What if Noctis were redesigned?

Noctis is different from Xiaogan in that Xiaogan can frontload her damage more. She can also store stacks, and more stacks beyond 3 will boost her damage. However Noctis can only consume 3 droplets at a time for his CA. It is not like he can consume 4, 5, 6, or more to buff his CA (and thus frontload his damage more).

Similarly Alhaitham can frontload his damage by getting 3 mirrors instantly. The more mirrors he has the more damage he does. In HSR, Lunae is a character that can frontload damage by consuming up to 3 sp.

So for Noctis do be more like Xiaogan or Alhaitham, he would need some mechanism to consume more than 3 droplets/stacks. For example, perhaps consuming 4/5/6 droplets will increase the scaling on his CA as well as its dmg bonus/crit/aoe. This way his CA can have different 'levels' similar to Xiaogan or Alhaitham. Absorbing 3 droplets gives a lv0 CA, 4 droplets lv1, 5 droplets lv2, 6 droplets lv3.

Technical note: as for the numbers, it could also be absorbing 3/6/9 droplets. His E creates 3, his Q creates 6.

This can then make supports that give droplets (like HMC) potentially more useful. Droplets can then almost act like skill points in Genshin.

And a support like Sparkle from HSR could be easier to create: Sparkle can generate more droplets (sp) for Noctis. For example, her ultimate can create say 4 droplets (or even 12). And she can increase the number of droplets/sp that exist on field from 15 to say 21. Here we are using the conversion that 1 sp = 3 droplets.

However, the question becomes: can Noctis choose how many droplets/sp to consume? Currently, in Genshin he would just automatically absorb all droplets nearby (up to 3 at a time). It is not like he can only absorb 1 or 2 droplets when a 3rd exists.

Alhaitham's frontloading

As discussed, sp in HSR operate similarly to cooldowns in Genshin. So Alhaitham's C1 of reducing cd essentially allows him to frontload his damage more often (have higher uptime on 2-3 mirrors). It is almost analogous to if Lunae had more sp to spend each turn.

One main difference between Alhaitham and Lunae is that Alhaitham has hyperbloom. Otherwise, if playing Alhaitham in a pure quicken team, his ability to frontload his personal damage (that can crit) is actually similar to Lunae (his E mirror dmg has 3 forms similar to how Lunae's basic attack has 3 forms).

Alhaitham's C6 encourages frontloading damage even more, increasing his crit value when he has more than 3 mirrors. Moreover it always returns 3 mirrors no matter how many he uses for the Q (this is eerily similar to Vicky's C6 which also returns 3 crystal shards to her if she uses her 6 stack E: it is Alexandrina Victoria's legacy in action). This last effect can probably just be a C0 effect: at C0 his Q always gives 3 mirrors (it will just do more damage if he consumes more mirrors for the Q).

But Alhaitham is a perfectly fine character at C0, and as they say his best constellation is Nahida's C2.

But on the theme of designing frontloading characters (with flexibility), we can propose some additional ideas involving Alhaitham's constellations, synthesizing the ideas of Noctis' droplet mechanic and Xiaogan's 3-6 stack mechanic as well.

First let's elaborate on the idea of getting more than just 3 mirrors. Xiaogan has a 3 stack E mechanic like Alhaitham, and she can also get stacks beyond 3, holding up to 6. Each stack beyond 3 will increase her E's dmg bonus.

Alhaitham's C2 currently makes each mirror give more EM, so this is further emphasizing how more mirrors give more damage as Alhaitham converts EM to more dmg bonus and base damage scaling (given his dual scaling). An additional effect could be: each mirror beyond 3 will increase his crit rate/crit damage by X% (max of 6 mirrors), and his Q gives one extra mirror (up to 4 mirrors).

His C4 is some team EM effect, which is fine for hyperbloom. Another idea could be (based on Noctis C4): while he has mirrors, he can gain 1 extra mirror (every 5s or 6s). Or it can be after triggering a reaction he gains 1 more mirror (can happen once every 5-6s).

And then his C6 could make his E mirror damage ignore 60% def or something of the sort. Or it can be something like a Noctis draconic stack effect: for each reaction triggered by the team, his E will ignore 20% def, stacking up to 3 times (so quicken, aggravate, spread, hyperbloom, burgeon, EC, burning can each count towards that stack).

Or if one wants to buff reaction (hyperbloom) damage, the E can shred up to X% res (this would buff both his personal dmg and hyperbloom/burgeon dmg).

