5th Gen Breeding Guide (WIP)

[TeamofNoobs]

http://www.psypokes.com/lab/ivguide.php

You should also post this link in there because it shows a very easy way to check if a pokemon possibly has 31 IVs, without having to take it to the Battle Subway/Frontier. I talso tells someone who doesn't know about IVs that much more information. It says if it has multiple 31 IVs, it picks a random stat, but I'm not sure if it picks a random 31 IV stat, or if it just picks any stat, no matter what the IV, I'm probably gonna check it.

 

[cyberphoenix]

This is helpful, but there is a problem with the number of Ditto you need to catch to get one with a 31 in each stat. You said it took about 36 but the actual number is 79 because after each 31, you don't want another 31 in that stat so the probability of getting a 31 in a stat that you don't already have is lower.

 

[TeamofNoobs]

Can you show the math for this, I don't know how either numbers were calculated, I just want to know which one is right. First about 6 in every 32 wild pokemon will have a 31 IV somewhere, but I am not sure how to put that into an answer for how long it will take to get 31 IVs in all places.

 

[Sprocket]

Think about it like this. Every time you encounter a wild Pokemon, the game rolls a 32-sided die six times to determine the IVs (0-31). You have six chances to roll a 31, so your odds are 6/32 = 3/16 = 18.75% chance of at least one IV being 31. Inverse that (16/3), and you get 5.333... So...out of every 5.333 Pokemon, you are likely to find at least one with a perfect IV.

3/16 is just larger than 1/6 (.16666...), so for simplicity sake you can say about every 1 out of 6 Pokemon.

 

[TeamofNoobs]

Yeah, I think that is right, but you might get a repeat, like two 31IVs in speed or defense or whatever.

I thought of it as 6 bags, each bag has 32 marbles, and one of those marbles is golden, and the odds of you getting one golden marble out of any of those bags is 6 in 32.

 

[DysfunctionalDragons]

You have six chances to roll a 31, so your odds are 6/32 = 3/16 = 18.75% chance of at least one IV being 31. Inverse that (16/3), and you get 5.333... So...out of every 5.333 Pokemon, you are likely to find at least one with a perfect IV.

3/16 is just larger than 1/6 (.16666...), so for simplicity sake you can say about every 1 out of 6 Pokemon.

Previous lurker registering to say that probability does not work this way. Probability is multiplicative, not additive. You don't 1/32 six times to get 6/32; If it worked the way you said, then in 32 different "rolls" for the IV you'd be guaranteed to get a 31, which is not how it works. Instead, you need to raise right number to the right power. Of course, (1/32)^6 calculates the probability of ALL IVs being 31, which is not what we want.

Think of it this way: a good way of calculating probability in an instance like this is calculating the odds that the given event won't happen, and then subtracting that from 1.
There's a 31/32 chance that any given IV of any given Pokémon will NOT be 31. Let's say we're catching Dittos. If I catch any old Ditto, the odds of it NOT having an HP IV of 31 is 31/32. If we want to calculate the odds of it having HP and Atk IVs that are both NOT 31, we multiply them:

(31/32)*(31/32)=0.93848

Pretty good chance! What about having six IVs, NONE of which are 31? Which, by inverse, you could calculate the chances of your Ditto having ANY 31 IV (one or more)? Then you use this formula:

1-(31/32)^6=0.17345

Which is also pretty close to 1/6, and the actual correct value for the odds. In this case it doesn't matter too much, since it's close to the value you got, but if you're doing further calculations then you really need to calculate probabilities the right way, or you'll run into trouble. Otherwise you'll get people whining that they already hatched 8,192 eggs, but why haven't they hatched a shiny yet? Because they don't really understand probability. I don't blame them, as it's not intuitive stuff.

Anyway, in terms of what's changed between Gen IV and Gen V, the step counter for the eggs doesn't seem to reset when you talk to the day care man. I had an egg hatch when I walked about 20 steps away after taking an egg from him. Needs confirming though.

 

[cyberphoenix]

Okay, so this is how to fully calculate how many Ditto are needed.

Getting the first 31 IV has probability 1-(31/32)^6=0.17345.
Getting a 31 IV if you already have a ditto with 1 perfect IV is 1-(31/32)^5=0.146785 because we don't care what the last IV is anymore.
If you have 2 perfect IVs getting a 3rd different one is 1-(31/32)^4=0.119262
4th is 1-(31/32)^3=0.0908508
5th is1-(31/32)^2=0.0615234
6th is 1-31/32=0.03125

This means getting the first Ditto will take 0.17345^-1=5.76542 Ditto.
Once you have 1 perfect IV, the second Ditto will take 0.146785^-1=6.81269 Ditto.
The 3rd IV will take 8.38492 Ditto.
4th is 11.0071
5th is 16.254
6th is 32.
Ignoring the small chance of getting a Ditto with multiple 31's the total number of Ditto to catch is the sum of these which is 80.2241 Ditto to get Ditto with a 31 in each IV.

