Finally, the fifth deadeye of the game is here. I’ve been waiting for her for 7 months now. Except she has a few slight issues with her kit (yes I’m saying it straight from the start).
https://preview.redd.it/5zi7odhiwd4b1.png?width=1600&format=png&auto=webp&s=73ca2477d91066c08a463f70924313213396ca9e

Overview

Lunacub is a 5* Deadeye Sniper focusing on Camouflage. With the ability to stay camouflaged nigh-permanently, she is capable of staying deep in the enemy lines to deal some damage before everyone else, making it the next best thing to global range.

Stats

• Offensive stats:

Deadeye Snipers have one of the best ATK stats in the entire game, losing only to Mystic Casters and Crusher Guards.
Similar to all those top ATK branch, they have a pretty slow attack interval to “compensate” at 2.7 seconds, one of the slowest of the game.

• Defensive stats:

I have already spoiled that she has extremely long duration camouflage, which boosts her “defensive” stats by a lot. But even outside of that, her HP is strangely high and is one of the highest of all Snipers, losing only to Spreadshooters, Flingers, and Fiammetta.
Why is her HP that high? Because for some reason they decided to give her +300 HP as trust bonus, on a unit designed to have nigh-permanent Camouflage. It’s the same trust bonus as Andreana – another Deadeye – at +300 HP +40 ATK, so it’s not unprecedented, but why not going for the full ATK like Firewatch.
For a number example, at max level and max module, she gets 1996 HP. That’s extremely high for a ranged unit.
Anyway aside from that her DEF is pretty average for a sniper.

• Cost:

Deadeye Snipers have a fairly high cost for a ST sniper. For Luna, that will be 19 at base and 21 at E1 onward.

Range

Deadeye have one of the best range in the game.
Deadeye range at base (left) and E1 onward (right)
They have the ability to reach a bit further forward than Marksman and 2 tiles further to their side which really enhance their positioning aspect because they’ll usually be able to reach the enemies even with awkward placement. Especially for Lunacub, she has even less of a positioning issue thanks to her Camo.

Trait

Prioritise attacking the enemy with the lowest DEF in range first
The Deadeye’s trait is a weird one. It’s unpredictable unless you have a solid grasp of the enemy’s stats or you are in your second attempt of the map already. It can be helpful because Casters and most ranged enemies tend to have lower DEF than their melee kin so you can take them out earlier. It’s not always true, but it’s a good starting point.
But it’s not all bad. It also prevents them from ever being distracted. There’s a typical strat that they would spawn ranged enemies following behind a defender or an elite which would usually lead to most ST units to aim at the defender instead of the ranged one, even when both are blocked. These snipers however, can snipe out those ranged enemies in advance. The same goes for enemies with Taunt (“more likely to be targeted” or “draws enemy fire” in the game’s description), the Deadeyes ignore that and only care about the DEF stats.
They also count DEF debuff/buff as well. If you have Shamare or something to reduce DEF, those guys are definitely more likely to be targeted by a Deadeye. The reverse case is someone like a Guerilla Herald, who buffs DEF for everyone but themselves and have low DEF. The Deadeyes usually target these Heralds first.
Anyway, the targeting will be annoying to get used to.

Talent: Wilderness Instinct

Available at E1: Gains [Camouflage] while this unit’s skill is not active
At E2: Gains [Camouflage] while this unit’s skill is not active; Attack interval reduced slightly after skill activation

