Which elements are the strongest?

[Liliat]

il y a 2 minutes, Horsting a dit :

Also +10% crit damage in above assumptions means more than +1% damage when taking flat reduction through defence into account

Not quite: the crit damage multiplier applies to both your own attack and your opponent's defense (it's counterintuitive), because Crit multiplier is applied on Damage, not Attack. If you do x2 crit damage at 140k attack vs 40k defense, you get 2x(140k-40k) = 200k damage, which is exactly x2 compared to the non-crit attack (140k-40k = 100k).

il y a 4 minutes, Horsting a dit :

Only the relative value, the absolute bonus remains the same (as long as the opponents defence is not higher than your attack).

Correct

[Horsting]

6 minutes ago, Liliat said:

Crit multiplier is applied on Damage, not Attack. If you do x2 crit damage at 140k attack vs 40k defense, you get 2x(140k-40k) = 200k damage

Oh, that is counter-intuitive indeed, many thanks for clarifying this. Just to be sure: This applies to crits themselves, not only the extra crit damage from red, so also orange becomes weaker compared to dark because of this, right? I'll also watch and verify the next challenges.

Edited June 1, 2023 by Horsting

[Liliat]

il y a 18 minutes, Horsting a dit :

This applies to crits themselves, not only the extra crit damage from red

Yes, this applies to all attributes that affect crit damage: the base +100%, the extra +35% red passive, and the extra +70% red team (I don't think there are other sources).

So that indeed makes all crits "a little bit weaker", in that they will never really help you breach a high defense (If your attack is equal to or lower than the opponent's defense, you will always do 0 damage, even with crits).

[DvDivXXX]

38 minutes ago, Horsting said:

Just to be sure: This applies to crits themselves, not only the extra crit damage from red, so also orange becomes weaker compared to dark because of this, right? I'll also watch and verify the next challenges.

Yes.

In general Orange is weaker than Dark or Blue because even though it's a significant buff to your crit chance, it's just that. On rounds where you don't roll a crit, Orange still effectively does nothing, even though of course you get crits more often thanks to Orange. All other effects apply all the time, only crit chance (Orange), crit damage (Red) and of course slight crit chance boost slash slight debuff to your opponent's crit chance (Purple) are RNG dependent and not a permanent buff every round, some of which get even better on crit turns.

Regardless of Red, your own crit turns are improved by Dark, Blue, and technically also Yellow (proportionally at least). Your opponent's crit turns are mitigated by Green and technically also White (proportionally as well). Also, as long as you don't lose on the spot when your opponent crits, your next turn's recovery is greatly improved by Blue times Dark (also helped by Yellow a bit). I still rate Orange highly because it's a significant boost to your crit chance, and more crits is great.

It's worth noting that there's some anti-synergy between Blue and Green, for scoring the most points in a sure win especially: Blue heals you based on the damage you deal, not your max HP, but Green increases your max HP. And the closer you are to your full max HP when you win the fight, the closer you get to the maximum points you can get. That's why Green tends to fall off at high level, although it's super useful when you're not guaranteed to win the fight. Likewise, the higher everything else (your own player level, your equipment, boosters, your girls' stars, levels, blessings, crit chance etc.) the stronger Dark and Blue get.

Overall, that's the gist of my version of the tier list. There are more nuances here and there and it varies as you go on your journey through the game, like many other things (for instance chlorellas can be serviceable or even good in the early to mid game, but they're really nothing more than fodder to sell when a tasks requires it by the late late game; conversely, cordyceps or AME are not worth using until the late game, where they truly shine: better to stash them early on etc.).

Edited June 1, 2023 by DvDivXXX
updated

[Horsting]

3 hours ago, DvDivXXX said:

In general Orange is weaker than Dark or Blue because even though it's a significant buff to your crit chance, it's just that. On rounds where you don't roll a crit, Orange still effectively does nothing, even though of course you get crits more often thanks to Orange.

The fact that crits and crit bonuses do not always apply makes then IMO less sexy, as I like smaller outcome margins and higher certainty, but this alone does not make orange or red weaker than dark (and relatively weaker than yellow) for the average rating. That crits and hence crit bonuses apply to the resulting damage instead of the attack, and hence the defence reduction applies by the crit multiplier as well, however makes then indeed weaker.

3 hours ago, DvDivXXX said:

I still rate Orange highly because it's a significant boost to your crit chance, and more crits is great.

