A Discovery I made

Status
Not open for further replies.

Kal

Hello Smogon. This is my first post, and I'm not sure where to put this, but...
I discovered what role Lucky Eggs play in breeding.

Once upon a time, I heard a rumor on GameFAQs that Lucky Eggs had "a role" in determining a child's stats.
In trying to make myself a Ninjask, I bred eight generations of Nincadas. Based on the rumor that Lucky Eggs are beneficial, I made the Father hold one each time. (The mother was holding an Everstone, to pass on the Jolly nature that I wanted).

I kept track of each generation and their IVs in a nice little chart, but its formatting makes it impossible to reproduce in this post. So I'll post a minor truncation of the data.


I started with a Jolly Ditto ("Jolly") and a Ninjask ("Cascada"). I kept the first seven Jolly offspring, and gave them 14 Rare Candies (that being the number I had on me), then checked their IVs.
I found each child's Average IV by average together the ranges of their possible IVs. I average together each litter's Average IVs to find the Generation Average. I compared this number to the Parent Average, which is the average of only the two parents.
Jolly and Cascada had seven children, named 1, 2, 3, 4, 5, 6, and 7. Their Average IVs were, respectively, 13.25, 16, 4.75, 10.33, 10, 18.75, and 17.42. The Generation Average was 12.93, and the Parent Average was 13.11. This is a difference of only 0.18!

I made a total of four generations with the lucky egg, and had those same parents breed also without the Lucky Egg. This is the results I found:

Parents: Jolly + Cascada (Everstone and Lucky Egg)
Children: 1, 2, 3, 4, 5, 6, and 7
Parent's Average: 13.11
Children's Average: 12.93
Difference: 0.18
Standard Deviation: 3.56

Parents: 1 and 6 (Everstone and Lucky Egg)
Children: A, B, C, D, E, and F
Parent's Average: 16.00
Children's Average: 15.92
Difference: 0.08
Standard Deviation: 3.67

Parents: D and E (Everstone and Lucky Egg)
Children: Alpha, Beta, Gamma, Delta, Epsilon, Zeta, Theta, Iota, Kappa
Parent's Average: 13.91
Children's Average: 13.86
Difference: 0.05
Standard Deviation: 3.97

Parents: Kappa and Alpha (Everstone and Lucky Egg)
Children: Do, Re, Mi, Fa, and So
Parent's Average: 14.59
Children's Average: 14.82
Difference: 0.23
Standard Deviation: 2.01

Parents: 1 and 6 (Everstone only)
Children: Phantom, Clones, Sith, Hope, Empire, and Jedi
Parent's Average: 16
Children's Average: 17.26
Difference: 1.26
Standard Deviation: 3.90

Parents: D and E (Everstone only)
Children: Land, Creature, Instant, Sorcery, Artifact, Enchantment, Tribal, and Planeswalker.
Parent's Average: 13.91
Children's Average: 17.41
Difference: 3.50
Standard Deviation: 4.14

Parents: Kappa and Alpha (Everstone only)
Children: Point, Circle, Ellipse, Parabola, Hyberbola, Line, Cross
Parent's Average: 14.59
Children's Average: 17.48
Difference: 2.89
Standard Deviation: 3.88



What I concluded from this: If a father holds a Lucky Egg, the average of the child's IVs will be very close to the average of the parents' IVs! You can see that the Lucky Egg children had a small difference between Generation and Parent IVs, plus a small standard of deviation. Those without the Egg had random variation.
Is this new and/or helpful?
 
Uh, that could just be coincidental, and probably is. Notice how the averages are all in the 12-16 area? Yeah...

Try more conclusive tests, like breeding with different averages that aren't in what you would expect a generic average to be like.
 
So... it's a coincidence that a Lucky Egg made the inherited IVs follow the parents' IVs across the entire range, and that a lack of a Lucky Egg made for no correlation at all?
Forthcoming is the generation whose parents had 23 or more IVs on average. I'll let you know what comes up.
 
Interesting... Ill test this later with my perfect IV ditto (friend hacked it for me) adnd a feebas... feebas has terrible IV's... so attach egg to ditto and evertstone to feebas.
 
