Jump to content

Which elements are the strongest?


Master-17
 Share

Recommended Posts

This post contains the following image, showing the distribution of the elements of girls by strength, according to the author of the post.:

daadadad.png

However, I got different information. Trying to pick up a rainbow team instead of a dominant one, as I did when putting up the strongest girls at the moment, I changed the last one in the team among equal in strength, but different in element. Together with the script data showing the probability of winning against the same opponent, the following sequence was obtained (from the strongest to the weakest):

1 - Dominatrix

2 - Physical 

3 - Sensual

4 - Exhibitionist

5 - Submissive (more chances of victory but less average reward) 

6 - Eccentric

7 - Playful

8 - Voyeur

Here are the teams and odds:

1.thumb.jpg.2ad82f554d8ad277a336d61939ded02e.jpg

There are no dominatrix here, but I added a team with a counterbonus.

For the sake of the purity of the experiment, I made the same comparisons against the rainbow team.

2.thumb.jpg.4822fb359391e4c8f502b8f2d0177485.jpg
Got pretty much the same results. Dominatrix and physical are in 1-2 places (first have a better average result, second have more chances to win). The rest go in the same order.

What do you think of the results?

  • Like 1
  • Thanks 4
  • Thinking 1
Link to comment
Share on other sites

Il y a 5 heures, Master-17 a dit :

What do you think of the results?

You need much larger samples, because rounding plays a HUGE role in the way battles work. 1 HP can make a difference in some cases. In some other cases, 30000 HP won't change the outcome.

There are also other factors: against high-defense opponents (Like Pantheon), Dominatrix is even stronger (and eccentric even worse). Again low-defense opponents (Like mono-equipped players), Crit Damage might be a bit better.

Sensual becomes very strong at high level, especially against weaker opponents, because you can heal most of the damage you take, which guarantees 24-25 points battles. Conversely, against much stronger opponents, where you take so much damage than healing doesn't even kick in, it's pretty worthless. In theory, yellow could become better than dark against opponents with a super high defense (But... Their defense would need to be higher than your attack, so you'd probably still get a 0% chance of winning 😛 )

That's why tier lists like this may only give a general idea. You need to check various elements on a case by case basis, trying different teams for each fight.

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

  • Moderator
9 hours ago, Master-17 said:

1 - Dominatrix

2 - Physical 

3 - Sensual

4 - Exhibitionist

5 - Submissive (more chances of victory but less average reward) 

6 - Eccentric

7 - Playful

8 - Voyeur

Definitely not. Dark and Blue are god-tier high above the others. Next tier includes Orange and possibly Green, then there's a gap until Red. And then a massive gap between Red and Purple, then a smaller gap until White, and then a gaping gap until your reach the absolute garbage-tier that is Yellow.

  • Thanks 1
Link to comment
Share on other sites

il y a 8 minutes, DvDivXXX a dit :

then a smaller gap until White, and then a gaping gap until your reach the absolute garbage-tier that is Yellow.

Yellow and White are really equivalent, though. If 2 players have exactly the same teams, they cancel each other perfectly. And Yellow has more counter potential than White (Yellow is good against the strong Blue element, but White counters the weak Purple).

  • Like 2
Link to comment
Share on other sites

  • Moderator

I disagree. White is in the closed loop with Dark and Purple, it very often comes into play in PvP. Dark teams prompt opponents to make Purple counter teams, which in turn can be countered by a White team. And many weeks, Blessings alone lead to a plethora of Purple-tagged opponents regardless. And that loop's counter gives a +20% crit chance bonus, which is huge!

Yellow "counters" Blue, and that comes into play often enough, but it isn't quite as big of an impact, considering they're both in the big 5-color loop and it isn't always worth weakening your team by inserting Yellow girls just to get a mild bonus against a Blue team which is often strong or at least resilient due to the extra Lifesteal. And that loop's counter gives +10% "damage" (I really think they mean attack though, much like in the Dark element's description, but spreadsheet experts could confirm either way) and a chlorella. Which is a far cry from the +20% crit chance you can get from counters in the tight loop.

As for the effect itself, both are weak for sure, and maybe in a lab with hypothetical pure White vs pure Yellow teams they cancel each other out, but not in real combat situations. In terms of active bonuses, adding a White girl to your team does something, not something great, but something. Adding a Yellow girl hardly does anything.