In numbers:

• With 100.000 attack, +5% crit damage and an opponent with 30.000 defence, +2% crit chance means (100.000-30.000)x1.05x0.02=+1470 damage in average.
• Assuming 20% crit change, +10% crit damage in this scenario means (100.000-30.000)x0.1x0.2=+1400 damage in average. This is in my case with 2 orange girls, but one of them, a common 5 star, is only there because of current blessings, while usually I guess one does not have more than 1 orange girl in any team but the orange counter team. So usually the bonus is less. BUT with +20% counter bonus crit chance, the bonus of course raises again by that 1400 damage, so then red becomes quite strong.
• Same scenario, dark +2% attack means (100.000+70.000*1.05*0.2)*0.02=+2294 damage in average, +2000 on every hit, +1470 on crits, which are the +294 in average. From 33% crit chance on, red bonus surpasses dark (in average), so in a dark-white-purple counter team, against opponents with significantly lower harmony + some orange bonus (with harmony alone, 29% crit chance is max), red at least should be at least tested/compared with dark (if similar power and otherwise exchangeable) on individual fights.

4 hours ago, DvDivXXX said:

It's worth noting that there's some anti-synergy between Blue and Green, for scoring the most points in a sure win especially: Blue heals you based on the damage you deal, not your max HP, but Green increases your max HP. And the closer you are to your full max HP when you win the fight, the closer you get to the maximum points you can get.

That is not an anti-synergy, it is just no synergy (while any kind of damage bonus does have a synergy with blue). Note that more ego also means that the opponent isn't able to reduce your ego percent as much, by the exact percentage that your relative healing bonus through blue is reduced. So these two bonuses to not affect each other for their average rating. I think it is generally better to compare absolute values based on an example: If your opponent deals 100.000 damage and your are able to heal 20.000 via blue bonus, more or less ego does not change something about that.

I would say one can best compare green with damage and defence bonuses based on amount of rounds and hence on overall amount of damage increase, respectively overall opponent damage reduction, compared with the ego bonus. So with 6 rounds, the absolute ego bonus must be higher than 6 times the absolute damage bonus and 5 times the absolute defence bonus (if you win in round 6) to be compatible. Probably I should write some spreadsheet with a reasonable example (one which matches most player scenarios as good as possible), apply this math and have the result discussed for a possible update of the performance handbook 🙂.

[DvDivXXX]

3 hours ago, Horsting said:

That is not an anti-synergy, it is just no synergy (while any kind of damage bonus does have a synergy with blue). Note that more ego also means that the opponent isn't able to reduce your ego percent as much, by the exact percentage that your relative healing bonus through blue is reduced. So these two bonuses to not affect each other for their average rating. I think it is generally better to compare absolute values based on an example: If your opponent deals 100.000 damage and your are able to heal 20.000 via blue bonus, more or less ego does not change something about that.

Green makes your total HP a higher number. Blue heals you for a percentage of the damage you deal. The extra points you get for a victory are based on the percentage of your total HP you still have when the fight ends. So the smaller your total HP pool, the bigger the proportional impact of Blue on it, and the more points you can squeeze from a clear win. In that context, adding more Green bonus reduces your average score. Again, that's for late late game high-level PvP. When fairly much everybody has their passive bonuses maxed out for all 8 colors anyway, and the variations I'm talking about are for active bonuses from your team instead.

As I said, Green really shines in harder fights where your main goal is to avoid losing, not to fine-tune your sure wins to increase your average score your league week. That happens less and less as your account grows in power and you climb up the divisions, though. Green is one of the best colors in the early to mid game, and it doesn't have a real synergy issue with Blue at that point. But it eventually becomes less useful and even something to limit if possible when you get to the point where you'll almost never lose a fight and the main focus becomes scoring as much as possible.

Overall, these aspects balance out to make Green one of the stronger elements in the game as a whole in my eyes (next to Orange behind the two god-tier ones).

Dark is unquestionably much stronger than Red in any realistic situation, I assure you. I think you might be overestimating kind of theoretical situations or "on average" things that don't translate as well as one might think neither in the average battle situation, nor in the wider context of weekly PvP ranking system based on your average score compared to the rest of the playing field sample of ~100 players you're pitted against. Dark is a sure thing all of the time, and it scales with crits when they do happen. Red only adds a bit more to crits and nothing at all most of the turns. And most turns in most battles are without crits. Even if Red made your actual attack triple rather than double in case of a crit, it wouldn't make it as valuable as Dark in most individual battles, due to the randomness of getting a crit or not and getting it where it makes a big difference for the outcome or not. Which in turns would make Dark better than Red for week-long competitions as we have them.