[Camouflage]: Prevents being targeted by enemy attacks if the unit is not blocking (but will still take Splash damage)
This is the main identity for Luna, being a unit with a nigh-permanent Camouflage buff.
As stated, camouflage prevents enemies from targeting the user, but they will still take Splash damage. This Splash damage refer to most attacks that deals damage in an area.
https://i.redd.it/yh87frcrwd4b1.gif
But what’s not expected is that some attacks that looks like it should be doing damage in an AoE still wouldn’t hit camouflaged unit. For examples, Sui’s thunderbolt, Degenbrecher’s area disarm, Dorothy’s… boss thingy that deals AoE damage that split evenly between all allies, etc. My guess is that due to how these skill works, they would still have to find the unit first… or something.
Idk why either
Altars – the cold one from chapter 6 and the true damage one from chapter 7 – for some reason doesn’t affect Camo despite being completely an AoE effect. This is made worse with the fact that the true damage altar actually hit invisible units, the status that is in essence “better Camo”. But this is the only exception so far, since the cold altar doesn’t hit either of those.
I think they just forgot to fix and is now too late to fix
There are also some AoE skills that only activate if the enemy sees someone to attack. This usually means Camouflaged unit can avoid those skills, with the only exception so far is IS#2 Phantom’s AoE attack which can target an ally through Camouflage.
^{-if you don't see a GIF that means I was too lazy to run IS#2 again to get a recording but you can take my words for this}
Her Camouflage is tied to her non-skill state though, so does this mean that if her skill is up, she will be vulnerable again? Nope because she has that extra HP hah. That and her skill is also designed accordingly with this talent, though we will have to wait until the Skill section for that.
Anyway, the second clause of the talent is that her attack interval is reduced when skill is active. This is a 15% reduction to her base amount of 2.7 seconds and is calculated before any Attack Speed (ASPD) buff. Her new attack interval with skill up, without any other ASPD, is 2.3 seconds (in 30 FPS context, that’s 69 frames, which is nice).
Spoiler I guess, but her module level 3 increases this reduction to 25%, which makes her attack interval 2.03 seconds instead.

Skills

• RIIC skills:

Always available – Lazy Cub: When stationed at a Dorm, self morale recovery per hour +0.7
This is not much useful, considering that her other base skill isn’t used constantly either. But this essentially means that Luna will recover morale fast after working in the base.
Available at E2 – Wilderness Survival: When this Operator is assigned to the Workshop to process Skill Summaries, the production rate of byproduct increases by 80%.
Admittedly, this is the highest % chance to get byproduct for Skill Summaries (aside from Nine-Colored Deer). But, like I said, you won’t use this skill and craft skill books that much. As in, eventually, your Skill Summaries acquisition rate will be slow enough that you will only be able to craft once every long while, unless you're actively farming CA-5 (like refresh-with-OP level of actively). Side note, if you have Hellagur, he has access to this skill straight from the get-go.

First skill: Time to Hunt

• Stats at level 7:

ATK +60%
18 seconds duration, 24 SP cost, 8 initial SP
Auto Recovery, manual activation
​

• Masteries:

M3 gives ATK +100%, 20 SP cost, 10 initial SP
​

• Advance Details:

Uhhhh…
This is exactly like ATK Up γ, except with a much shorter duration and cooldown. Presumably, the rotation was reduced with her talent in mind, to minimize the time when she’s vulnerable to attacks.
​

• Usage:

Despite that, I don’t think the shortened cycle is enough to put Lunacub deep in the enemy lines. 18 seconds is a decently long time to get killed, even with her amazing 300 extra HP (no I will not let this go).
Technically, if she’s still vulnerable to attacks when this skill is up, we can fix that by properly timed the skill when there are no more ranged enemies around. But doing that negates the upside of short cooldown.
This skill doesn’t look too promising, not because it’s inherently bad. Rather, it pretty much removes Luna’s main identity unless you’re extremely good with skill timing and all that.
And furthermore, Andreana also has an S1 with the same +100% ATK, with similar attack rate (2.2 seconds vs Luna’s 2.033 assuming both have mod3), and Andreana’s ASPD is always active and not just during skill, and, Andreana has DEA–Ⲭ which gives her potentially even more damage.
You can definitely use it because it is Luna’s only method of gaining even more ATK, so that she can actually kill stronger enemies (or at least almost kill because now the short duration can actually hurt).

Second skill: Umbrial Ambush

• Stats at level 7:

ASPD +110; Gain Camouflage for 5 seconds; Within the skill duration, defeating an enemy grants Camouflage for 7 seconds.
25 seconds duration, 54 SP cost, 25 initial SP
Auto Recovery, manual activation
​

• Masteries:

M3 gives ASPD +140; Camouflage for 8 seconds; Defeating an enemy grants Camouflage for 8 seconds.
50 SP cost, 30 initial SP
​

• Advance Details:

With 110 ASPD and her E2 talent effect, her new attack interval with this skill is 1.1 seconds, or 33 frames in a 30 FPS scale. At M3 with 140 ASPD, that interval is 0.967 second, or 29 frames. And with mod3 M3, that interval is 0.833 second (25 frames).
Luna gains 5 (8) seconds of Camouflage upon using this skill and 7 (8) seconds upon getting a kill during the skill duration. These duration don’t stacks.
That is to say, Luna will only remains in Camouflage for the next 7/8 seconds after getting a kill, regardless of when it happened.
Luna herself has to be the one delivering the killing blow to receive the Camouflage.
​

