Guide Psa 3: The Complete Saga, Ain Discovers Wynncraft Is A Lie.

I - The Phantom Manas

As the first of the recent set of discoveries, @hppeng and @Neptune found that the Mana Regen ID (which displays as x/4s in game) actually ticks every five seconds, and natural mana regen ticks once every second. Here is a link to hppeng’s thread where this is all discussed. This was speculated previously, but seems to be more well-known now.

@Saya has also put out a video on the Mana Regen ID post-Gavel, which you can check out here.

Spoiler: video

This is part of a set of videos some of us at Atlas Inc want to put together to help make the mechanics of Wynncraft and class building more accessible.


II - Attack of the Detlas Test Dummies

If you’ve seen a build using wynnbuilder recently, you’ve probably noticed that Mana Steal and Life Steal have also had their numbers changed, to /3s. @highbread made a thread on this here, which I encourage you to read for the details on how we tested this and the results. For a TL;DR, a couple of us smacked test dummies for a few hours and did some math and realized that Mana Steal probabilities are WAY higher than we assumed. Salted confirmed the suspicions here in highbread’s thread.


III - Revenge of the Ticks

A little more info on similar IDs: Poison and Health Regen (hpr). Poison ticks do work as described (that’s a first), although there is a slight delay before the first tick. The timeline is:

Melee Hit - 1.25 seconds - 1st psn tick - 1 second - 2nd psn tick - 1 second - 3rd psn tick.

If you hit a mob once, poison is effectively /3.25s, but in practice you will be constantly hitting a mob to the point where the delay doesn’t matter, and your poison dps becomes /3s.

For hpr, which isn’t given a duration in game, the ticks happen every 4 seconds (as was previously thought). This makes sense, as it has been changed a few times over the years—nice to know it was implemented correctly. The amount natural hpr heals by is still quite a mystery. We know the number increases as your hp decreases but decreases as your mana decreases. Yeah, weird. This is likely a holdover idea from the implementation of hpr as an out of combat mechanic (as it used to be). I suspect we will keep trying to figure out that formula as time goes on.


IV - A New Trope

Now for the new weirdness. @Gigavern and @Bart (MC) have mentioned a few times in Atlas Inc’s discord that the values we use for attack speeds and their hits/sec are inaccurate, especially at higher attack speeds, though they couldn’t tell us real values. We have done some initial testing, and can confirm that a) our values are inaccurate and b) it’s really strange.

Important note: Spell multipliers remain the same. Super fast weapons still get the 4.3x multiplier on spell, but the melee attack speed is different.

I’ll explain more of the testing process in the next section (including when we got a mod to macro for us on video), but here is a quick overview first:

1. sfast, vfast, and fast speeds all deviate significantly from their multipliers
   
   
2. Archer and Shaman appear to attack at slower speeds
   
   
3. Higher cps doesn’t necessarily correlate to higher dps.

V - The Macro Strikes Back

So, here it is, the attack speed info. We initially were testing this by getting some people with sustained high cps (@Highmore did a lot of the work) to manually melee and record it on video, to which I counted the attacks by ear and divided by the total time elapsed. This is useful info (as discussed later), but highbread came in clutch with a new method to get much more reliable results. Through the use of a non-looping left click macro, we were able to test the left click classes at a consistent 32 cps. At the time, we couldn’t get a consistent right click for Archer, so we called in backup from @NagisaStreams who came in hot with a 100cps Archer macro. (Note that not all macros are allowed under Wynncraft rules. If you are unsure about one, ask a moderator, like Nagisa.)

To make the process easier, we used builds that did the same amount of damage per hit with 0 dexterity (using weapons like crackshot and imperious), which let me take the final health of the test dummy and get an exact value for the number of hits without having to do it by ear.

For the data:

• Super Fast, which has a 4.3x spell multiplier, has an attack speed of approximately 3.97 hits/s on Mage. We *assume* that Warrior and Assassin share this value, but didn’t take enough data to confirm that. Shaman clocks in at 3.75 hits/s, and Archer lags far behind at 3.40 hits/s.
  
  
• Very Fast (3.1x spell mult) had a wide range of results, but I believe it has an approximate attack speed of 2.81/s on Assassin (and assumed on Warrior/Mage), 2.76/s on Shaman, and 2.63/s on Archer. These numbers could be much closer to each other around 2.70/s.
  
  
• Fast (2.5x) had our weirdest set of data, with trials reaching quite different results. In talking a bit with some of the IM team, we think that Highmore’s 2.28/s on Assassin is close to the real value, but much more testing is needed here. (Without the IM info, I would interpret our data as 2.10 to 2.15 hits/s.)
  
  
• Normal (2.05x) had only 2 trials from me, where I clocked an even 2 hits per second on warrior.
  
