I tried Embossed in the rift today, 12 breeds, and despite the cloning odds being 30% I did not see anything close to those numbers. All attempts failed, which leads me to believe that the 6% adjusted chance is most likely correct. More numbers are needed as I could have had a terrible luck streak, but another day or two will really prove what's happening.
So it looks like something special happens with the Epic cave, and the chances remain at 30%. The other caves look to be 6%.
The social cave is much harder to test, numbers are always much lower in that cave. Trying to determine if a dragon is getting it's base 2% clone, or adjust 1.5% is really difficult without 1000s and 1000s of results. I've been looking for a dragon with 20-30% 2 parent cloning chances that has enough elements to really adjust the numbers. So far the best I can find is one that takes it from 20 to 15%, and that's still so close that we'd still need many results. I would take a guess that social follows the same rules as the rift and normal cave, but it would be nice to prove it.
Cave: Between 15% and 21% Embossed depending on what the latest 36 hr breeds turn out to be. (5 to 7 out of 33). This is across all parks with various breeding partners (Rainbow, Triple Leap Year, Leap Year) at level 10 or 15. Some caves are upgraded, some not. (I'm keeping track of all this when I'm logging my results; specifics available upon request).
Island: 34-38% Embossed, or 11-12 out of 32. As above, different breeding partners but all level 15; some islands upgraded, some not.
I have Embossed in five alt parks now, but I only have time to check them once per day. It's going to take a little while.
So I did the testing on Embossed + Rainbow in one of my alt parks. Here are the results with 12 breeds: Dodo Embossed 3 Sandstorm Plasma Arctic Conifer Molten Tempest Lagoon Frostbite
I will continue with this park (the only one that has a rifted Embossed) until the next event. Although I may take a few breeds for a rifted Fury in between. 😉 Edit - just want to add that each parent dragon is at level 16.
So I did the testing on Embossed + Rainbow in one of my alt parks. Here are the results with 12 breeds: Dodo Embossed 3 Sandstorm Plasma Arctic Conifer Molten Tempest Lagoon Frostbite
I will continue with this park (the only one that has a rifted Embossed) until the next event. Although I may take a few breeds for a rifted Fury in between. 😉 Edit - just want to add that each dragon is at level 16.
So I did the testing on Embossed + Rainbow in one of my alt parks. Here are the results with 12 breeds: Dodo Embossed 3 Sandstorm Plasma Arctic Conifer Molten Tempest Lagoon Frostbite
I will continue with this park (the only one that has a rifted Embossed) until the next event. Although I may take a few breeds for a rifted Fury in between. 😉 Edit - just want to add that each dragon is at level 16.
(...) But if the question actually boils down to "Does the EBI produce regular hybrids when Set%' > 100?" then I suppose any breeding combination with Set % > 100 will do? I'm thinking I'll try Rainbow+Poison in all breeding locations (Cave, Island, Rift, Social) in my parks to see if there's a noticeable difference with the % of regular hybrids.
Update.
I did the test with Poison(4)+Rainbow(15) to see if I would got a different % of regular hybrids (excl. Poison) in different breeding locations (cave, island, rift, social). The result was negative. In the rift only 2 of 219 breeds were reg. hybrids, and none were Poison. Outside the rift, 0 of 132 breeds were reg. hybrids. In the social cave I got ca. 14% Poisons, and that was the only place I got any Poisons. (Not a surprise. Poison has a 10% parent cloning chance in the social cave compared to 1% in the rift and cave/island). The Poisons in the social cave may have come at the expense of epics (in the rift I got 20% epics, in the social cave only ca. 10%) but I don't want to jump to conclusions here. My focus was the % of reg. hybrids, and I ended the experiment when it was obvious Poison(4)+Rainbow(15) did not result in 10-20% reg. hybrids anywhere, unlike Poison(4)+Masque/Canvas(15) in the social cave.
After 2x ~150 tries with Poison(4)+Masque/Canvas(15), I suspect the way the odds ('set %') are scaled down (when the total set % is above 100%) is more complicated than just one factor for everything. If all odds, parents' included, were reduced by the same factor, I should have seen ca. 70% rare dragons in the case of Masque. Excluding the parents' odd and scaling down the rest, I should expect 63% rare dragons. But the observed number was just over 50%. Same thing with Canvas: 76% rare dragons were expected if the odds scale down with the same factor, but the actual result was 63%.
I suspect the % of reg. hybrids decreases proportionally to the increase in total set %. Canvas has 5 elements and the total set % is ~130%. I got 14% reg. hybrids with Canvas+Poison. Masque has 6 elements, the total set % is ~220%, and I got 8% regular hybrids with Masque+Poison. In the Rainbow+Poison test (all elements) I got 0% reg. hybrids in 63 tries.
I'm trying to think of a way to take this further. But I also feel like I need a break after the Valeice event.
Post by Fletch_smf on Mar 11, 2021 15:36:37 GMT -6
The Snooz dragon has a 10% cloning chance in the CBC, and adds all 10 elements into the mix when trying to clone it. If people want to track their attempts we might be able to shine some light on whether the 10% is affected, or not, by the overall set percentages of all the possible dragons.
During the week that Magnolia was available I've been running a test in the Social Cave: Cactus(4)+Midday(20) vs. Plant(4)+Midday(20).
Both combinations contain the primary elements {plant, lightning, cold} and the {epic/sun} element from Midday. At first I was just testing Cactus(4)+Midday(20), and getting a lot of Cactuses as expected. (Cactus has a 10% single parent cloning chance in the social cave, and is also a regular, possible result with those elements. Midday has a 1% cloning chance and was not breedable). But then I got curious about what would happen if I used a parent that could not be cloned; basically, which other dragons would I get more of? So halfway through the week I switched to Plant(4)+Midday(20). It was always Cactus or Plant as the "local" dragon. I also ran some test in the rift with Cactus(4)+Sun(15) vs. Cactus(20)+Sun(15) vs. Lycan(20)+Sun(15), but I'm still analysing the numbers, so more on that at a later time. But below are the Cactus vs. Plant results.
Cactus(4)+Midday(20)
#
%
Midday
4
1.8%
Cactus
49
22%
Hybrids (excl. Cactus)
87
40%
« average pr hybrid dragon: 8.0%
Rares
69
32%
« average pr rare dragon: 5.3%
Epics*
9
4.1%
*Magnolia
4
1.8%
total
218
*Sun/Moon
4
1.8%
*Rainbow
1
0.5%
Plant(4)+Midday(20)
#
%
Midday
2
0.9%
Cactus
20
8.6%
Hybrids (excl. Cactus)
111
48%
« average pr hybrid dragon: 9.4%
Rares
80
34%
« average pr rare dragon: 5.7%
Epics*
19
8.2%
*Magnolia
9
3.9%
total
232
*Sun/Moon
9
3.9%
*Rainbow
1
0.4%
So, while I don't want to over-interpret the results, if the average for hybrid dragons (in this situation) is ca. 9% without Cactus as a parent, then we're getting more than just 10% extra Cactuses when it's a parent. The extra Cactuses (beyond the ca. 9%) seem to come at the expense of all the other non-parent dragons (as opposed to e.g. just the other hybrids). The increase in Epics is interesting. I'm considering what to do next.
The Snooz dragon has a 10% cloning chance in the CBC, and adds all 10 elements into the mix when trying to clone it. If people want to track their attempts we might be able to shine some light on whether the 10% is affected, or not, by the overall set percentages of all the possible dragons.
When I tested Poison+Rainbow in all caves, I got ca. 14% Poisons in the CBC in ~65 tries.
Your margin of error on the 218 and 232 attempts is around +/-5%. You're seeing 14%, so it's going to land somewhere between 10 and 20. We need around 5000 results to get it down to a 1% margin of error.
