possible) then hit a Vessel of Chaos for all uncompleted echoes to reroll into better rewards and increase the Unique ones you need. Then, when you have enough good rewards, you hit Vessel of Memory and it lets you re-do all node you have completed already.

You can check a more detailed guide for target-farming here:

2 Likes
6

In addition to the entire roll for each node is random, so sometimes you will only see a few and other times you will see more. The number the OP saw was indeed a little low, but that’s b/c of the RNG factor of each node rolling what it did.

7

I know all this but imho is still too litle reward for time spent. Espescially if you have prophecies that grants 14 unique rings for 9001 favor. It is so much faster to farm bosses with this prophecy than clearing monos… and that is sad to me because I personally enjoy clearing maps more than killing bosses… but that just me… :tipping_hand_woman:

8

Oh, I supposed you didn’t know how to properly use Vessels because you never mentioned them in your whole analysis about the number of valuable nodes in a timeline.
And that’s detrimental to the argument of there being very little valuable nodes to farm, since Vessels are there specifically for this.
If there were too much, say, unique nodes on a fresh timeline, then you’d be overflowing with unique nodes when you did the strategy properly and hit 2~3 Chaos/Memory in a row.

Also, there are people that don’t play CoF, so they have no idea what you’re talking about here:

9

You are right I am always thinking from the SSF perspective cause it is my standard way of play in LE. IMHO for CoF monos are not optimal way to target-farm. That is my statement for today, so I am waiting for next big patches to change this system or give more variety to it. We will wait and see I suppose :slight_smile:

1 Like
10

Does reroll chance mean that the Unique rolls as a Red Ring, however that Red Ring then gets rolled again with a 2% chance of staying a Red Ring and a 98% chance of becoming a different Unique?

11

Which takes high investment in effort and knowledge though and doesn’t help overall progression for players. The rewards give a bit of variety… but with the factions this method has become fairly worthless.

As CoF your general target is to finish prophecies, they’re in all aspects superior to a high setup method.

As MG you go for generic drops which could’ve value, you don’t target farm (obviously you might do, but in the big picture it’s not the case overall)

This leads to the fairly… wonky mechanic it was formerly to become lackluster simply.

And that’s the factual outcome when put in reference to other methods, it fails to do its job by now.

1 Like
12

Yes that is how reroll chance works.

1 Like
13

Which is another bad thing actually in my eyes.

It’s an issue both for balance and also for performance. There is a reason why commonly such things are done via ‘weight’.

If we imagine that there’s 20 rings possible to drop and 10 of them can re-roll then that code will become a bit of a mess already.

How do you discern which item drops exactly how often? After all if you add new rings it’ll change the statistical possibility not proportionally to the re-roll chance and total rings available but also with the re-roll chance for other rings. However does someone expect to actually balance around that then?

And also… if the program gets that code it’ll go through the function, decide ‘whoops, we got our value but we don’t want it’ then throw it away and start from scratch. That’s fairly nonsensical.
It causes the same code to re-run again for now reason, it needs to decide again between all rings, even worse if it needs to remove that option and re-roll without the red ring option available, because that causes unnecessary extra steps for the program to do.

Instead of that using a fixed ‘weight’ value, adding all of them together and deciding with a single number outcome which one it actually produces is vastly more efficient and also gives direct feedback to the coder in charge of balancing.

Might even be one of the big reasons why loot creation causes such FPS issues at time. No wonder when you could theoretically run the same line of code dozens of times rather then once. It’s nonsensical.

14

It’s a simple calculation loop all done in memory. Modern computers can run thousands of these per second. So I very much doubt that is the reason for the FPS drops.
If you had to access the database or the game files on each loop, that might happen, but not on a simple calculation like this.

It’s much more likely that it’s the rendering of multiple items at once, since that is a task that more demanding for the PC.

15

We don’t know what’s the rat-tail attached to this decision though. Yes, it can be looped thousands of times per second… but what if there’s a whole amount of functions running through it first before even getting to that? Then it turns from thousands… to hundreds… to one.

Also why are you trying to defend objectively bad coding? There is no instance where this choice provides any positive. It’s bad for scaling, it’s bad for the initial implementation, name a singular instance where it’s better then the alternative and I’m fine with it.