Coiled Flame said:
Interesting... Ill test this later with my perfect IV ditto (friend hacked it for me) adnd a feebas... feebas has terrible IV's... so attach egg to ditto and evertstone to feebas.
Click to expand...
If the model holds true, you should get a Feebas with middle-of-the-road IVs.
Also note that the weakest Lucky Egg fit was when my Ditto was involved. That might indicate that a female and male are necessary, and not ditto. Time will tell?
 
Time and a bit of further testing. This could have some pretty interesting consequences if it turns out to be non-coincidental.
 
If that's true, a 31/31/31/0/0/0 parent with a 31/31/31/0/0/0 parent, for example, will never yield a 31/31/31/x/x/x child with IVs other than 0 for the last three stats? I highly doubt that.
 
If the model holds true, you should get a Feebas with middle-of-the-road IVs.
Also note that the weakest Lucky Egg fit was when my Ditto was involved. That might indicate that a female and male are necessary, and not ditto. Time will tell?
Click to expand...

So you are suggesting that a perfect IV ditto + a crappy IV feebas will result in a middle of the road IV? That is, in my esteemed opinion, crap. Why would you bother with that? From what I've read you'd have more luck with just randomness in getting reasonable competative IV's.

Ditto acts as a male/female, even in IV passing. I got a wicked Umbreon (like 29 HP, 30 D and 31 SD) IV from breeding a crappy as crap Eevee with a wickedly defensed Ditto. That indicates to me that, according to your results, using the lucky egg is crap compared to randomness. Who should bother?

I'd also like to submit this: Your AVERAGES seem to be close to the parents averages. Fantastic. Well, how about this: If you want an awesome Attacking/speed, lets say scizor, and your average is 16, you can end up with a shitty scizor with high IV's in like SD and SA and crappy IV's in ATK and SP. That's not really what you wanted now, is it? Perhaps look at more than the crappy averages which mean nothing cause pokemon are bred for certain IV's, not averages, and look at how it impacts IV's you are specifically looking for over several generations.
 
Looks like pure coincidence, stretched into something that you think is a discovery.

Just so you know, all babies gain exactly 3 random IVs from their parents, that remain *identical*. This means that, for example, two perfect-IV parents are guaranteed to have children with at least 3 perfect 31 IVs. Are you trying to say that in such a scenario, this would function differently, perhaps giving 4 perfect 31 IVs as a guarantee?

The problem with your "discovery" is that there is far too much variance in the numbers to find out if it's even remotely true. The parents' averaged IVs were below average, and all the children had IV averages that could be considered "average" or lower; These are likely results similar to what you'd get if you caught random Pokémon in the wild. I don't see how any of these numbers have any sort of proof to your theory...
 
What it means: if a 31/31/31/0/0/0 mates with a 31/31/31/0/0/0, the average of both their IVs is 15.5. Their children will average about 15.5 in their IVs, for example 17/19/16/14/15/12.
 
Kal said:
What it means: if a 31/31/31/0/0/0 mates with a 31/31/31/0/0/0, the average of both their IVs is 15.5. Their children will average about 15.5 in their IVs, for example 17/19/16/14/15/12.
Click to expand...
So if both the parents average 31, you're saying the children will always be very close to 31?

Average. IVs. Are. Bullshit.
You know what averaging the IVs does? It gives you an "average" number. Middle of the road. It means that your average IVs are much more likely to be close to 15.5 than to anything else, because the numbers are RANDOM, and the center of the numbers that can be randomized is 15.5.

I'd also like to note that if you hear something weird on GameFAQs, it's very, very, VERY safe to assume that it's bullshit.
 
wow, the community doesn't seem to like anything new.
fine, it's probably better that I don't say anything again.
 
The Smogon community generally likes exhaustive testing on every case to rule out other hypotheses, or simple misunderstanding, before declaring something as true. Not unlike the scientific community, I might add...
 
It's not anything new.

I think it's worth testing, although we want a ridiculously huge sample to test this properly.

Like, 100 magikarps or something.
 
i can assure you that your findings are crap because i already tried the theory with lucky egg + sea incense...mind you it was a year ago
 
Theres a simple way to test your theory. Have someone hack two 31/31/31/31/31/31 parents and breed them with lucky egg attached, then base conclusions on the results.
 