Link to comment
Share on other sites

  • Moderator

On leagues Blue is the top element most of the time. I ALWAYS check my expected average score with my current Blue team because it tends to be better than the "better" team against most opponents (except the strongest of them). Of course, I can only talk about my experience and my stats. Dark is the absolute best against strong opponents, and even Orange is better than Blue against the strongest, but that hardly comes into play in the league (you try to battle against unboosted opponents, right?).

For the same reason, White tends to be better for me than Yellow too. If you see Yellow as a poor man's Dark, you can see White as a poor man's Blue. Of course, they're awful compared to the other elements except Purple.

An as other people have said, every battle can be a world and ultimately you need to check with the script to be sure. But if you don't want to use all your hours in this game it's good to have shortcuts that work 99% of the time.

  • Like 3
Link to comment
Share on other sites

Statistics don't lie

10 hours ago, Liliat said:

You need much larger samples

OK. 2 more samples. With full information. First with rainbow wquipment:

1.thumb.jpg.55d81cb8a70b27b3898e5e5b69cfe021.jpg

No change from previous measurements. I would even say that the physical ones are moving even closer to the first place. Less chance of losing and more chance of getting better results. Sensual ones approach the dominant ones, but still they are slightly inferior to them because of the greater chance of defeat. Exhibitionists are only slightly behind due to slightly lower average results. Submissive are even a little weaker. Eccentricplayful and voyeur can lose with a more devastating result, but the chances of a better result for each subsequent one are lower than the previous one.

Second with monoequipment:

2.thumb.jpg.fba14297d81e586c1a8a00124f150cdc.jpg

Physicals are in the lead again here, but dominants are less likely to lose big and more likely to get medium-high results. Unexpectedly, eccentric and playful come forward, thanks to good chances for medium-high results. But at the same time, they are much more likely to have a major defeat than sensitive and exhibitionists. In the outsiders are submissive and voyeur due to less chances of a serious victory.

All results based on replacing one character of equal strength but different elements.

  • Like 1
  • Thinking 1
Link to comment
Share on other sites

  • Moderator
24 minutes ago, Master-17 said:

Statistics don't lie

OK. 2 more samples. With full information. First with rainbow wquipment:

1.thumb.jpg.55d81cb8a70b27b3898e5e5b69cfe021.jpg

No change from previous measurements. I would even say that the physical ones are moving even closer to the first place. Less chance of losing and more chance of getting better results. Sensual ones approach the dominant ones, but still they are slightly inferior to them because of the greater chance of defeat. Exhibitionists are only slightly behind due to slightly lower average results. Submissive are even a little weaker. Eccentricplayful and voyeur can lose with a more devastating result, but the chances of a better result for each subsequent one are lower than the previous one.

Second with monoequipment:

2.thumb.jpg.fba14297d81e586c1a8a00124f150cdc.jpg

Physicals are in the lead again here, but dominants are less likely to lose big and more likely to get medium-high results. Unexpectedly, eccentric and playful come forward, thanks to good chances for medium-high results. But at the same time, they are much more likely to have a major defeat than sensitive and exhibitionists. In the outsiders are submissive and voyeur due to less chances of a serious victory.

All results based on replacing one character of equal strength but different elements.

Well, seeing your screenshots I can see what's happening. I never see those odds and averages unless I'm unboosted and my opponent is boosted. And of course I never fight that way. I have averages above 20 99% of the time (and that's without AME, but with 3 cordyceps).

In your situation Sensual is not that great and Dark and Physical shine. It can even make sense to use some mono equipment. It reminds me of my situation in CxH, there I don't use Sensual as much as in HH, just like I don't use it against Pantheon Walls.

By the way, have you made sure those are unboosted opponents?

  • Like 2
Link to comment
Share on other sites

  • Moderator
1 hour ago, Master-17 said:

Statistics don't lie

No, but humans need to decide on relevant criteria before making them, and then humans need to interpret the results based on these criteria.

Here, it's not a question of how many samples you make. Your opponent is ridiculously stronger than you, and they're Rainbow-tagged, so the differences you get by switching around your own color tag are literally meaningless. You could run a billion simulations on this and it would still serve no purpose.

Have you seen your results, dude? Comparing 68% odds to win with 65% odds to win, let alone a plethora of intermediate results in between is a sheer waste of time, and doesn't say anything about the usefulness of active bonuses in general. Not even for you specifically, actually. All you've proven is that you're going to have to skip this battle until either Phase 2, or until you reach a higher player/girl level and/or better team to have a shot against this type of opponents.

If you want to make a relevant test, pick an opponent that you have 100% odds of beating regardless, and then check how many points you're likely to score according to the script's simulations in various configurations. That would already be much more useful. But EVEN THEN, it would NOT be a test of the actual strength of any specific active bonus against another, the way you're doing it. You seem to be changing exactly one girl in a very weak and mishmash team and consider that a good premise to test which color's active bonus is the best as if in a vacuum. It doesn't work like that at all.

All the bonuses are added percentages on some stat you already have (*), and they interact with one another, with synergy in the best cases (Dark and Blue go ridiculously well hand in hand, not coincidentally) or with anti-synergy in the worst cases (Green and Blue can clash together for instance). If you have a strong attack to begin with, Dark will have a more significant impact than if you don't. And so on.

Honestly, it pains me to see you spend what looks like a fair bit of time and efforts into making these tests when they're poorly conceived and won't get you any useful insight. Back to the drawing board, M-17. For your own good! ❤️

Edit: (*) Actually, Yellow is the outlier for this too: instead of a bonus on a stat you already have, it gives a theoretical handicap on a stat your opponent already has. One of the stats that has the least impact and variance out of all of them (because Defense as a stat is just far too low compared to Attack to be as relevant as it was probably meant to be).

By the way, in 99% of situations in the current version of the game, full Rainbow equipment outperforms any set-up that includes 1 or more Mono item. Sad, but true. Only in very niche cases is it ever worth changing your gear for a specific opponent in PvP, these days (almost always because your opponent themselves made a mistake and equipped several Mono items, giving them a very low Harmony which allows you to reduce your own for a bit of extra punch without much risk of it backfiring). So you should really only test with full Rainbow. Or test full Mono, but against a Pantheon team, where it's actually the best choice most of the time because your opponent will crit you into oblivion no matter what, so Harmony is no longer a factor.

  • Like 3
Link to comment
Share on other sites

  • Moderator

This is cool.  I can tell you put a lot of effort into it.  But like Liliat said, you need larger samples.  That doesn't mean more data for one opponent.  It means thousands of opponents.

Realistically, I wouldn't bother changing one girl to see if it'd help vs. a single player.  I do that for the Pantheon, but not for leagues.  Also, if I were to switch girls out, it'd be a totally different team, of at least 3 girls of a counter element.  Not just 1 girl.  I get that you're trying to see which element is better, but I recommend 3 girls, not 1 girl.  I also get that you were trying to limit the girl power variations.  It might just be too difficult to test, unless every girl is unblessed & lvl 700 or 750.

You should test this w/ 3 girls of each element, for 8 different teams.  Maybe a 9th rainbow team too.  Versus many different opponents, not just one.  The reason is that, in this particular case, a certain element might help against a specific opponent, b/c your ego gets a slight boost, or your atk/crit chance helps you win earlier in the ~100 simulation fights.  You can check how many fights the sim did in the console, (right click -> inspect element -> console).

To understand the grand scheme of things, looking at elements objectively, you'd need much more data, vs. a wider range of opponents.  Different total girl powers compared to your own (which affects atk), different boosters, different element counters, easy un-boosted opponents, opponents boosted w/ 3 cords, 3 chlorella, etc.  It's honestly too much for one person to do.  That's why the sim is handy.

Also, I'm pretty sure that 163.7k team opponent, is boosted w/ at least 2 cords, for that 118k atk.  That's why Kenrae kept hinting that.  That's fine, but you have to take into account, it would lessen the benefit of using exhibitionist or submissive girls even more, to fight an opponent w/ boosted atk.  Boosting HP or Def matters less to an opponent w/ boosted atk.  If you fought that same player, while they weren't boosted, the sim would spit out different results.

There's a script for checking if they're boosted: https://github.com/45026831/hh-booster-detector/blob/main/hh-booster-detector.js
10%, 20%, or 30% to atk/ego/harmony, is 1-3 cordyceps/chlorellas/jujubees.
13-17% to atk, plus a small boost to the def/ego (3-12%), is a mix of boosters, and likely 1-3 ginseng.
You want to see no stats boosted.