 

[DysfunctionalDragons]

Actually, it's a little more complicated than that, because there is no magic number you can catch that will give you an 100% chance of getting the Ditto you want. You can get it pretty close, however, and even calculate how close you're getting.

For example, we determined that the odds of catching a Ditto with a perfect IV in any of it's stats is about 17.345%, or 1 in 5.76542. But if you caught 6 Dittos, what's the actual probability that one of them will have a perfect IV? For this you nest the probabilities.

1-((31/32)^6)^6=0.68112

You can use this formula for calculating probability at any step of the process. Chaining Dittos in Diamond and Pearl, and want to know good your chances are if you catch 40 Dittos?
For any 31 IV at all, it's a bit overkill: (note that it's still not certain)

1-((31/32)^6)^40=0.99951

But for a specific 31 IV, it gives a very respectable 71.915%:

1-((31/32)^1)^40=0.71915

In fact, that specific-31 formula is what you need to check the probability of getting a complete set. To have a full set, you need:
At least one 31-HP Ditto AND at least one 31-Atk Ditto AND at least one 31-Def Ditto AND at least one 31-SpAtk Ditto AND at least one 31-SpDef Ditto AND at least one 31-Speed Ditto
Which means you're taking that probability of having at least one Ditto with specific 31-IV stat and raising it to the sixth power. That gives you this formula:

(1-(31/32)^n)^6=p

Let's try that 80 cyberphoenix got for n, and it comes out to this:

(1-(31/32)^80)^6=0.61082

So, even if you need an average of 80 captures to get that complete set, about 39% of the time you're going to need to go back into that tall grass and look for even more. Probability stinks, doesn't it?

 

[Sprocket]

And that is why I'm an engineer, not a mathematician.

 

[DysfunctionalDragons]

Uh huh. I understand. But it would be really nice if your guide (which has a lot of useful information in it, by the way) had accurate numbers, and most of the math in it right now is wrong. Do you want me to run through it and recalculate everything for you, or do you think you can handle it?

 

[TeamofNoobs]

Wait, i think that the 1 in 72 is right because the first one is 6/32 which is rounded up to 1/6, then 5/32, which is rounded up to 1/7, then 4/32 which is 1/8, then 3/32, which is rounded up to 1/11, then 2/32 which is 1/16, then 1/31, which is 1/31, then we add the denominators, to get 72, so we have 1/72., but that's probability, so it could be more or less. when I say rounded up, I mean the denominator is rounded up, because 6/32 is 1/5.33..., which is rounded up to 6.

 

[DysfunctionalDragons]

Wait, i think that the 1 in 72 is right because the first one is 6/32 which is rounded up to 1/6, then 5/32, which is rounded up to 1/7, then 4/32 which is 1/8, then 3/32, which is rounded up to 1/11, then 2/32 which is 1/16, then 1/31, which is 1/31, then we add the denominators, to get 72, so we have 1/72., but that's probability, so it could be more or less. when I say rounded up, I mean the denominator is rounded up, because 6/32 is 1/5.33..., which is rounded up to 6.

I can't make sense of your post. Here, let me explain why 6/32 wasn't right to begin with, because this isn't a situation in which you can just add the probabilities and get the right answer.

Look at the Wikipedia page for probability and scroll down to the table under Mathematical Treatment called "Summary of Probabilities". It shows some basic statements and how to calculate them mathematically. I don't know how to make the fancy set notation on a forum post, but I can run down how this applies to Pokémon.

A is a probability (let's say, the probability of the HP IV of being 31). It is what it is (1/32). B is another probability (how about the Atk IV being 31?). It is what it is (Also 1/32, and completely independent of A). And so on, for C, D, E, and F. "Not A" is the probability of the HP IV not being 31 (1 - 1/32 = 31/32). "A or B" is what we're interested in (well, "A or B or C or D or E or F", but I'm trying to keep it simple here), but since A and B are not mutually exclusive of each other (both values can be 31, after all), we can't just add them.

In other words, we want either the HP IV to be 31, or the Atk IV to 31. They can both be 31, but we can't just add the two because in the instance that they both are 31, then that 31 would be counted twice which skews the probability.

This gets a little weird when you start taking many different IVs into account, because there are a lot of different possibilities to account for. Conveniently, though, we have another way of stating "A or B or C or D or E or F".

If we can make the statement: "There is not a single randomly-generated IV on this Pokémon that is 31"

Then the statement: "There is at least 1 IV with a value of 31" is NOT TRUE.

