It is just coincidence, but on-color feeding is always more efficient, so you are very wise to stick with it
Ive always fed whatever hero im leveling everything (except 2* and 3* trainers) and have always maxed special before 3rd ascension. I know its not as efficient but im incredibly impatient.
On color until the special is maxed, then, any color to speed up the leveling.
Once we’re at tier 3 or tier 4, the 20% hp loss in hp for off color feeding is negligable.
Except for the trainer heroes where I always do on color feeding as the hp loss is too much to ignore.
I plan to max every *4 before 3.60. So start by 10x at a time *1 feeder on-color, and the result, Wilbur at 6/8, Scarlett at 5/8.
So I do not want stuck, then, I try…
Wu Kong 3.01 at 5/8, I start feed 1x at a time *1 on-color… at 3.30 finish skill 8/8, then start with 10x at a time on color, because more eficiene food cost.
Boril also like Wukong, now it is 8/8 at 3.34.
Proteus 7/8 at 3.10, I hope get 8/8 before 3.60 (but this one, I will ascend to tier 4), no problemo.
I had a Rudolph I was trying to power level for the last 3* raid tourney get to 3/50 with 4/8 specials and it took an ungodly number of attempts at 60-75 percent to get those 4 specials - something like 18-22 attempts. Literally hundreds of feeders.
It sucked, but it’s just random.
Since last version 1.80 (maybe), after hero level maxed, special skill is more easy, just use 5x *2 on-color, and it is 100% guaranteed skillup.
What’s also irking that if your specials are not full and you’ve actually hit the current ascension ceiling and couldn’t go on, you are stuck having to leave that hero alone until you ascend him/her.
3* Heroes say are at 2^40 and couldn’t ascend because short of mats (I know, only farmables, but you never know when they player, who might be newer, would be short of a chainmail shirt or tall boots) and if his/her specials are not full at that point, the level up button is greyed out). Perhaps a more common situation might be for 4* Heroes at 3^60, but specials somehow are not full, and they are short of non-farmable mat, be it glove/compass or their elemental mats), they can’t work on their specials any more until they are ascended past that point. My Wilbur is approaching that point and he is only 6/8, and I don’t have the Gloves to ascend him in my inventory, not until I get them from the next Farholme.
That is similiar to my plan and I do plan to have 8/8 special skill at 3/60 especially for the heroes I’m not sure to bring to last ascension
6/8 and 5/8 at 3/60 with 1* same-color feeding? That’s very bad luck…
My worst record is Ameonna 7/8 at 3/60. I used mix of 1* and 2*, sometimes 3* (a level before ascension [1/39 or 2/49 or 3/59] I usually use mix of 2* and 3* to max the probability )
Using only 1* do maximize the mean of number of special up occurences, however it also increase the variance due to lower probability (more frequent good luck but also more frequent bad luck).
10x or 1x or various around 1 - 5 at a time or how much heroes at a time?, thats what I’m concerning about now.
That used to be the subject that’s been heavily debated. When you lack the material and resources to level / ascend a hero, you kind of want to give your hero as much time to work on his/her special before he/she hits the current ascension ceiling.
For me, I still prefer to feed 1* of the color 1 at a time, sure, it is 2% each time, but if I have 10 of 1* feeders of the same color, instead of only one chance of 20% (that can get it to go up one at the most), I would rather have 10 chances of 2% each, who knows, although unlikely, I might even get more than 1 special level up.
This has become less of an issue since a few versions ago where a maxed hero gets 5X chance. For 4* heroes, like I said in my above post, I am looking like I am about to run into my current ascension ceiling for Wilbur, whose special is currently at 6/8. If he hits that 3^60 and the special did not go up, I won’t be able to work on his special any more until I get another pair of Gloves to ascend him. Although it doesn’t seem like 4* would have problem with special level ups very often, it does occasionally happen.
3* heroes are getting the most benefit with that upgrade, because all of their ascension material are farmable.
So far all my 4* s achieved skill 8/8 before 3/60 and all 5* s before 2/60. My leveling is consciously steered towards achieving that.
Starting everyone with 10x 1* on color, 20% chance, continue on and on, and if I notice any draw back, I start playing around with the percentage of the chance, sometimes 22% -30% by mixing 1* & 2*, or even lower to 13 % -18 ( mixing more 1* on color and a few 1* off color) never with trainer heroes before skill 8/8
At the end all maxed before the final tiers. Once skill 8/8 is achieved, I sometimes kitchen sink feed, but to be efficient, I always feed with 2s and or 3 on color at higher tiers.
Thats what I’m concerning about… I always 10x feed at a time on-color to *4, and mostly not getting 8/8 at 3.60.
While at the same time I also developing *3 which is I do not care about skill, because easy to skillup after maxed (and without farmable mats). So I only feed *2 on-color to my *3 rosters.
So if I get *1, I keep until 10x to leveling my *4 heroes (on-color), and if I get *2 some times keep around 1-3 and feed to my *3 heroes (on-color). The results for *3 majority get skill 8/8 before reaching 3.50 but some has 7/8.
Thats why, I rethink for leveling other *4 heroes. Which is I will prefer using around 1-5 at a time *1 on-color.
First time it happened, so felt it necessary to post here.
My leveling habits are always the same and the special skill is always maxed before 3.60, at least for my 4*
What is sad, though, is that Caedmon will remain at 3.60 for quite some time, as i’m short on gloves
It is always better to feed it 10 at a time. There is a mathematical approach to that.
Suppose you have 50 1* of the same color.
Feeding 5 at a times:
Probability to have at least 1 level up:
1 - 0.9^10 = 0.651 = 65.1%
Feeding 10 at a times:
Probability to have at least 1 level up:
1 - 0.8^5 = 0.672 = 67.2%
The smaller the probability, the higher the variance, the smaller the guarantee.
Using my method I succedd in bring 16 out of 17 to 8/8 at 3/60.
Here is the secret:
When your hero is at first ascension, make sure it land at level 39, just before it maxed. Then accumalate 10 2* of the same color, level it to 40 with 40% chance.
At 2nd ascension, stop at level 49 and do that again.
At 3rd ascension, stop at level 59.
Basically, it allow you to have 3 more chance to increase the skill (without increasing their level more than 1) with 40% probability each time.
About Mathematical approach is on debate.
This graph, mention that 10x at a time and 1 at a time is same.
Now, thats make sense with this exception, thought thats way around 1-5 heroes at a time is still more or less or have same chance, because it is still RNG. And I still leveling *3, so I do not want use *2 to my *4 heroes, until maybe at tier 4, because it mostly more food costs, so then my development at that time (if all *4 are at 3.60) then I stop leveling *3 (almost done). So next *2 feed not for *3 heores instead for *4 heroes tier 4. And continue *1 heroes for new *4 heroes Tier 1.
So conclution for me, for stoburn hero that di not skillup doing 10x at a time, I feed around 1-5 heroes at a time until 8/8.
I saw that once, that is a good math. But look carefully:
One 1* at a time
Ten 1* at a time
It prove my conclusion is correct. Ten at a time is better than one at a time. One at a time due to its lower probabilty, have higher variance and less guarantee (it has higher chance to end at 7/8 and lower chance to end at 8/8 compared to 10 at a time).
OK, but I don’t want repeat my result for other heroes like my scarlett and wilbur. Maybe just my account RNG is not suitable with those methode I think thanks for the suggestion thought