  • Like 1
  • Thanks 2
Link to comment
Share on other sites

From the same thread, and just below the chart shown in the OP is the following :
Best Elements 2.png

Which makes it clear this is based on a top players experience, and that lower level players will probably not get the same benefits, and different elements may have a better effects for them.

This is may current league sheet to show how many different teams I am using this time :

League example.png

Edited by Incubys
  • Like 2
Link to comment
Share on other sites

On 9/13/2022 at 6:25 AM, Ravi-Sama said:

There's a script for checking if they're boosted: https://github.com/45026831/hh-booster-detector/blob/main/hh-booster-detector.js
10%, 20%, or 30% to atk/ego/harmony, is 1-3 cordyceps/chlorellas/jujubees.
13-17% to atk, plus a small boost to the def/ego (3-12%), is a mix of boosters, and likely 1-3 ginseng.
You want to see no stats boosted.

Thanks for the link, looks good (when I get used to what it all means !). Also I hate you, as I'm going to be spending even more time in this game assessing who to attack in Leagues ! lol (But mainly thanks as it will be a great help)

  • Haha 3
Link to comment
Share on other sites

  • Moderator
On 9/14/2022 at 12:23 PM, Incubys said:

Thanks for the link, looks good (when I get used to what it all means !). Also I hate you, as I'm going to be spending even more time in this game assessing who to attack in Leagues ! lol (But mainly thanks as it will be a great help)

Actually, this script cut the time I used in leagues by a lot. It does for me something I already did myself, but using a lot more time.

  • Like 4
Link to comment
Share on other sites

  • 8 months later...

I am wondering why playful is always rated much worse than submissive. I'm a new player, 26 days old and level 254, still lacking one mythic equipment slot and full 5 star leg girls team(s), so I cannot serve much relevant data at this point. The performance handbook only compares playful to dominatrix, which is ridiculous as there is no question that dominatrix is always much superior, as attack is always much higher than defence, and all relevant boosters push attack. But comparing it against submissive (where the handbook just says it's "okay"), there are two aspects:

1. One aims and mostly fights weaker enemies, which usually have less defence, which makes playful weaker. On the other hand, boosters push attack and it is probably common to have high win chance and points against enemies with similar defence but much less attack due to missing boosters. In leagues, I'm currently on the 2nd place in Sexpert I (just because I'm active, while actually at or below average power in this league) where the level range is large, and there is hence a large number of enemies which I must loose against, due to much higher level, better girls etc. So while moving up the leagues and in mid level/power range, this benefit of submissive over playful becomes less. Loosing still gives points, so it is still relevant how damage we deal.

2. One has the first shot, which benefits playful, since the enemies defence applies one more time than the own defence. This benefits playful when one wins (mostly) while it does not benefit submissive against strong enemies where one must loose (like point 1 above). The benefit is larger on shorter fights and I'm not sure how long fights are in average on high level scenario, to rate this against the submissive benefit when own defence is higher than enemy defence.

So it is own defence vs enemy defence and length of fights which play a major role, and currently, at my level stage, I would actually rate the benefit of playful, applying often for one more round/shot, higher, than the benefit of submissive, only applying if the own defence is higher. To get any meaningful data, I need a 5 star playful leg girl to compare against Dorothy. I hope to get Murane soon, there is not much crap rewards left :).

Any thoughts, am I missing something?

  • Like 1
Link to comment
Share on other sites

il y a 1 minute, Horsting a dit :

I hope to get Murane soon, there is not much crap rewards left :).

Careful: Murane is a L5 Playful, but she's weaker than modern L5 girls. She only has 23.75 base power, vs 25.00 for all the recent girls (Less than 2 years old)

il y a 4 minutes, Horsting a dit :

Any thoughts, am I missing something?

I agree that Sly's ranking is not ideal, specifically for Playful girls. The reason he rated them so low is because of the comparison with Dominatrix, but as you say, that's not fair. There are situations where the choice is not between Grey or Yellow, but between White and Yellow (Specifically because blessings may favor girls of these 2 elements).