In certain specific battles, all of them PvE, rules and circumstances vary enough as to change the order of things. High-level Pantheon bosses will crit all the time and you'll crit every other blue moon no matter how much Orange or Purple you have, so you're better off ignoring crit chance entirely. Red matters even less in that context, obviously. Boss Bang runs, as soon as your reach the account power threshold where that becomes viable, are always the easiest with a full Blue team barring anything else, because you can't heal between fights and the 31% lifesteal maxed-out passive and active Blue bonuses solve that problem and then some. The tier list and global rankings of elements overall are generally aimed at PvP, where there's more room for each bonus type to shine (even if Blessings and other weekly variables often have an even bigger impact and we're typically judging elements against each other in a vacuum, or say a week with very small blessings that allow for basically all combinations in theory).

Anyway, this has been a fun exchange. We don't have to agree on everything and your perspective may change over time anyway (mine too, I'm open to new things and approaches, but as a newer player you have a lot more of variations and transitions ahead of you). As far as general guidelines and evaluations go, I feel we went above and beyond expectations already. Good luck have fun!

[Horsting]

17 minutes ago, DvDivXXX said:

Green makes your total HP a higher number. Blue heals you for a percentage of the damage you deal. The extra points you get for a victory are based on the percentage of your total HP you still have when the fight ends. So the smaller your total HP pool, the bigger the proportional impact of Blue on it, and the more points you can squeeze from a clear win. In that context, adding more Green bonus reduces your average score.

There is a quirk in this logic. You do not get more damage from your opponent only because you have more ego. With more ego, the opponents damage reduces your ego by a smaller percentage as well. So even that blue does not heal as much relative to your total ego, since it heals the same absolute ego, it has no weaker effect. With more green you will end the fight with more absolute and more relative ego and hence get more points if a 10% breakpoint is reached. Thinking in proportional values here (for damage/healing/defence etc) leads to false logic: More damage (from you) means more absolute healing via blue, so that is a positive synergy. More ego has zero effect on the amount of healing, so there is zero synergy, neither positive, nor negative. But this does not mean that other colours, which do have a positive synergy or work differently, become better compared to green in endgame.

23 minutes ago, DvDivXXX said:

"on average" things that don't translate as well as one might think

The "on average" things are all that counts for a generic rating of these bonuses and when doing strategic decisions like which girl to max out first. If in a certain team, red has a stronger average effect than dark, then you will reach higher expected scores against most opponents (E[x] value of HH++), while of course on an individual opponent it may be different. But probably in above math I missed an aspect. And the circumstance in which these 33% crit chance apply are rare anyway, so I do not say that it is good to focus on red, just saying that it is possible that it is stronger than dark in a certain team against opponents with low harmony and/or small cycle counter team. I currently do not have a strong reg girl and/or any weak dark or purple opponent in the league to show a good example, but will keep my eyes open, just out of curiosity 😊.

57 minutes ago, DvDivXXX said:

We don't have to agree on everything and your perspective may change over time anyway (mine too, I'm open to new things and approaches, but as a newer player you have a lot more of variations and transitions ahead of you). As far as general guidelines and evaluations go

Definitely. I do not even have a fixed perspective here, just doing some rough math based on assumptions, which may be wrong or the formulas lack important aspects. Trying to find a way to understand and generate a better generic colour rating, which does not imply simulating every possible outcome of a fight (like HH++ does).

[DvDivXXX]

I know you're more in the spreadsheet formula world than in the day to day effects and interactions in the game itself (you couldn't draw any meaningful conclusion from the latter right now even if you tried, as the early game is so different and incomplete from the late game fuller and more stable picture). That's a big part of why I'm stopping there. I was sharing my years-long high-level PvP experience more than anything else (also informed by reading, chatting and testing out many different theories, ideas and strategies from other players over the years, including some amount of spreadsheet and maths-based stuff, but I always take that with a pinch of salt). YMMV.

Edited June 2, 2023 by DvDivXXX

[Horsting]

I do not say you are wrong regarding your experience and conclusion, but you cannot say that the method of (trying to) using formulas and statistics over individual daily experience is wrong or inferior. Statistics do not lie, all the technology that drive your daily life strictly depend on formulas and statistics, never on individual experience. But what can be wrong is the formula itself (wrong assumptions, missing aspect, mine is most likely too simplified), how statistics are derived from it (false logic, typos, ...) or the conclusions you derive from it. And conclusions based on daily experience may be correct while the derived explanation may be wrong 😉.

Btw, I'm discussion friendly, but I never mean it disrespectful or offensive ✌️🙂.

The HH++ battle simulator is a great resource as well, which verifies also the logic we discussed: https://github.com/zoop0kemon/hh-plus-plus/blob/a157a1f/src/modules/BattleSimulatorModule/Simulator.js#L21
Could be adapted to run some constructed battles, vary parameters to catch all results between breakpoints, and find/adjust formulas which match the average of the simulator results, to finally create a spreadsheet based on those formulas with some sensitivity analysis for all bonuses. Or giving up on such analytic formulas and build a little website with player + opponent inputs which embeds the HH++ simulator to show exact results/probabilities + sensitivity analysis 🤔. Such or something similar does not exist already, does it?