Given the data we do have internally, we can take some good guesses based on where the number lands for many, many dragons. If we expect a '15' and we're seeing 17 with a +/- of 5, then we can be fairly confident that 15 is correct.
Number for cloning Cactus is not directly in the data. I've heard the 10% from several sources, including our previous tests, as the 'default' value when it is not found. Based on the 14 you're seeing, I wouldn't be surprised if the number is 15 and not 10, but it could be any of those numbers in that range and it doesn't yet disprove 10. All of the data is very interesting though!
During the week that Magnolia was available I've been running a test in the Social Cave: Cactus(4)+Midday(20) vs. Plant(4)+Midday(20).
Both combinations contain the primary elements {plant, lightning, cold} and the {epic/sun} element from Midday. At first I was just testing Cactus(4)+Midday(20), and getting a lot of Cactuses as expected. (Cactus has a 10% single parent cloning chance in the social cave, and is also a regular, possible result with those elements. Midday has a 1% cloning chance and was not breedable). But then I got curious about what would happen if I used a parent that could not be cloned; basically, which other dragons would I get more of? So halfway through the week I switched to Plant(4)+Midday(20). It was always Cactus or Plant as the "local" dragon. I also ran some test in the rift with Cactus(4)+Sun(15) vs. Cactus(20)+Sun(15) vs. Lycan(20)+Sun(15), but I'm still analysing the numbers, so more on that at a later time. But below are the Cactus vs. Plant results.
[Tables copied below]
So, while I don't want to over-interpret the results, if the average for hybrid dragons (in this situation) is ca. 9% without Cactus as a parent, then we're getting more than just 10% extra Cactuses when it's a parent. The extra Cactuses (beyond the ca. 9%) seem to come at the expense of all the other non-parent dragons (as opposed to e.g. just the other hybrids). The increase in Epics is interesting. I'm considering what to do next.
This is great stuff. Firstly, thanks for helping me see mistakes in my Magnolia post. I have fixed those now. Borrowing your table format, I have put in the numbers from my tables
Actual
Cactus(4)+Midday(20)
#
%
Midday
4
1.8%
Cactus
49
22%
Hybrids (excl. Cactus)
87
40%
« average pr hybrid dragon: 8.0%
Rares
69
32%
« average pr rare dragon: 5.3%
Epics*
9
4.2%
*Magnolia
4
1.85%
total
218
*Sun/Moon
4
1.85%
*Rainbow
1
0.5%
Projected
Cactus(4)+Midday(20)
%
Midday
1%
Cactus
18.5%
Hybrids (excl. Cactus)
43%
« average pr hybrid dragon: 8.5%
Rares
30%
« average pr rare dragon: 5%
Epics* (excl Midday)
7.5%
*Magnolia
3%
total
*Sun/Moon
3%
*Rainbow
1%
* Double Rainbow
0.5%
Limited/Epics are a few % under what we expect here.
Actual
Plant(4)+Midday(20)
#
%
Midday
2
0.9%
Cactus
20
8.5%
Hybrids (excl. Cactus)
111
48%
« average pr hybrid dragon: 9.4%
Rares
80
34.4%
« average pr rare dragon: 5.7%
Epics*
19
8.2%
*Magnolia
9
3.9%
total
232
*Sun/Moon
9
3.9%
*Rainbow
1
0.4%
Projected
Plant(4)+Midday(20)
%
Midday
1%
Cactus
10%
Hybrids (excl. Cactus)
51.5%
« average pr hybrid dragon: 10.25%
Rares
30%
« average pr rare dragon: 5%
Epics* (excl Midday)
7.5%
*Magnolia
3%
total
*Sun/Moon
3%
*Rainbow
1%
* Double Rainbow
0.5%
Limited/Epics are much closer to what we expect. It could be that we are just seeing the variance swinging down with Cactus and up with Plant, which is making the actual seem more pronounced than the projected.
Your margin of error on the 218 and 232 attempts is around +/-5%. You're seeing 14%, so it's going to land somewhere between 10 and 20. We need around 5000 results to get it down to a 1% margin of error.