While this doesn't massively change the outcome (Playful and Submissive are still part of the worst elements), it's still worth taking into account. As you said, the difference between Yellow and White is mostly situational. Sometimes defense is better (because your opponent is much weaker than you), sometimes attack is better (because your opponent is more evenly matched). Either way, the difference will be small on average. On very specific situations, it can make a massive difference (because of how rounds happen, sometimes getting a mere +1 attack or +1 defense means you beat your opponent 1 round earlier, or survive 1 extra round, resulting is a massive win rate and/or score bonus), but generally, I'd say these 2 elements are comparable. And they're both weak.

I recently added element-based sort in the Harem++ script, and there I also ranked Playful above Submissive and Voyeur (But it was pretty much an arbitrary decision for the bottom 3 elements. I think it's more important to decide what's best on a case-by-case basis, rather than find what's best on average for these low-tier elements. Because "on average", you want to ignore them completely anyway).

  • Thanks 1
Link to comment
Share on other sites

2 minutes ago, Liliat said:

Careful: Murane is a L5 Playful, but she's weaker than modern L5 girls. She only has 23.75 base power, vs 25.00 for all the recent girls (Less than 2 years old)

Oh, very good hint, thanks for this! I already recognised and wondered why some girls of same rarity and stars are weaker than others, also e.g. the 3 starter girls vs other common ones. Why were those never aligned?

6 minutes ago, Liliat said:

I think it's more important to decide what's best on a case-by-case basis

Agreed, when I have the time, I anyway swap out those close girls on every fight, compare the HH++ forecasts and use the strongest combination, which is sometimes quite surprising because of the +1 round +1 crit etc chances you mentioned, which can make a large difference on an individual outcome. But due to lack of strong yellow girl, it is more about white vs green (current blessings, no 5 star green leg so far) vs violet (yeah, Chichi at 550, Dorothy still at 500, need 20 white gems for her to shift the decision). And of course, counter bonuses and blessings throw everything around anyway. But since the performance handbook does this (absolute) ranking, and not everyone uses HH++ or finds the time/mood to compare on every fight individually, that ranking should be as correct as possible. And there I would tend to yellow > white > violet.

16 minutes ago, Liliat said:

Sometimes defense is better (because your opponent is much weaker than you), sometimes attack is better (because your opponent is more evenly matched).

That is my point 1, but I would actually rate point 2, that the enemies defence applies often one more time (hence yellow benefits one more time), much higher. How many rounds do high level/league fights last in average? I mean if one wins the fight in round 6, assuming equal defence of both opponents but one wins easily because of full attack boosters, yellow is in average 20% better than white. If one looses, both are the same (in average). Since, again, boosters are usually about attack and also counter bonuses do not affect defence, I'm not sure whether one can reasonably assume that the own defence in average is 20% higher than the opponent ones. Most correct for all players in average should be actually equal defence, which eliminates any benefit of white over yellow completely (only in average, of course).

Link to comment
Share on other sites

Il y a 1 heure, Horsting a dit :

How many rounds do high level/league fights last in average?

I'd say between 4 and 6 (without crits). 4 rounds against super-weak opponents (2-shot with crit), 6 or 7 rounds against unboosted strong opponents (Assuming you go 3x Cordy + 1AM). It can probably go to 7~10 rounds (still without crits) when using less boosters (As most players do regularly. Only a few players can maintain 3x Cordy + 1AM at all times)

Link to comment
Share on other sites

Sounds quite similar to how fights are in my level range and league. Indeed hard to maintain all boosters at all time when not spending kobans for cordyceps and AMs. A shame that PoPs give chlorella only  😢 , or does this change with higher level? I managed to win cumback contests 1-2 times a week, but its getting harder the higher the level with people starting to use strategically saved rewards etc. The resulting AM lasts max 2 days, even when trying to avoid spending them on seasons as much as possible. But well, kiss storage for kisses and especially tokens is limited, so without loosing kisses from regeneration it cannot be fully avoided.

If we assume that we challenge all opponents 3 times, hence also fight strong(er) enemies at some point, and we won't be able to catch every one of them without boosters, then we assume 7 rounds in average, this makes yellow 16.6% more effective on wins. If we now assume 85% wins (not sure what is realistic when doing all challenges), it remains ~14% more effective. And if we assume as well that the average defence of all opponents match ours, white has no benefit from that end. Of course it remains true that on individual fights things can be very different.

It must be taken into account that yellow generally only benefits us, same as black, brown and red, if we manage to defeat the opponent one round earlier because of it (or if we loose). So on most individual fights, white will leave us with a little more ego, and hence a chance for a point more, while yellow leaves us with much more ego in some cases, but with above assumptions larger average benefit.