• Usage:

As anyone would guess, this skill is designed to make Lunacub completely covered with Camouflage while still be able to use skill. As long as she kills something, she will get Camouflage for the next 8 seconds.
But that’s exactly where the issue is. She has to kill things constantly during her skill to achieve this permanent Camouflage. (In a perfect scenario, she would need to kill only 3 enemies if it’s perfectly spaced out every 7.9 seconds.)
She doesn’t get any +ATK with this skill, it’s pure ASPD. Meaning despite her high base ATK, that’s all she wrote. In general, her ATK is similar to a Marksman with +100% ATK, while her skill attack interval is similar to a Marksman (or shorter with mod3). Wait a minute, a Marksman with +100% ATK and Camouflage??? Oh my god RIP Ap…
Jokes aside, this idea forces Luna to only use this skill when there are kill-able mobs around. Which sounds fine, skill issue and all, but because of the lack of ATK it severely reduces the type of mobs that she can kill within 8 seconds, relegating it to mostly trash mobs.
And this is why I have always use “nigh-permanent” when talking about her Camouflage. In theory, she can roll over trash mobs and feed her Camo constantly. In practice, that uhhh… depends on a lot of things: skill issues, wave flow, amount and flow of trash mobs. Basically, even the most skillful player can still find scenarios where they can’t use the skill regardless.
Maybe the 50 SP cost is because you’re not going to be able to use this skill constantly anyway.

Modules

DEA–У

• Stage 1:

HP +120
ATK +55
Trait upgraded: Prioritise attacking the enemy with the lowest DEF in range first; If the attacked enemy remains alive, gain 1 additional SP
Analysis: As with Firewatch’s module, this module seeks to improve her skill rotation. Sadly, HG knew they are going to give her this DEA–У module, so they masterfully “balanced” her primary skill by increasing its SP cost to 50.
Anyway, Lunacub was released during a time when HG had already realised that SP cost should just stay low. With her DEA–У module, her S2 50 SP cost can be brought down to a minimum of about 36? But then, why not give her 40 SP cost and then give her DEA–Ⲭ instead? I know I have discovered that DEA–Ⲭ doesn’t give that much boost at mid-range but something is better than nothing.
Well, that’s enough rant for this section.
You can technically use S1 and have an extremely low downtime though, similar to Firewatch S1 (same SP cost and all), though with a much shorter skill duration. Admittedly, her S2 is her identity skill and this mod helps “fix” the SP cost. So yes, it’s a very important mod for Luna.
Important note: the attacked enemy has to lose HP in order for DEA-У to give SP. That is to say, if the target dodged or negated the damage with Barrier (Big Sad Lock, SSS Sloma tower boss…), these Deadeyes will not get SP.
​

• Stage 2: (stat boosts do not stack, only replace)

HP +145
ATK +63
Talent upgraded: Gains [Camouflage] while this unit’s skill is not active; Attack interval reduced slightly after skill activation (-20%). {Upgraded from -15%}

• Stage 3:

HP +160
ATK +68
Talent upgraded: Gains [Camouflage] while this unit’s skill is not active; Attack interval reduced after skill activation (-25%). {Upgraded from -20%}
Analysis: This is a pretty sizeable boost to her attack interval. As mentioned before, it decreases her S1 attack interval from 2.3 seconds to 2.033 seconds, and her S2M3 attack interval from 0.97 seconds to 0.83.
Module upgrades materials are pretty limited, so in the grand scheme, this module is not worth it because Luna herself is not worth it. But for Lunacub specifically, this module is very good if you want to stick with her. It covers her S2 SP cost, gives her even more attack rate for both skills, and provides her with more ATK which she will definitely need if you use S2.