Not sure if I misunderstand, but the ca. 14% I mentioned was for Poison in Poison+Rainbow in the social cave. (14% in ~65 tries). This was part of a different series of tests (mentioned above, Mar 7). The point was to see if the number for the single parent chance in the social cave for a regular hybrid (10% for Poison) got reduced to almost 0% like other reg. hybrids. (It didn't). But I should emphasise that I definitely don't consider "14% Poison" to be a bankable result. In the context of the Poison+Rainbow test it was just a short (sloppy?) way of saying "not almost 0%".
Regarding the 'Magnolia Experiment', it was just something I did out of curiosity while at the same time trying to get a Magnolia twin for my main park. I wanted to see if there was something to see, basically.
This is great stuff. Firstly, thanks for helping me see mistakes in my Magnolia post. I have fixed those now. Borrowing your table format, I have put in the numbers from my tables
Cactus(4)+Midday(20) ... Limited/Epics are a few % under what we expect here.
Plant(4)+Midday(20) ... Limited/Epics are much closer to what we expect. It could be that we are just seeing the variance swinging down with Cactus and up with Plant, which is making the actual seem more pronounced than the projected.
Thanks for doing the calculations. To be completely honest, I didn't think of Rainbow (and Double Rainbow) before I bred the first one myself. In theory I knew it was probably possible (with sun/epic as the fourth element), but I don't think I've ever seen it happen outside the Rift.
Regarding the Epic/Limited numbers - yes, the difference is intriguing, but the numbers are a little too low to draw firm conclusions. I'm tying to think of another good way to test what happens to the "parents' percentages" when the parents can't be cloned.
I tested Cactus(4)+Sun(15) vs. Cactus(20)+Sun(15) in the Rift during the week that Magnolia was available. I wanted to see if the level of Cactus had any influence. Without further ado, here are the results:
Sun(15)+Cactus(4)
#
%
Sun
22
7.2%
« Parents combined 17%
Cactus
30
9.8%
Hybrids (excl. Cactus)
152
50%
« average 9.9%
Rares (excl. Rifties)
89
29.1%
« average 4.8%
Rifties
3
1.0%
Epics* (excl. Sun)
10
3.3%
*Magnolia
6
2.0%
total
306
*Moon
0
0%
*Rainbow
3
1.0%
*Double Rainbow
1
0.3%
Sun(15)+Cactus(20)
#
%
Sun
20
5.8%
« Parents combined 18%
Cactus
43
12.4%
Hybrids (excl. Cactus)
150
43%
« average 8.7%
Rares (excl. Rifties)
107
31%
« average 5.2%
Rifties
2
0.6%
Epics* (excl. Sun)
24
6.9%
*Magnolia
10
2.9%
total
346
*Moon
0
0%
*Rainbow
11
3.2%
*Double Rainbow
3
0.9%
Taken at face value, the results suggest you get twice as many epics (6.9% vs 3.3%) with the higher level Cactus. Also a few more rare hybrids. This comes at the expense of regular hybrids (down from 50% to 43%). But you also get extra Cactuses (the parent) - up from 9.8% to 12.4% - which, notably, is a bit higher than the average for other reg. hybrids – 8.7%. I'm on the fence about this, it could just be variance. But interestingly, the total parent percentage doesn't vary much (17% vs. 18%).
Footnotes: Rifties – I tried to avoid them by changing traits etc., but sometimes the alignment was against me. Moon – I know it complicates matters a bit when Moon/Sun are interchangeably breedable and Sun is a parent, but ~90% of the results are "daytime" result so that Moon wasn't breedable.
'... overwhelming goodwill, and witty humour, profound grace and compassion blended with the spice of relentless taunts.'
Note: This is an independent forum and is not affiliated with Deca Games, the developer of DragonVale.
It was created by and is managed by DragonVale fanatics.
All who observe the following brief guidelines are welcome to this great community: no personal attacks & no spam.