But all theoretical thoughts. I hope to be able to do some statistics once I got Murane and managed to max her out. With girls equipment I should be able to compensate her lower base stats for a fair comparison.

Link to comment
Share on other sites

  • Moderator

8 months later... ^^

(That's fine, though, this is a continuation of the same discussion and it references the old one. But in general please be careful with necro).

-----------

This is an old self-quote but still relevant and it addresses both the necromancer's question and Liliat's personal rankings of the bottom three elements, with which I strongly disagree:

On 9/8/2022 at 1:14 PM, DvDivXXX said:

White is in the closed loop with Dark and Purple, it very often comes into play in PvP. Dark teams prompt opponents to make Purple counter teams, which in turn can be countered by a White team. And many weeks, Blessings alone lead to a plethora of Purple-tagged opponents regardless. And that loop's counter gives a +20% crit chance bonus, which is huge!

Yellow "counters" Blue, and that comes into play often enough, but it isn't quite as big of an impact, considering they're both in the big 5-color loop and it isn't always worth weakening your team by inserting Yellow girls just to get a mild bonus against a Blue team which is often strong or at least resilient due to the extra Lifesteal. And that loop's counter gives +10% "damage" (I really think they mean attack though, much like in the Dark element's description, but spreadsheet experts could confirm either way) and a chlorella. Which is a far cry from the +20% crit chance you can get from counters in the tight loop.

As for the effect itself, both are weak for sure, and maybe in a lab with hypothetical pure White vs pure Yellow teams they cancel each other out, but not in real combat situations. In terms of active bonuses, adding a White girl to your team does something, not something great, but something. Adding a Yellow girl hardly does anything.

re:

5 hours ago, Horsting said:

I am wondering why playful is always rated much worse than submissive.

+

5 hours ago, Liliat said:

I recently added element-based sort in the Harem++ script, and there I also ranked Playful above Submissive and Voyeur (But it was pretty much an arbitrary decision for the bottom 3 elements. I think it's more important to decide what's best on a case-by-case basis, rather than find what's best on average for these low-tier elements. Because "on average", you want to ignore them completely anyway).

White and Purple are still both much more relevant than Yellow regardless of their respective effects (which are all very weak anyway), if only for the fact that they're in the small RPS loop with Dark, while Yellow is in the larger 5-color group with the weaker bonus (and the smaller odds of coming into play due to Blessings). I still think both slightly improving your own defense and slightly improving the margin of crit chance between you and your opponent are better effects than slightly reducing your opponent's defense, so Yellow is bottom of the barrel regardless, but the relevance of counter tags is what really seals the deal imho.

I hadn't checked Sly's old tier list or thought about it in a long long time, but I'm pretty sure he made it before Ken helped clarify how much we initially misunderstood Red's effect and how weak it really is, so the part where Sly ranked Red equal to Orange in the second tier even Sly himself would probably disagree with nowadays if he still played. Apart from that, I still mostly agree with the rest.

Well, I stand by this, actually:

On 9/8/2022 at 12:40 PM, DvDivXXX said:

Dark and Blue are god-tier high above the others. Next tier includes Orange and possibly Green, then there's a gap until Red. And then a massive gap between Red and Purple, then a smaller gap until White, and then a gaping gap until your reach the absolute garbage-tier that is Yellow.

Anyway. There are more and more variables added to fights as we go, but this is still relevant and good to think or rethink about every now and then.

  • Thanks 1
Link to comment
Share on other sites

@DvDivXXX

Many thanks for your input. You state that white is more relevant than yellow because of the counter loop. I totally agree, but I was actually referring only to the synergy bonus ranking, not about which colour comes more often into play as a counter team. However, that is actually another important aspect indeed when the question is which girl to level up first.

Since the large (5 colours) counter loop does not only increase attack but also ego both by 10%, it is not that bad. I nearly never had a case where choosing my strongest rainbow team (earlier it was blue, but I lack a blue 5 star leg girl so far 😢) was better than using a counter colour team. So I use counter teams basically whenever the opponent has a colour, or even if it has two colours, and HH++ nearly always shows better chances, sometimes only a little better, but better. However, most opponents in my league and seasons are rainbow, second most have the colour which get blessings bonus (like large amount of yellow teams until this week) if any.  