Concluding thoughts

Lunacub is designed to stay in Camouflage and snipe enemies from really dangerous position. Unless if there are AoE environmental hazards (altars excluded) or enemies that deal AoE damage without needing to target anything (like the mushroom from SN). If she can feed off of mobs, she can stay in Camouflage permanently

• S1 is her only +ATK source. This makes it worth a consideration when Luna is not in danger but you need more damage to kill anything other than trash mobs. That in turns make her unique talent irrelevant, or at least half-irrelevant, and is essentially a side-grade Andreana S1. That sounds sad doesn’t it. You can use it while abusing the camo off-skill (and that +300HP) to dodge the low traffic ranged attacks (why I said half-irrelevant), but if you are at that strategical prowess already, you don’t need this guide.
• S2 is Luna’s identity. She can go extremely aggressive because of her Camo. With some fodder, Luna can stay almost permanently in Camouflage. The lack of any ATK boost limits the kind of mobs that she can kill, so she’s relegated to mostly weak fodders. She also prefers to be placed away from other DPS or it’ll be a struggle of kill to get Camo. The only major issue is AoE damage (and reveal but we haven’t seen those enemies for a long time until the recent floor 3 boss map in IS#3: Requiem Aeternam and Eternal Wrath, 2 same maps but different boss).

For most people, their first comparison to Luna is Firewatch. Both are Deadeyes that revolves around staying undetected. But Firewatch is far more powerful. Invisible is a definite step above Camouflage. Her Invisible has 50 seconds duration with a downtime of around 15-20. She deals bonus damage to ranged enemies, removing the need for Invi/Camo, though she may or may not target them first. Luna can have permanent camo, though that will be without skill or pray that she kill enemies properly with S2. It is much easier for Luna to be placed aggressively compared to Firewatch, but Firewatch gives a much better reward if you can make it.
Another comparison is someone I made a joke about way earlier: April. Both are Camo Sniper that revolves around being capable of going deep in the enemy’s line and kill something (maybe). With both of their S2 up, they have similar stats: 1200 ATK and about 1 second attack interval. The advantage here isn’t any skill rotation or DPS or whatever, but that April has guaranteed 26 seconds Camo duration whereas Luna needs to get kills. This guaranteed duration allows her to spend time killing (or at least wounding) the stronger enemies unlike Luna who only have a third of that time. Also, April only needs that deployment slot when she’s ready, while Luna has to hog it the entire time. April has an easier time moving around the map. Luna’s only advantages are range, which are rarely seen and/or are not hard to work around, and more importantly her off-skill state for long duration fighting (like I said, something is better than nothing), for example against global range mechanics like Faust or Talulah.
But those are comparisons made while focusing on their innate mechanics of Invi/Camo. What about their designed role? Who else is designed to be stationed deep in the enemies line, having specific talent that will allow them to stay alive in the thick of things, to deal damage before anyone else? That’s right, it is time to talk about Kira…
Throughout this whole thing you might have noticed that I have been constantly saying that Luna is incapable of killing anything but trash mobs. That’s technically not true of course. But given the design of Luna’s kit, her most unique role is to stay undetected behind the enemy’s line to deal damage, and in order to do that she needs to kill an enemy every 7.9 seconds (btw I hope you understand why I keep saying 7.9 instead of 8).
Given this time restriction and the fact that she gets no extra ATK boost, the only thing she can kill within that time are basic mobs. Basic here actually covers a wide array of weak fodders: slugs, doggos, literal basic mobs, their archer version, and their caster version, with some occasional drones. So as much as I keep saying “limited” to trash mobs, it’s a bigger list than it seems. The downside here is that it’s trash mobs, there are plenty of ways of killing trash mobs that don’t involve even 6*.
If you are not using her unique talent, you are essentially using the worst Deadeye Sniper, if not just one of the worst 5* Snipers, with either of her skills (except for maybe Firewatch S1 against pure melee but she has an easy fix for that). No I’m not kidding. I’m not saying you are wrong to prefer using her. I’m just saying to expect accordingly.
So plop her deep in the enemy lines, time her skill correctly, and you can have a unit that filters out trash mobs before they reach your frontline. That’s her best and only value.

Outro

So uhh, how’s your 10 dupe Lunacub doing? Anyway, I really love the Deadeye branch in general, so I was kinda thrilled to see a new one announced 7 months ago on CN server. Too bad what we ended up getting is a ST Kirara but thankfully can actually survive and kill things. Still a great addition to my team though. Anyway thanks for reading I guess?
​
Sellout
All other guides in this post here.