Alhaitham could have used something like a crit rate ascension (as he gets enough dmg bonus from his passive). Also his E mirror aoe is already quite good, but his 2-3 mirror hits can probably increased to something at least Xiao plunge level (currently Alhaitham E lv3 has something like 4m radius, whereas Xiao Q high plunge has around 6m radius). But even now, similar to Lunae, the more mirrors Alhaitham has, the more aoe he gets.

At C0, his E could have an added effect: each mirror hit (after the 4th) will increase his crit dmg by 6-13%, up to a max of 4 stacks. This count will be reset when he leaves the field. Or it does not need to be after 4 consecutive hits. It can start with the first mirror hit.

Moreover, each mirror hit will increase his dmg by 5-11%, stacking up to 6 times.

Now all this would greatly increase his personal dmg in quicken teams. But Alhaitham is already a very good character at C0 because he applies dendro fast for hyperbloom teams. Alhaitham's constellations are really for buffing his quicken damage (they do not really buff hyperbloom damage much as that depends more on the electro trigger).

'Geo version' of Alhaitham

Say Lunae is geo in Genshin. His kit would probably be most similar to Alhaitham due to the mirror flexibility mechanic. But if Lunae is geo, his scalings would need to be high enough given that Alhaitham has quicken and hyperbloom which can increase his team dps. Geo dps (and anemo, physical) need to rely on raw scalings as they do not have additive/multiplicative reactions or things like hyperbloom. For example, Xiao can do around 10 nukes each with around 400% scaling during his Q (and then his C6 allows like six E nukes of ~400% scaling per sceond). So a 'geo version' of Alhaitham (or Noctis) would need appropriate scalings to compensate.

At lv10, Alhaitham lv3 E has around 360% atk scaling (3 hits each with around 120% atk scaling, ignoring the EM scaling for now). This is already pretty high without reactions (close to Xiao plunge scaling actually). His lv3 E can spread two times (due to unique 2-hit icd), so this increases his scaling even more (to say something like 500-600% scaling depending on EM). Now hyperbloom would increase his team dps even more (depending on EM of electro trigger, which would usually be around 1k EM for any serious hyperbloom team).

So a geo version of this would probably want something like 200% atk scaling (or equivalent) per mirror hit if they want to compete as geo cannot spread, melt, vape, or proc hyperbloom.

And most of the serious anemo/geo dps have some way to buff their scalings: Xiao has high scaling on plunge, Itto buffs atk by a lot (which effectively buffs his scaling), Scaramouche E gives multiplier to NA scaling.

By contrast at lv10, Noctis CA has around 14% hp scaling per hit, which is around 140% atk scaling per hit. With his draconic stack passive, he can multiply this by up to 1.6 for a bit more than 230% scaling per hit. But even though he can technically vape, because he has icd and applies hydro a bit faster, it is not like he can vape all his CA hits (maybe 2-3) due to icd. Even if in an ideal scenario where every 3rd hit reacts (normal icd), say with someone like Dehya, then the dps increase from vape is around 4/3, which effectively results in a bit over 300% scaling per hit on average.

So a 'geo' version of Noctis would probably want something like 200-300% scaling per hit to account for how geo does not have access to the reactions that hydro has.

Backloading

Perhaps the opposite of frontloading (getting all your damage out at the start, possibly with longer cooldown) is backloading or building up a nuke that will happen later. Examples of backloaded damage include Jing Yuan in HSR. But even in Genshin, the Raiden Shogun (building Chakra stacks) and Eula (burst stacks) are examples of backloading dps. Even Navia is also a backloaded character in the sense that you need to build up crystallize shard stacks if you want her E to do bigger nuke damage. Navia can frontload her damage by instantly using her 2 charges of E, but she can also backload her damage by building stacks for larger E.

So which is better: backloading or frontloading? The Raiden Shogun is a very backloaded character (needs burst energy, stacks for larger burst) and is a very good character meta-wise. So it does not make sense to say one is better than the other. Rather, they are useful for different situations. Perhaps in overworld and few waves of mobs, a frontloader can clear faster, but for longer fights and more waves there may not be much of a difference. In fact to give a backloader an edge their nuke should have larger scaling. For example, Eula's nuke can reach 5000% scaling which is much more (around 3-7 times more) than Navia's frontloaded nuke (700-1400%). This is to compensate for backloaded damage, while giving them an edge in longer fights (as opposed to instantaneous ones).