3 hours ago, DvDivXXX said:

I still think both slightly improving your own defense and slightly improving the margin of crit chance between you and your opponent are better effects than slightly reducing your opponent's defense

And this is what I am disagreeing with according to my thoughts. In average, whether you make 100 more damage has the same effect as when you get 100 less damage, i.e. both the opposite of the opponent making 100 less manage. If both defence values are the same, than yellow and white have the same benefit in each round. But yellow (your damage/opponents defence) applies one round more often when you win, making it something around 15% better (depending on amount of rounds), as long as I am not missing/misunderstanding something.

Indeed red turns out to be weaker than ranked. But I'm actually wondering why: In theory, if both teams have equal harmony, crit chance is 15%. Orange bonus currently give me another 5%, hence 20% overall crit chance. +10% damage in 20% of hits means +2% damage in average, which equals the +2% damage of dark. Or does it actually mean 210% crit damage instead of 200%, so not +10% but just +5% overall damage on crits? Is this what Ken helped clarifying?

Edited by Horsting
Link to comment
Share on other sites

Il y a 5 heures, Horsting a dit :

And this is what I am disagreeing with according to my thoughts. In average, whether you make 100 more damage has the same effect as when you get 100 less damage, i.e. both the opposite of the opponent making 100 less manage. If both defence values are the same, than yellow and white have the same benefit in each round. But yellow (your damage/opponents defence) applies one round more often when you win, making it something around 15% better (depending on amount of rounds), as long as I am not missing/misunderstanding something.

Then it depends what you're looking for. If it's purely "Points", then Defense and Attack will behave differently. Attack works with "Breakpoints": if you can win in 1 less round due to extra attack, you gain a lot of extra points. Otherwise, you gain nothing. Defense and Ego also work with such breakpoints when it comes to the chances of winning a fight, but also provides extra points if you reach a "score breakpoint". I don't know the threshold to gain 20/21/22/... points in leagues, but if your defense allows you to reach it, you'll gain more point (even with the same %win rate).

This means that while yellow might be better overall, the chances for it to actually change anything are lower (but when it does, it should be a significant buff). Whereas white might have higher chances of giving you a smaller buff. Probably. (Note: at high league level, the win rate is almost 100%; so only score matters. Losses happen from time to time, but that's usually less than 1% chance)

Critical hits (Orange, Purple) will give you extra points/extra win rate no matter what, where every single point will give you a small bonus to your expected result; but will never trigger any breakpoint. So, lots of small bonuses, but no big one.

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

  • Moderator
8 hours ago, Horsting said:

Is this what Ken helped clarifying?

I can't research the thread now (it's just a bit less ancient than Sly's guide ^^) and Ken quit the game as well since then. But basically, the way to read "+sth% crit damage" isn't the optimistic one but the more mundane one (I'm tempted to say these buffs are additive rather than multiplicative, but don't quote me on that ^^).

With 0 active or passive Red bonus, a critical hit does double damage compared to a regular hit, let's call that "200%".

With a maxed out passive Red bonus of "+35% crit hit dmg" (which is what you get permanently for having 100+ Red girls in your harem, regardless of your active team or tag) is added to the already doubled crit damage, not multiplied with the total damage you're doing that turn. So when you crit, you're doing 235% of your regular damage, up from 200%.

With a maxed out active Red bonus on top of that (from having 7/7 Red girls in your team), you're adding another +70 to that 235, so in case you happen to get a crit hit, you're now doing 305% of your regular damage, up from 235%. So the maximum boost it's possible to get from the Red bonus, both active and passive, is that when you happen to get a crit hit, you're doing triple damage instead of double damage, and a good chunk of that comes from Red's passive already. And even if it were quadruple crit damage as originally assumed, Red does literally nothing the rest of the time. Boosting your crit chance with Orange, your regular damage with Dark (which also counts for crit hits) and/or the lifesteal you get from all your damage with Blue all have a much bigger impact than Red does, and they all scale much better, on top of working every turn (except turn 1 and when you're already killing it anyway, for Blue).