Therefore:

Not ("not A" and "not B" and "not C" and "not D" and "not E" and "not F") = A or B or C or D or E or F

Not ("not A" and "not B" and "not C" and "not D" and "not E" and "not F") can be shown mathematically with (31/32)*(31/32)*(31/32)*(31/32)*(31/32)*(31/32)

That's where the (31/32)^6 came from.

Make sense?

 

[hsb39]

Just checking something with the conversion of ratios to fractions, you said:

"1:125, or 1 out of every 125 eggs", but from my limited knowledge, 1:125 = 1/126.

Think of 1:2, that is not 1 out of every 2, clearly, as the first has a smaller chance of occurring than the second. It is actually a 1 in 3 chance, as there are a total of 3 possibilities. Similarly, for 1:125, there are 126 different chances, the one for "1" and the one hundred and twenty five for "125". So that means that it is 1 out of every 126 eggs.

More generally, x:y = x/(x+y)

 

[ozzy]

What exactly is set when you receive the egg? I thought the nature was set when the old man gives you the egg but after resetting, I got a new nature and a different set of IVs.

I saved right before getting a Larvesta egg,
the first time it was a Tmid nature but after resetting, the nature changed to Sassy.

Does this mean I can't soft-reset for different IVS for the same nature?

 

[Sprocket]

To all of you bitching about the math, I'll go over it again this week when I revise the guide as a whole. Honestly that's not as important as the rest of the guide. 'Kay?

What exactly is set when you receive the egg? I thought the nature was set when the old man gives you the egg but after resetting, I got a new nature and a different set of IVs.

I saved right before getting a Larvesta egg,
the first time it was a Tmid nature but after resetting, the nature changed to Sassy.

Does this mean I can't soft-reset for different IVS for the same nature?

Correct. Unlike previous generations, in this generation everything is determined when you pick up the egg: the species, gender, nature, ability, and IVs.

 

[cyberphoenix]

I know the math isn't that important and the rest of the guide is extremely helpful, however it should be close, and I only brought up the issue with Ditto because it was something that was off by a large amount.

 

[UpsideDownHitmontop]

I've been reading up and breeding and apparently in the older versions one stat was passed from a parent then a second stat other than HP then a third other than HP or Defense. Does it still work this way? For example, if I am breeding with the power anklet the first stat passed would be speed, so when the second stat to be passed is chosen, would HP be excluded?

 

[Yeah Galo Sengan]

Just a head's up, I have been breeding a Dream World female Dratini with a male Axew and hatching Dream World ability Dratini roughly 40% of the time, I am assuming only Ditto makes it impossible to breed for abilities.

 

[NijaSkills]

Unfortunately, the ditto catching method that you suggested no longer works as well. Ditto keeps its own catch rate no matter what it transforms into, meaning that turning it into magicarp is only useful for giving ditto no attacking moves. An effective method I found was to use munna as a synchronizer, and let it carry the smoke ball, then use yawn on the ditto as it transforms, and switch to my false swipe user when it falls asleep and hope it doesn't use yawn. I carry a bunch of awakenings just in case. Since munna is lv 11 or so, ditto has stupidly low defense, so false swipe easily brigs it down. Then it's an easy catch, since it's asleep at 1 hp.

 

[macle]

Guys please don't ask questions about breeding here. Use the Simple question thread in Dragonspiral Tower forum.

 

[Strants]

Since I assume there would also be reason to have a 0 speed ditto (for trick room teams), here are the values for zero speed IV's:
Neutral Nature
Level|Speed|Precision
53|55|Precisely 0
54|56|Precisely 0
55|57|Precisely 0
56|58|Precisely 0
63|65|Precisely 0
64|66|Precisely 0
65|67|Precisely 0
66|68|Precisely 0
67|69|Not precise

Hindering Nature
Level|Speed|Precision
53|50 Not|Precise
54|51|Not Precise
55|52|Not Precise
56|53|Precisely 0
63|58|Precisely 0
64|59|Precisely 0
65|60|Precisely 0
66|61|Precisely 0
67|62|Not Precise
68|63|Not Precise
69|64|Not Precise
70|64|Precisely 0

Helpful Nature
Level|Speed|Precision
53|61|Precisely 0
54|62|Precisely 0
55|63|Precisely 0
56|64|Precisely 0
63|72|Not Precise
64|73|Not Precise
65|74|Not Precise
66|75|Not Precise
67|76|Not Precise
68|77|Precisely 0

 

[cantab]

If anything, stats for low speed IVs are more needed, since the IV judge won't tell you those.

I'm still unsure as to the best way to capture Dittos, but I'm having good success leading with a Synchroniser that knows only Thunder Wave. T-Wave, then throw Dusk Balls. The capture isn't 100%, but usually only takes one or two balls. