champions Champions: Girls not trigger class effect the whole first round

[Horsting]

Probably it is just bad luck combined with selective perception, but does someone else have the impression that on first rounds of champion (including club-champion) performances, it is too common that none of the girls trigger their class effect, even that the chance is stated to be above 40%?

It also usually happens only on the very first performance against a champion/stage, and normalises on subsequent performances. Just happened again on my first performance against the current club-champion after the 24h waiting time: The effect chance is stated to be 43% but none of the girls triggered it in the first round. The chance for this is 0.57^10 = 0.36%. The same happens regularly (not always!) on the other champions, with class effect chances of 48%, which should happen only in 0.52^10 = 0.14% of cases. Also this time, on the second round it immediately normalised, with first two girls triggering their effect and then every now and then as expected. And the second performance also went through just as expected.

I will start to do some statistics about this now, i.e. how often it happens and more importantly how often it does not happen.

Edited July 16, 2023 by Horsting

[DvDivXXX]

My formula for champion fights is a bit simpler: Perform + Skip + Ok = done.

The only part I give any thought to these days is the draft when I start a new run (mostly for the CC; my regular champs enjoy their retirement on the same old draft except when I end up pushing one of the weaker ones too far during PoA or another "do champion performances" task).

[Horsting]

25 minutes ago, DvDivXXX said:

Perform + Skip + Ok = done.

The only part I give any thought to these days is the draft when I start a new run

Exactly. Only for the first performance against a non-trivial (CC and the 3 highest regular ones), after creating the draft, I often watch at least the first round, to check whether the plan works out, i.e. CH defence and KH healing vs champion attack (and in case matching poses) to survive as long as possible. This is how I recognised the strange effect that often there is not a single defence bonus and healing triggered the whole first round. When this happens, I wait for the first girls of the second round and check the second performance to verify that effects are then triggered as normal.

32 minutes ago, DvDivXXX said:

my regular champs enjoy their retirement on the same old draft except when I end up pushing one of the weaker ones too far

I actually thought that at some point it may be best to intentionally let the champs forget the impression to forever do performances on the same stage and champ level, instead of raising the difficulty. Do I understand correctly that this is what you are doing? Currently, a champion performance is giving 1-2M Ymen, doing this only every 1-2 days, or once a week on the first two champs, to equalise the 10% daily impression decrease, is too much lost Ymen in my current situation. Also the 10 kisses on final stage are at least not nothing, helped with last SE and might save me in KC, when I bring all champs to stage 5 before KC starts. Also I get more champ tickets than I can possibly use only on CC. But when the champ levels further raise while 5* girls at some point are capped, they could become too difficult to defeat at a daily basis with regular ticket income and/or give too few Ymen per performance to be relevant 🤔.

[DvDivXXX]

52 minutes ago, Horsting said:

forever do performances on the same stage and champ level, instead of raising the difficulty. Do I understand correctly that this is what you are doing?

No, there's a cap and I'm pretty sure all my regular champs have reached it. What I'm doing is simply not using them unless I have a reason for it. Until you've unlocked all their resident girls, you can't really ignore them, but once you catch them all, you could fight them less often to keep them on the same difficulty level longer.

52 minutes ago, Horsting said:

check whether the plan works out, i.e. CH defence and KH healing vs champion attack (and in case matching poses) to survive as long as possible.

That's not a good strategy against champions. You're better off dealing as much damage as possible in 5 rounds than stalling them for 10+ rounds and doing less damage to them in the end. Always match poses if/once you can, especially the first 5 and I'd even say the first 3. You'll do much more damage per ticket with 3 moderately strong girls matching the first 3 poses than with a mishmash of 5-10 stronger girls that don't match poses. This becomes evident quicker against HC bosses, but it's true for any champion fight.

EDIT:  

1 hour ago, Horsting said:

The effect chance is stated to be 43%

I don't think you're reading this correctly. Bear in mind champions operate under the legacy KH/HC/CH battle system, for the most part. Blessings still apply to individual girls, but only in terms of the third of their "base sum stats" that matches their own class. Elements and the bonuses thereof don't apply at all. The numbers you see on top that I'm afraid you might be mistaking for "chance for special effect to trigger" are actually just your harmony and the champions' harmony. And there's no crit chance passive bonus (that's an elemental perk and Elements don't exist in champion fights).