You're getting a crit every now and then by default. At best, with a maxed-out passive Orange bonus, you have a flat 7% chance from that (hence, third-best element) plus the low-ish crit chance you normally get from your low base chance give or take the bonus or handicap you get from the difference between your Harmony (which Purple boosts) and your opponent's Harmony. It depends on where you're at in the game and what type of fight we're considering, but for late-game high-division PvP fights, that's typically not a lot. Fights don't last dozens of turns for this to "average" significantly or meaningfully: they last a few turns each. Even when you do get a crit, it's not a given that it'll make or break the fight then and there (except maybe in the early game or against an opponent that you were already beating handily regardless).

So yeah, Red is basically a win-more mechanic that's too low and inconsistent to really help us win more when we have other buffs that are unconditional and also scale with crits (or improve your crit chance). The passive is nice to have, but not that great compared to just the base crit damage mechanic on its own. And the active is just not worth it (it's not nothing if you happen to have other reasons to use Red girls in your team, but not worth putting Red girls in your team just for it). So it's not great. Better than the bottom three overall, but nowhere near the second tier (Orange and Green) let alone the top tier.

Afterthought: damn, I would have spent a lot less time searching for my old discussion with Ken on this topic and quoting it than I did remembering, re-checking and typing all the above! ^^

  • Thanks 1
Link to comment
Share on other sites

5 hours ago, Liliat said:

Attack works with "Breakpoints": if you can win in 1 less round due to extra attack, you gain a lot of extra points. Otherwise, you gain nothing. Defense and Ego also work with such breakpoints when it comes to the chances of winning a fight, but also provides extra points if you reach a "score breakpoint".

Explained in much better words then I did above, thanks for that. This is why I tried to use the phrasing "in average" as much as possible, as on most individual fights white will be slightly better than yellow, when the next breakpoint is not reached. All this ranking of course can only be about averages.

About the score breakpoints in leagues, as far as I could see it is linear at least, so if 10.000 lost ego means one lost point, 20.000 ego means two lost points. Ah, here is how HH++ calculates it: https://github.com/zoop0kemon/hh-plus-plus/blob/a157a1f/src/modules/BattleSimulatorModule/Simulator.js#L125

  • So on a win it's 15 + 10*leftEgo/maxEgo, rounded up. So every full 10% lost HP you loose one point.
  • On a loose it's 3 + 1 for every full 10% ego the opponent lost. As it is linear, this makes rating bonuses easier, since same absolute bonus means same (change to get) additional points in average.
     
1 hour ago, DvDivXXX said:

Afterthought: damn, I would have spent a lot less time searching for my old discussion with Ken on this topic and quoting it than I did remembering, re-checking and typing all the above! ^^

Hehe, but it was worth it. So indeed dark +2% damage means actual +2% damage increase while red +10% damage means only around 1% damage increase, depending on crit change, of course. Still better than yellow in most cases, but generally I do not like the randomness and hence increased range/uncertainty of outcome when it comes to crits.

 

2 hours ago, DvDivXXX said:

plus the low-ish crit chance you normally get from your low base chance give or take the bonus or handicap you get from the difference between your Harmony

15% in case of equal harmony, so not that low. With assumed +5% overall orange bonus this is how I get to +1% overall average damage increase for the first +10% red bonus. 10% of 200% times 20% = 0,05 * 20% = 1%. Very rough of course, but it gives an idea which moreless matches my observations so far.

  • Like 2
Link to comment
Share on other sites

il y a 2 minutes, Horsting a dit :

So indeed dark +2% damage means actual +2% damage increase

Even better: 2% attack means 2.x% damage increase, because the opponent's defense is subtracted last.

If you attack with 140k vs 40k, you do 100k damage. With +2% attack, you get 142 800 vs 40 000, which is 102800 damage (2.8% better).

The higher the opponent's defense, the higher the bonus for attack/damage.

  • Like 2
Link to comment
Share on other sites

2 minutes ago, Liliat said:

Even better: 2% attack means 2.x% damage increase, because the opponent's defense is subtracted last.

You are right of course. I was ignoring this effect for simpler comparison since it this applies with the exact same absolute value to all types of bonuses which increase your damage, also to crits and increased crit damage, without affecting the raking: Also +10% crit damage in above assumptions means more than +1% damage when taking flat reduction through defence into account, but the absolute average bonus damage (and hence difference between colours) remains the same.

3 minutes ago, Liliat said:

The higher the opponent's defense, the higher the bonus for attack/damage.

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

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share

×
×
  • Create New...