Hello,
I am leveling my mage as fire, and i noticed what I consider a bug in the ignite damage calculation.
Ignite is supposed to deal lets say 5% damage of a given spell in 8 seconds, with its duration refreshed upon any new application, by of course adding the damage to the ignite debuff.
This means thats, whatever the scenario of the fight, if the ignite duration is lets say 60 seconds, based on 1 million damage done with direct critics spell, the damage should be 1M*5% = 50k damage.
Thats exactly whats written in the tooltip.

I noticed in does more (a VERY LARGE amount more) damage, why ?

When you refresh ignite, the damage per tick is calculated at the current time t, which corresponds to whatever have been applied the last 8 seconds + the new ignite (5% of last spell crit) .. but it will last for 8 seconds !!!
Every seconds, ignite damage SHOULD be recalculated with based on past stack, but it does not.
This means that with a high ignite tic, casting more spell results in a net (LARGE) dps LOSS.

example (screenshot (in french), on a dummy, all spells are crit, 18,4% mastery):

total direct damage = 44,4 + 16,7 + 8,7 = 69,8i should do 69,8 * 18,4% = 12,84k.
I did 45,5k: thats 254.36 % MORE damage than it should be.
​
https://preview.redd.it/lcw7pkyoi24b1.png?width=1061&format=png&auto=webp&s=0695f85ed22f08d1e17105effae3a6c68a98af20

The worst part ?

At some point my ignite does, lets say, 1K damage per second.If i do nothing (literally dont touch my keyboard), the ignite will do 1k*8 = 8k damage (as I said, debuff damage is NOT being recalculated by the server unless you apply a new ignite).
If I cast a new spell, it will be refreshed with a ~random~ (it is not, but you will see why I say its random), lets say 200 damage per tic, for 8 seconds.
For this case (which just happened on a dumy), I will do, 1k (first second) + 200*7 = 2400damage, instead of 8K, a loss of 70% of ignite damage fot the next 8 seconds, at this point is it worth it to cast a new spell ?
​

Why is it insanely bad design ?

Because that means that at every second, an optimal scenario is possible, and that scenario is impossible to calculate as a human (you need to record every direct crit done in the last 8 seconds, do a very big calculation), which will tell you wheter or not it is worth to cast a new spell.
Personnaly, I am NOT rain main, I can't do 6 figures calculation 10 times per seconds in my head to calculate wether or not I should cast a spell, or just jerk off for 5 seconds.
​
Am I missing something, or blizzard dev skip calculs course in their uni ?
​
Edit : also, notice that all the figures in my post are directly linked to the way I test, if I had have press spells faster or slower, the damage would have been different. I had 240% more damage, but it could have been 30%, or 150%, or 300%, it all depends of at what exact time I press new spell. There is -obviously- a cap to this, which can be reached after 8seconds of non stop crit of pyroblast launched the fastest way possible.

So I went out and find the formula and did some calculations. She needs around 111% effect hit rate to have some consistency (90% success rate) BUT if you get her eidolon that reduce 20% effect RES than you only need 60%+. The difference is huge (as much as body relic Mainstats) and if they don't make any more changes. That one Eidolon will be the difference between Support and Sub DPS

Hey guys. I'm just making a post after going to multiple threads where I wrote essentially the same thing and I figured a lot of people are just confused about Aphelios' skill points and how to properly allocate your skill points as you level up.

Background:

In this thread I will explain some things and show some math behind Aphelios and his skill points that he can put into one of three stats; Q, W, or E up to 6 times each.Q = 4.5 ADW = 9% attack speed with a 0.64 AS ratio ( resulting in 0.0576 or 5.8% more attacks per second)E = 5.5 lethality which can give flat armor pen from 3.42 at level 1 all the way up to all 5.5 at level 18.
The formula for how much flat armor pen you get from each point of lethality is as follows:
FLAT ARMOR PEN = LETHALITY × (0.6 + 0.4 × level ÷ 18) resulting in the table below I pulled from the armor penetration wiki.
Figure 1: % of lethality you get per point per level

Armor Explained:

While AD and AS are a lot easier to understand, to truly understand how lethality and in turn, how armor penetration works, we also have to understand how armor works. Just to keep it simple here is a chart I made in excel of the damage mitigated by certain armor values. The steeper the curve, the more % dmg we get out of penetrating armor at that point on the curve. The flatter the curve, the less flat penetration matters.
Figure 2: Damage Mitigated by Armor
Using the values from the graph above, I can show what armor penetration does in terms reducing the damage mitigated, which essentially just means increasing the damage dealt, for specific values of either flat or percentage based armor penetration.
Figure 3: % More Damage from Various Armor Pen Numbers
Since lethality scales with level, I picked values that are easy to show on the graph, and roughly translate to values you could get in game. 15 flat armor penetration can roughly translate to 18 lethality (think collector or Youmuu's) at levels 10-11, and 5 armor penetration can roughly translate to 5.5 lethality, or a point in your E at levels 13-14. Armor penetration by % is not an estimate as it's directly purchasable and doesn't scale with level. I will talk more about lethality and armor penetration later but this is all you need to understand before going forward.