If you see 43% on your side and 7% on the champ's side, it simply means you get the lion's share of the bonus "crit chance" (or "special ability trigger") you and the champion share, and which amounts to an extra 50% of your respective base chances. Even if you have 50% and the champ 0%, it doesn't mean your girls have a 50% chance to trigger their ability when they attack. It means you have 150% of your base odds for that to happen in the old system, and the base odds are something like 20 or even 10% IIRC. You're definitely never even close to getting special hits 40+% of the time ever.

Edited July 9, 2023 by DvDivXXX
updated

[Horsting]

40 minutes ago, DvDivXXX said:

No, there's a cap

Good to know 🙂.

40 minutes ago, DvDivXXX said:

What I'm doing is simply not using them unless I have a reason for it. Until you've unlocked all their resident girls, you can't really ignore them, but once you catch them all, you could fight them less often to keep them on the same difficulty level longer.

I have all girls, but as said, major reason, even more than the girls were (as even their 5* have very low stats), is that they are a massive Ymen income. Without it, PoA Ymen goals, and rolling EPx1 thousands of times to collect GXP and Aff to max out Yesette and Selena in a few days would be impossible at my level 😁.

4 minutes ago, DvDivXXX said:

That's not a good strategy against champions.

Of course I do check/consider overall damage and try to max matching poses. But for me, usually maxing out the duration via CH+KH girls also maxes the overall damage, at least against CH, KH and mixed class poses: The HK bonus is usually way weaker than the other two, CH usually best, but as it triggers as well for the next girl, an alternating combination of CH and KH is in nearly all (my) situations the best team. Also collecting strong girls from two classes when re-rolling the draft has a higher chance than collecting strong girls from a single class. Only on hard HK champs, Sanji Alban later/last stage(s) and CC Chad, especially if their girls have matching poses, their damage is sometimes so high, that CH (and KH anyway) bonus has a too low effect on the duration and maxing out damage on first spots with matching pose (HK) girls becomes the better strategy.

E.g. our current CC Chayotte (It is always the same for all clubs, right?) is level 205 and has not a single matching pose. Class effects trigger, as said, with 43% (after first round of first performance 😉). KH healing is slightly more than his damage, but CH reduces it to 0 for two girls. Mixing in some CH girls dramatically increases duration and multiplies the overall impression per challenge, as my ego up and down gets quite close to equilibrium. A streak of multiple girls not triggering their effects is what usually brings it to the end. As this highly depends on the champs damage, I guess with some matching poses and a much higher CC level, the strategy to just have the strongest matching pose KH girls on the first spots becomes better.

With "round" above and in title I btw mean a full round with all 10 girls doing their thing once. Could be misunderstood with 1 round = 1 girl attack + champ answer.

[DvDivXXX]

These aren't league fights, the only thing you score is damage done to the champ, and the champ cannot heal (except once every 24 hours, for 10% of his max HP, if you let it take that long; for regular champs only: CCs don't have this mechanic as their runs can extend over multiple days).

I assure you, surviving longer =/= doing more damage. You seem to underestimate just how big of a damage boost you get from matching poses.

Also, against champs specifically, some older L5s can still outperform newer ones, especially the ones you got from the regular champs, as their main stat (KH for Any, etc.) is typically much higher than the other two, and it's the only stat that's taken into account for their total power against champs, unlike everywhere else.

I have no idea where you get your 43% from, but it sounds wrong to me.

Edited July 10, 2023 by DvDivXXX
updated

[Horsting]

4 hours ago, DvDivXXX said:

These aren't league fights, the only thing you score is damage done to the champ, and the champ cannot heal (except once every 24 hours, for 10% of his max HP, if you let it take that long; for regular champs only: CCs don't have this mechanic as their runs can extend over multiple days).

I know, but I do not understand what you want to tell me with this. With healing I mean the KH girl effect, with duration I mean the amount of rounds/girls hitting the champ before the single performance ends. Of course I do not aim to let a champ survive longer than 24h. You know me: If what you read from me looks like I misunderstood a mechanic from a regular (not new event or so) game element, you misunderstood me most likely, maybe I didn't write clearly what I mean 😉.

4 hours ago, DvDivXXX said:

I assure you, surviving longer =/= doing more damage. You seem to underestimate just how big of a damage boost you get from matching poses.