L2m6w5 said:
Theres a simple way to test your theory. Have someone hack two 31/31/31/31/31/31 parents and breed them with lucky egg attached, then base conclusions on the results.
Click to expand...


Actually, I think three groups of parents will be needed for this.


Group A: 31 across the board

Group B: 15 Across the Board

Group C: 0 Across the board.
 
Surgo said:
The Smogon community generally likes exhaustive testing on every case to rule out other hypotheses, or simple misunderstanding, before declaring something as true. Not unlike the scientific community, I might add...
Click to expand...
It's called the intelligent way to do things.

Kal, I want to make this is simple and non-offensive as I possible can: Get real evidence in numbers too large to be simple oddities, and use parents with IVs that are actually good. If you do both of those things and the numbers come out similar, then maybe, just MAYBE, you're onto something.

Think about this:
If both parents have straight 31 IVs, the lowest IVs the baby can have is 15.5 average (31/31/31/0/0/0). This means that in ALL cases, it will be closer to the parents IVs than to the opposite extreme. Similarly, if the parents are straight 0 IVs, then the minimum of 3 IV spassed down from the parents means the max IVs you can get on the buggers would be 3 31s and 3 0s, meaning they can ONLY be from 0-15.5. Get what I'm saying yet?
The child's IVs already do average out near the parents' in a lot of cases, especially when the IVs of the parents are mediocre or average out "middle of the line". If the parents are exactly half 16 and half 15 for all their IVs (15/16/15/16/15/16), thus averaging 15.5:
The most the babies can have is 16/16/16/31/31/31. This averages to 23.5, only 7 higher than the parents average.
The least they can have is 15/15/15/0/0/0, averaging 7.5, which is 8 under the parents' average.

And these are the absolute extremes, which have likelihoods in the area of .00000001%. The MUCH more likely scenarios would average up to 3 or less away from the parents' average.

So, basically, your results aren't even a statistical oddity. Your results are EXPECTED with no outside influence. If you want even the most remote chance of proving this, you need a MASSIVE control test and a MASSIVE experiment test. Something along the line of two parents with perfect IVs and no egg breeding 100 children, and two parents with perfect IVs with the egg breeding 100.

In such a massive experiment with numbers that are much farther from "average", your result is either going to be far more defined or far less defined, and in either case it's proven right or wrong.

Have I explained my reason for not believing you well enough for you to stop bitching and do some real tests?
 
Wait... you found the Averages of the Children? >.> Can you give us *actual* IVs instead then. First give us the parents' IVs and then give us the children's then if we find any correlation we might just believe you. But you have to give *actual* IVs because average IVs don't help. Now, don't feel offended by this. I can understand what it feels like when you feel like you've discovered something big, it's just that... we need more than that. If you really do want to prove yourself right, then we need tests, big ones.
 
I concur with hypnotist, give us retardedly hard typed out IV's, individually packaged and served to us.

Also, we need Desperately to prove this with two test groups: one with Lucky Egg and the other without, using the same parents. sample sizes in this case of like 25 would be fine, but we need a comparison first hand without bias (go university level stats class! working for you!)
 
Humor me, I am stupid.

Your hypothesis is
If a father holds a Lucky Egg, the average of the child's IVs will be very close to the average of the parents' IVs!
Click to expand...

This is hard for me to grasp. Does this mean that one iv of a pokemon will the average of that pokemons parent's ivs in that stat? So if my pokemon M has 31 hp iv and my pokemon F has 31 hp iv, then they're baby, pokemon B, will have 31 hp iv? Since 31 is the average of 31 and 31? Please elaborate for me, as I am very interested in new game mechanics, especially ones having to do with breeding.
 
TVboyCanti said:
Humor me, I am stupid.

Your hypothesis is


This is hard for me to grasp. Does this mean that one iv of a pokemon will the average of that pokemons parent's ivs in that stat? So if my pokemon M has 31 hp iv and my pokemon F has 31 hp iv, then they're baby, pokemon B, will have 31 hp iv? Since 31 is the average of 31 and 31? Please elaborate for me, as I am very interested in new game mechanics, especially ones having to do with breeding.
Click to expand...

Essentially, that's what he's saying. I'm not sure that I believe it just yet, but some testing never hurt anyone.
 
Status
Not open for further replies.
Back
Top