[DvDivXXX]

On 6/2/2023 at 4:42 PM, Horsting said:

Btw, I'm discussion friendly, but I never mean it disrespectful or offensive ✌️🙂.

Me neither, nor did I perceive this back and forth this way. Your thirst for knowledge about the game is nice. It's just that we already went way deeper than I had in quite awhile (and these notions are long-established for me, although I keep an open eye for new perspectives and ideas) I don't want to go in circles. Perhaps most importantly, you've started this quest by mentioning you've only just started playing the game less than a month ago, and this game is a years-long marathon (except for people with massive amounts of disposable income they're willing to throw at the game to fast-forward the first few years of progress, which has only been possible for about a year or two and we've already seen quite a bit). Apart from that, for a regular player (F2P or with a small, occasional budget), the early game alone is still the first couple of years, I'd say. It's been awhile since I've made a new account to see what the new player experience became, admittedly, and the game has changed and keeps changing in many aspects, some changes that made the very early game (or the tutorial phase, as I'd call it) much quicker, but mostly changes that moved the milestones much farther out of reach for what I'd call mid game, late game and very late game. So, to me you're still in or barely out of the tutorial right now. It's great to have a long-term strategy and focus as early as possible, but only if you stick around for years to see any of it come to fruition. If you get bored with the game in a couple of months and end up quitting, then all that theoretical data and spreadsheets won't be useful to you anymore (unless, as I half-suspect, it's a fun side quest for you to do in and of itself).

On 6/2/2023 at 4:42 PM, Horsting said:

Such or something similar does not exist already, does it?

Yes and no. This here forum and some parts of the official Discord (and a number of private Discords) would be the closest. Tons of players have shared and are still sharing tons of data, assessing and adjusting odds for most things in the game (that are almost never clearly disclosed by Kinkoid but instead always figured out by the community), experiences and ideas for many years. We don't exactly have a separate spreadsheet with battle sims though. I wouldn't be very interested in that personally, but some might. The issue with such a project, as previously mentioned, is that the battle simulator does a great job at approximating and averaging your odds to win a specific fight in real time, as well as the expected average in terms of points (for league fights). It's not the be-all, end-all of all the situations you might encounter and it's not very useful in a vaccum either. Blessings can and regularly do have a much bigger impact on your odds to win and to score high than the individual active bonuses from your team members do. And more and more variables come into play (with the very recent and terrible addition of equipment for girls, more than ever, but I don't doubt we'll keep seeing more and more game-changing updates in years to come). So a purely theoretical spreadsheet of simulator results wouldn't be very useful for anyone to draw any applicable conclusions on anything, let alone worth it compared to the level of understanding and the relatively stable consensus we already largely have on elemental bonuses. IMHO.

But if you feel like doing that, knock yourself out. Good luck have fun either way.

[Horsting]

51 minutes ago, DvDivXXX said:

Perhaps most importantly, you've started this quest by mentioning you've only just started playing the game less than a month ago

Only because I started the game a month ago does not mean that I am not able to understand the mechanics behind it, and event less whether I am able to find suiting analytic formulas (or instead write an interactive simulator) to calculate average benefits or not 😉. What I am missing is endgame "data", which is the assumptions I made in the calculations above, which of course must be varied. E.g. I also do not know the caps for passive colour bonuses of girls which are not in your team. Those do not raise unlimited as I understood?

Practically, a spreadsheet/simulator would be fed by individual players with their stats and optionally opponent stats, and depending on short-term/long-term questions excluding current or including next blessings (not so trivial to do manually, actually 🤔). It would then show you the result based on battle simulation or analytic calculation + some sensitivity analysis, i.e. showing you results when individual stats vary by a certain degree in both directions, to cover break points. To come back to the actual topic: Based on this, the effect of colour bonuses can then also be derived, of course always only based on the input player/opponent data, never as absolute rating for everyone. For the performance handbook, if someone can update it, or a new thread to point players at, experienced players would then need to feed the thing with varying reasonable data, situations and stats which are common in different game states, and hopefully from that a helpful picture/ranking can be derived. For the individual fight, HH++ E[W] is the only thing that matters anyway.

As you probably recognised, it's fun for me to tinker with such kind of questions, so even if I would quit the game before finishing anything useful, I wouldn't see it as a waste of time. I mean, isn't clicking on some buttons to generate some pictures and numbers stored as bits on some server out there, a waste of time as well then? 😅
Solving tasks is probably what makes things interesting, and for me this includes answering questions by trying to analyse and understand mechanics, then optimise based on it.