Well, it is not hard to understand: Matching pose is exactly doubled damage for this one girl. If I do exchange e.g. 5 of 10 girls with matching poses (very unlikely to roll 10 matching poses anyway) with CH girls which do not match poses, but assuming they have same base attack, the overall damage is reduced to 75% (half of the girls do half damage). If the fight in turn lasts the double amount of rounds/girls, the overall damage is increased to 150%. I do the math and I did observe sufficient champ fights to get an idea at which point/against which champs which amount of mixed in CH girls (without matching poses) makes sense, if any.

4 hours ago, DvDivXXX said:

Also, against champs specifically, some older L5s can still outperform newer ones

I know, but what are you referring to?

4 hours ago, DvDivXXX said:

I have no idea where you get your 43% from, but it sounds wrong to me.

The big percentage that is shown beside your currently attacking girl class, while the performance is running. This percentage tells you (at least it should) how likely it is that your girls are triggering their class effect, i.e. defence bonus of CH, healing of KH and additional damage of HK. This is what I mean with "class effect" in the title and what the whole topic is about. Dammit, sorry for obviously writing so unclear. One of the problems is that I have no idea how these things are usually named. They are not mentioned in the performance handbook, have no title or what ingame and also the wiki is not mentioning them. Generally I was not able to find any topic which discusses champ girl class mechanics and strategies 🤔.

Edited July 10, 2023 by Horsting

[Master-17]

17 hours ago, DvDivXXX said:

My formula for champion fights is a bit simpler: Perform + Skip + Ok = done.

How much do you do with champions))) My formula is: Go to champion -> Press Perform -> Back -> Go to second champion -> ... (repeat N times) ... -> Go to main screen. And then if this is at the time of the contest for fights with champions)) If you only need one fight, the formula is: Go to the champion -> Press Perform -> Go to the main screen 

[DvDivXXX]

1 hour ago, Horsting said:

The big percentage that is shown beside your currently attacking girl class, while the performance is running. This percentage tells you (at least it should) how likely it is that your girls are triggering their class effect

You mean this, right?

If so, sorry to disappoint but it's your shared harmony bonus, adding up to +50% of your base odds, as I explained previously. Here, I have +43% to my normal chance to trigger a special attack (which is something like 10 or 20%), while Chayotte has a +7% bonus to his normal chance. It does NOT mean that my girls have a 43% chance to trigger their special attack, or that the champ's girl only has a 7% chance to trigger hers.

If you're referring to something else, please point it out (or show it).

As for the things I've mentioned that didn't seem clear to you why I mentioned them, I was trying to make sure you're clear on how the old battle system works, since this was long before your time that it was everywhere instead of just against champions, and many subtleties are easy to overlook. So for instance, I pointed out that the L5s you get from regular champions have very poor "base sum stats" in the new battle system used everywhere else these days, but they have great stats in the champion environment, because only the highest of the three stats is taken into account (as opposed to league for instance where it only matters how much the total of KH+CH+HC is). This also explains why sometimes you can see two modern L5s with exactly the same level and the same total power in the new system but one of them will be a bit or even a lot stronger than the other in a champion draft, simply because the part of her modern battle system stats that happens to be for instance CH is larger than that of the other CH L5. It makes no difference in a villain or a PvP fight, but it matters in the old battle system still used for champion fights. It might have been too much info at once, and/or you already made your research and spreadsheets and you didn't need these reminders. Better safe than sorry, just trying to help. ^^

[Horsting]

17 hours ago, DvDivXXX said:

If you see 43% on your side and 7% on the champ's side, it simply means you get the lion's share of the bonus "crit chance" (or "special ability trigger") you and the champion share, and which amounts to an extra 50% of your respective base chances. Even if you have 50% and the champ 0%, it doesn't mean your girls have a 50% chance to trigger their ability when they attack. It means you have 150% of your base odds for that to happen in the old system, and the base odds are something like 20 or even 10% IIRC. You're definitely never even close to getting special hits 40+% of the time ever.

Ah sorry, I saw your edit just now. That is very interesting. I did quite some counting when starting with champions, and the percentage seems to match quite well my statistics. On regular champions, it is indeed close to every 2nd girl (in average) casting their special ability. So if what you say is true, then the base chance times shown percentage adds up quite well to the shown percentage 🤔. Just for my understanding top clarify what you mean:

• There is a 10% or 20% base chance for the special ability
• Only the percentage shown is based on your harmony vs champ harmony, like crit chance in PvP but with 50% shared among both (instead of 30%).
• So the max special ability chance should then be either 15% or 30%? If so, there seems to be more to it, because the average chance on my champs is definitely higher 🤔.

Also if what you say is true, it means that the champ's girl should cast her ability not that much less than you, but what I observe on all champs it that she indeed nearly never casts it. 10% * 1.07 would be probably correct, but then my girls cast it definitely more than 10% * 1.43, and the champ's girl definitely less than 20% * 1.07. So either there is more buggy in my case than just the first round (what this topic is about), or the mechanic was updated.

10 minutes ago, DvDivXXX said:

So for instance, I pointed out that the L5s you get from regular champions have very poor "base sum stats" in the new battle system used everywhere else these days, but they have great stats in the champion environment, because only the highest of the three stats is taken into account (as opposed to league for instance where it only matters how much the total of KH+CH+HC is).

Jep I understood that. I only mentioned the bad base stats to say that after getting some new 5* girls, the champs girl's were not the main reason for me to fight the champs, but the Ymen became the major reason. Of course I was happy to get the girls as well, but there was no urgency for me to do that. The chance that a particular girl is rolled in a particular champion challenge is very low, so I do not consider it reasonable to get specific girls in the hope they will significantly help you with champs 😉.

Edited July 10, 2023 by Horsting

[DvDivXXX]

You got it right. I'm honestly not sure whether there are new bugs, and/or your sample is just too small to account for variance yet, and/or the mechanics have been changed since the last time I've checked (it wasn't announced but it might have been a stealth update).

One easy way to verify would be to go against a really weak champ and ensure you get all of the +50% and he gets +0%.

If his girl never ever EVER triggers her special attack, then you might be on to something and these numbers might be indicating something else than what they used to since forever and definitely back when I was still farming my regular champs for their girls and kept up to date with those mechanics (but this was a million year ago, before they even introduced CCs).

If his girl still triggers her special attack about as often as usual, while yours don't trigger theirs all the time (yours or mine, I could do this test for you since it should be much easier for me), then it still works at least somewhat in line with the old system. And that shared 50% is a bonus applied to the base chance to trigger special attacks, which is relatively low (the hard info is definitely somewhere in writing, maybe even as a table, but in a very old thread if you care to make a deep dive with advance searches, but I can't remember off the top of my head what it was or where it was established conclusively back then, sorry).

EDIT: Sorry @Horsting (pinging you so you notice the edit ^^) I've checked and I can't test this today, as I'm in pure attack mode with my boosters and I can't get 50%/0% against any champ as a result. On a quieter day this week, I'll try with some gins on, it should be fine (I'm not too far even with 4 cordys, so a couple of gins should do the trick).

Edited July 10, 2023 by DvDivXXX
edit

[Horsting]

Okay, I just started to note down the amount of rounds + own casts + champ casts + shown percentage for each challenge, and the number of cases where the whole first round ended without any cast on my end. Nasty that there is no way to count the ability casts than watching the whole challenge 😴. However, I'll try to do that for a while whenever I have the time to have more than just a rough feeling based on rare casual counting.

With less than 30% casts it starts to be not that unlikely that I was just affected by selective perception.

EDIT: @DvDivXXX  

1 hour ago, DvDivXXX said:

EDIT: Sorry @Horsting (pinging you so you notice the edit ^^) I've checked and I can't test this today, as I'm in pure attack mode with my boosters and I can't get 50%/0% against any champ as a result. On a quieter day this week, I'll try with some gins on, it should be fine (I'm not too far even with 4 cordys, so a couple of gins should do the trick).

Same here: I have 4 cordys currently, so only 42/43% on champs where I can collect data: Romero and Whaty are done too fast. 50% (I guess 49% is max, same as 29% harmony crit chance in PvP?) would be better, but everything significantly above 30% should give a picture when collecting enough data.

What I can already say now after ~200 attacks across all champs, the champs ability cast rate is about 4%, while the shown percentage is 7% and 8%. So it is even lower than the shown percentage while with a base chance of 10/20% it should be slightly above 10/20%.