As for rendering… likely! The major point which could cause it.
Could it nonetheless be this as well? Absolutely! Saying otherwise is nonsense.

16

Why would there be? Stuff drops, call the function. Pretty standard.

It makes no objective difference for scaling. If you made it a simple weight table red ring would have a 0.0001 weight, the same as it does now.

In fact, it’s actually easier to add new stuff. With a weighted table you have to adjust every single value whenever you add a new unique so it adds up to 100%. This way you just need to add the desired rarity for the item and not change anything else.

No. It’s like saying that rolling 20 random numbers each time causes a random drop when the operation weight is completely negligible in modern computers.

17

By your own words: “We don’t know how their codebase looks like”
It’s also pretty standard to use a weight-system for the exactly named reasons from my side. That’s lowest of lowest level for coding.
So that wasn’t done… you can expect other things to happen.

But it doesn’t roll ‘20 random numbers’, it rolls between ‘1 to infinity’ numbers every time or ‘1 to n’ times if it removes the respective possible outcome.

18

That doesn’t mean you can expect the code for determining drops to be called up when you move. Or when you assign a passive point. There are reasonable expectations which is that it’s called only when something drops to determine what it is.

Since there are 19 ring drops, it rolls 1 to 19 times. Less, because many items have a 0% reroll chance. So totally negligible. Even if you had 100 possible rings, it would still be totally negligible.

Like I said, the most likely reason for doing this instead of a weight table is so you don’t have to change every single item every time you want to add a new one. You just determine its reroll chance according to which rarity you want. Actually makes it easier to add new uniques and place them accordingly in their rarity tier.

19

But it could be called up after all values of the item are decided, which means the whole item creation process being done before the re-roll triggers (which would be dumb, but who knows?)
Or it could be embedded in a for loop of other general decisions and outputs a value which designates it as ‘failed’ and hence to be re-done from scratch, with whatever else might be in there.

There’s a myriad of dumb situations that could be done and it functions nonetheless.

The code design itself for that tiny aspect though is objectively bad.

Factually wrong.
If it doesn’t remove the red-ring from the viable choices the theoretical upper limit is infinity.
You should know that. You shouldn’t get that wrong even in your sleep.

19 rings. Number 19 is the red ring. It lands on 19, now it creates a value between 1 and 100, it lands on anything but ‘1’ or ‘2’, it returns. It rolls ‘19’ again, it goes and rolls the value… as often as needed.
Infinite… times… theoretically.

So this is a scaling problem. The more rings you have which offer a re-roll chance and the higher this chance is the more loops it’ll do, which is a exponential usage of resources.

What?
That’s also nonsensical!
A weight table has a end-value, every item has a weight. You add all values for the items together. The value returned and in the range for that item causes it to drop, end of function.

It’s worst-case 2 vectors which have the minimum value and the maximum value inside. MinVector’s next entry is decided by the same position of MaxVector’s former value + 1, MacVector’s value is the weight-range. The follow-up function hence dynamically resolves that.

It’s basic stuff.
Mind-numbingly basic stuff. We’re not talking about a bubble-sorting method created from your memory… and heck… you should know a simple bubble-sorting method in your sleep as well, and that’s vastly below that level of expertise.

My provided example is scalable, yours is not as too many entries cause the loop-time to exponentially increase towards infinite, mine doesn’t.
The only limiter is value type size limit, and if you reach a long long value’s max size because of that then whoever designed the system needs to get fired anyway.

20

It could many things. It could even be called on the splash screen. However, we’re assuming that the devs actually know what they’re doing, since this a basic type function and we have no evidence to the contrary.

Otherwise you could say the same thing about everything.

Note the “different unique”. It seems quite obvious that the unique gets taken out of the pool. In fact, I think Mike’s wording on this is “the item gets discarded and it rolls again”.
So let’s once again assume they know what they’re doing on basic level stuff.

You’re trying to find a problem where none exists. In terms of rarity scaling there is no objective difference between both systems and the system LE uses simply makes adding new uniques a lot simpler without having to change the whole table for every single item.