Optimal Start for Aphelios:

Alright now that we have the background out of the way let's talk about level 1 and his optimal start. Let's just assume some things to start so that we can simplify a few things.
First of all since Aphelios has no abilities level 1 you can calculate his DPS easily by just multiplying his AD by his AS. Secondly Level 1 E is bad value because you don't have any other sources of armor pen at 500g and you do not have any base dmg's from abilities to take advantage of the % bonus dmg you would get from negating armor level 1. Additionally the DPS calculations would be very messy because we'd have to know the champion you're hitting's base armor and how many armor shards they took so let's just avoid that entirely to make this easier.
Taking those into account we can start a relatively easy comparison between the different starts.
Given that you start cull (most common starting item), take legend bloodline (instead of alacrity), and take absolute focus (most common runes), I've created this DPS chart for level 1 auto attacks.
Figure 4: DPS chart for various rune shard/skill point choices.
As you can see, the overall winner is D. Now the only issue with option D is that despite it having the highest DPS at level 1, a lot of people don't like the idea of having less AD early because it'd make their abilities weaker going into level 2 and it'd be harder to last hit than they're used to. The second greatest build is A, just slightly ahead of the the typical start most Aphelios players go for. It is worth noting that option A has comparable DPS with a whole extra 5.4 AD when compared to option D. Let's see how the two best starts transfer into level 2.
Figure 5: Surprising?
Unfortunately at this point it becomes a lot more difficult to accurately decide since your own abilities are thrown into the mix, all of which scale more heavily with AD and lethality than AS. If we were ignore those attributes as before we'd actually get that W level 2 is more DPS as well, regardless if we do starting build D or A.
I'm pretty sure most Aphelios players here would agree that putting points in Q would be optimal for levels 2 and 3 since you most likely aren't able to recall and the AD makes all your auto attacks AND abilities deal more damage, while the points in W are barely higher in DPS, and come at the expense of ability damage. I personally run A1 in every single one of my games and I would recommend everybody else to do so, and I continue to put points into Q until level 4 to maximize my early ability damage, however once we hit level 5 or 6 we should all really be thinking about what we're trying to do in the game and what our stats are doing for us.

When to Q or W:

To just keep on the simple topic of a point in Q vs a point in W, if we want to maximize auto attack dps, what we have to keep in mind is which of these two values is greater than the other.
TO MAXIMIZE PURE AA DPS FOLLOW:

Q: 4.5 * (Current attacks per second)	W: 5.7% of AD

Here's a small image to help illustrate my point here:
Figure 6: Why Q vs W matters even when you have the same DPS initially.
The more complicated question is when will I want AD over AS for increasing my actual in game damage? What if I can't auto attack that often? What if I just want to land massive 5 man blue ultimate's and 1 shot everybody? What if I've built a few on hit items? This is when you should think about the context of the game and how you want to play it. 5.7% of your current AD is only the value of the point in W if you're auto attacking as often as you can. The 4.5 AD is a lot more universal as it applies to more than just your auto attacks so in the general case you can adjust the 5.7% based on how frequently you think you can consistently auto attack in fights.
In games vs a lot of mobile assassins, it's likely that the point in W will never be worth it as the most amount of hits you can get off are one or two autos, a Q, and your ult. This would be a situation where the value of the attack speed is closer to 0% of your total AD. Against a front to back team comp though, you'll see Aphelios players with 300AD, who have built a kraken slayer and Runaan's, putting another point in Q just to "max it" when 5.7% of their AD would be way more DPS especially with the on hit damage. When using Chakram, (white) attack speed is also heavily depreciated in value as the attacks per second is determined far more by your distance to your target, and if you crit, than bonus attack speed as well as the mirror chakram essentially functioning as a multiplier to your AD per AA.
Since the value is of attack speed is context dependent, and it's not very often that you're actually just auto attacking without using any spells, I think in a more practical scenario we can substantially reduce the value of attack speed, while still recognizing that it can still be more valuable than more AD if you have enough AD already. My general rule of thumb for my own play is that I'd be casting spells a bit over half of the time that I'm auto attacking, so my adjusted average value that a point in W gives me is about 2.5% of my AD. Taking that into account here are the new values to optimize stat choice.
FOR GENERAL USAGE HERE IS A RULE OF THUMB TO FOLLOW FOR DECENT STAT SELECTION:

Q: 4.5 * (Current attacks per second)	W: 2.5% of AD

All of that was just to talk about Q vs W but we have a whole other stat to talk about which is lethality, Aphelios' E. I won't do as much math unless it's requested for this topic in the interest of saving space but if you want me to follow up and show some more detailed math please let me know.

E, Armor Penetration, and Lethality:

Lethality, as previously demonstrated in Armor Explained: Figure 3, is best received when the opponent has low armor. Since lethality is more valuable when the opponent has low armor, the more armor penetration you have, the lower their armor, and therefore the more valuable each point in E becomes. Since lethality scales with itself and other forms of armor penetration so dramatically here is a general rule to follow for points in E.

• If you are in a situation where it's most valuable to put a single point in E, it is very likely you should continue to put points into E until you can't anymore.
• If you are already building lethality in a rush item (such as Youmuu's, serrated dirk, collector) you should be enhancing that lethality with points in E from levels 3-9 and should consider a heavier AD start (option A or C from Figure 4).
• Do not put points into E if you are already low on damage as points into your E only enhance the damage you're already dealing and especially do not put points in E if you are doing it to penetrate the armor of a 300+ armor tank.

Now you might be wondering, when should I start to put points into my E? Well just like before, it's a question of value for every point, at every level up. It's a bit harder as the % benefit changes based on the armor of the opponent, what other forms of armor penetration you've built, etc., so I won't focus on an exact number but rather some rules that I will list below.

1. You have a decent amount of AD (120+)
2. Are in a situation where you can't reasonably auto attack that often or don't need to because you're against squishy champions which can be bursted quickly with abilities.
3. Are planning on building an armor pen item soon or you are already building it, UNLESS the enemies you care about have 300+ armor.

Using these rules I can create a new general rule of thumb for what to put your skill point into:

Q: 4.5*(Attacks per second)	W: 2.5% of AD	E: 2 out of the 3 rules are met

This is already mentioned in rule 3 but points in E should be avoided if the enemy team is armor stacking. Flat pen is less valuable than your other stats regardless once they get enough armor.
Once you finished going down the lethality route, it's important to realize that your abilities will all deal substantially more damage, meaning you'd want more AD, but please recognize that it is still possible that a point in W could be more valuable, so just fall back onto the rules of thumb for Q vs W at that point.

In Conclusion:

While this is meant to be a general guide on when to put points into your Q, W, or E, I also want to point out that this doesn't disprove that certain stats are generally more optimal than others. There is not much harm in just maxing whatever stat you prefer to max, or following a simple rule such as Q>E>W, however the main reason I made this is so that we can all recognize one thing.
We are Aphelios mains. We do not max abilities like everybody else because we're stuck to just one "better" ability to function. We have 5 guns and 3 stats to choose from and I want us to stop asking each other which one to max in every situation or for which build. Depending on your items, or what the opponents build or pick, or what your team picks, your stat selection should change frequently. Our flexibility is a strength. We must recognize there is no reason to MAX anything at any moment. It's all about picking the optimal stat for the moment you need it, which can change at any moment.
Aphelios is not like other champions, we shouldn't talk or think like other champion mains. Thanks for reading.
Tl;dr; Stop maxing a stat because somebody told you to. There is no benefit to maxing anything, just get your stats when they're most effective. Never feel trapped to max anything in any order.
Edit: Just adding this post at the end as a footnote that pro players are starting to catch on. He doesn't max any stat by level 9 and instead chooses what he thinks is optimal for his DPS early on, and also commits to the rule of thumb for lethality which is to dump points into it once you start putting any points into it. Also he has 6 points in lethality at level 13 which is when the base damages for Aphelios are at their highest, which is perfect timing. Please check it out!