Counting the CH defence bonus correctly is a little challenging, since one cast is triggered on two girls. So one must pay attention to not count it two times, as well as not miss when the 2nd girl triggers her own ability 🧐.

Edited July 10, 2023 by Horsting

[DvDivXXX]

Okay, sounds good (although still too small a sample to go beyond variance).

4 hours ago, Horsting said:

(I guess 49% is max, same as 29% harmony crit chance in PvP?)

I'm 99% sure I already had 50% vs 0% a bunch of times. We'll figure it out soon enough. The thing is versus a human player's team, the difference in harmony will almost never be spectacular one way or the other (at least in D2-D3). Versus a champ, even without really trying, it's almost hard not to have a huge harmony advantage on him; so when you actually boost your harmony, it goes through the roof quickly.

[Horsting]

Did some counting and math. It was mostly with 4x cordy, so lower cast rates, and I did not have the time and/or too lazy to watch and count all fights, but it is sufficient to derive some info:

	Player	Champion			Player	Champion			Player	Champion			Player	Champion
Shown %	42	8		Shown %	43	7		Shown %	44	6		Shown %	46	4
Hits	81	80		Hits	956	955		Hits	509	505		Hits	298	294
Casts	36	1		Casts	401	37		Casts	237	12		Casts	142	5
Actual %	44,44 %	1,25 %		Actual %	41,95 %	3,87 %		Actual %	46,56 %	2,38 %		Actual %	47,65 %	1,70 %

I noted data per champion and per performance, but accumulated them here per shown (cast rate related) percentage. I think we do not assume that it works differently on different champions, and the individual rates were not too different either.

As can be seen, my cast rate is always pretty close to the shown percentage. What I already mentioned and got hardened, is that the champions' cast rate is always quite significantly below their shown percentage. Either I fall asleep a little too much while watching the by times long performances, to miss the champs cast by times, or there is more about it.

Let's pick the 43:7 results, mostly from Chayotte CC and from one Visor last stage iteration, as this is the largest data sample and the only one where my actual cast rate is below the shown percentage. Also it is the sample with the highest actual champion cast rate, and the one closest to his shown percentage, being important later:

• If the system was a base 20% cast rate increased by the shown percentage, the average cast rate should be 0.2*1.43=28.6%.
• Such discrete yes/no random experiments are binomial distributed, hence the chance to get a certain actual rate or amount of casts (individual and cumulated) based on the underlying average rate and sample number, can be easily calculated.
• In this case, with 956 hits, when assuming the 28.6% cast rate, the chance to get up to including 350 casts (36.6%) is 99.99999645%. My calculator just shows 1 when I try to calculate the chance for up to 401 casts (the actual amount I got). So the chance to get more than 350 casts or even more than 400 casts is basically zero. This proves the theory of the base 20% rate wrong, and no need to check base 10% rate, of course.
• If we assume that 43% actually means 43%, then my amount of casts 401 is 14 casts below the expected average of 956*0.43=415 casts. The chance to get an amount of casts within this range of 415+-14 resp. 401 - 429 is 88.5% (cumulated until 429 casts) minus 24.5% (cumulated until 400 casts) = 64%, hence within the standard deviation range which contains a cumulated chance of about 68%. So this theory is perfectly possible based on the data.
• What bugs me is the champion's rate. The 10% or 20% base rate is even more ruled out by the data, but the 7% means 7% is also unlikely. When assuming it, the chance to get up to including the 37 casts from 955 hits is only 0.0029%. For the other samples, while the sample size is smaller (which increases this chance), the champ's actual cast rate is even further away from the shown percentage.

I cannot exclude human error while counting, since it is a pretty boring job watching these by times long performances. But I tried to be concentrated, took care to not count the x2 defence bonus doubled or miss the next girl's own cast (who can do two casts then: defence from last girl and her own), and also catch the champion's casts. Also I didn't note any case where I started dreaming and was not sure whether I missed a cast or not, or something. While I am pretty certain that my own casts were counted pretty accurate, with an error hange of one hand at most, I am not so certain for the champion, as my eyes usually kept starring at my own girl. This may be the explanation for the systematically much too low champion cast rates.

Best would be to write a script do automate the counting. But if it was only about the champion, it does not really matter whether his rate is 2% or 8%. It is too low in both cases to make an important difference to stategies.