  
• Super Slow (0.51x) registered essentially the same value (0.52/s, with some uncertainty) from @H07oh. We saw that the slower speeds seemed to be matching their spell multipliers, so we chose not to prioritize those for testing.

We would love to have more data! If you want to help us test, ping me @touhoku#8699 either on Atlas Inc’s discord on the Wynn discord server. I will keep this thread updated with the true values.

For now, we have a running hypothesis courtesy of hppeng: that attack speed is determined by different tick cooldowns for weapons. Ignoring Archer and Shaman for the moment, the data seems to match the following cooldown lengths:

• sfast: 5 ticks (4/s)
  
  
• vfast: 7 ticks (2.85/s)
  
  
• fast: 9 ticks (2.22/s)
  
  
• normal: 10 ticks (2/s)
  
  
• sslow: 38 or 39 ticks (~0.52/s)

The data for Archer and Shaman does not fit with this at all, so it could be totally wrong, or projectile motion/bow mechanics could be throwing a wrench into things. Again, if you have the time to test, please help us out!


VI - Return of the Spellsteal

I want to start this section by saying that: nothing in the game mechanics has changed recently. Your builds won’t feel different, they just might make more sense. What we have found isn’t new mechanics, it’s that the mechanics that have existed and either incorrectly explained or not explained at all in game.

So, back to mana. A common term in class building is ‘net mana,’ which refers to the sum of your MR, MS, and natural mana regen ticks. Previously, this was an easy calculation. Add Mana Steal and Mana Regen, then tack on an extra 5 for natural. Now, it’s a different story.

In my opinion, calculating effective mana over a five-second cycle seems like the most natural, as regen is consistent, whereas steal is chance-based. The base formula for effective mana is then:

EM/5s = 5 + mr + (5/3)*ms

However, the attack speed weirdness comes into play here. We think that the estimates for Mana Steal being /3s still hold for the new attack speed values, as the differing numbers (sfast, vfast, fast) coincide with our data points in the steals testing that matched up least with our predictions, and new attack speed numbers make the prediction closer. Stuff starts to get strange when talking about a) Archer and Shaman, and b) user cps.

Because Archer and Shaman have slower attack speed values, but (we think) the same steal chances (this should be tested), Mana Steal is less effective on Archer and Shaman than it is on say, Mage, but only for fast attack speeds. The buildstyle hit the hardest by this is probably fast hybrid shaman, which often runs a TFA style dependent on Mana Steal. At slower speeds (anything below normal, and especially sslow), I believe as of now that Mana Steal is equally effective across all classes, ignoring the interaction with the Shaman beam split.

Even more worrying for fast hybrid (and tierstack builds in general) is how user cps plays into the number of attacks per second at higher speeds. Although super fast was estimated at around 4/s from our macro trials, prior to that, Highmore did a trial where he clicked a manual cps that ranged from 8-15 (average was around 11.5), and only got 3.44 hits/s with Assassin. It seems as if, without precise clicking (that is likely humanly impossible) or clicking along with a metronome (which highbread did for a trial), maximizing hits/s at high attack speeds is nearly unachievable.


If anyone has any questions about this, let me or any of the AIn members mentioned in the thread know! We might have answers, or it might inspire some more ridiculous testing in the future.

 

sorry for delaying the thread lmao

 

Wynncraft is a perfectly balanced game with no exploits.

 

me when the me when the wynncraft is sus

 

this thread would've come out 2 hours earlier if Saya's PC didnt suck :dome:

we got it out and that's what matters though :holyshot:

 

society is crumbling and so is our understanding of wynncraft

 

Let's find those diamonds

 

Waiting for the PSA prequel that shows where this all began and how did it all happen anyway

 

I'm glad we got everything clarified. Now wynn can be properly understood.




*crash*

Powders as ingredients, thorns and reflection proc, forest aconite rainbow spell stat, etc.: ALLOW US TO INTRODUCE OURSELVES.

 

working as intended

 

The lies are no more.

 

Well, not exactly lol

Still says /4s in game for now, and we only have a very vague idea of attack speed

 

you guys really did just wake up one day thinking "yeah maybe everything we know is a lie, lets test that"


for really though its impressive how you all found this stuff

 

when the atlas inc finds a way to get consistent 32cps "for testing purposes" ::::

 

we do a little (too much) trolling

 

I love how there is no 7-9

 

My life is in shambles now :(

———————

Great work! This is the type of stuff I always look forward to.

 

I set the deadline on the thread for today so that we wouldn't spend weeks on testing everything, but, 7-9 may end up coming not too long from now with all the random stuff people keep finding

 

please no sequel trilogy sucks

 

Does this affect anyone apart from class builders?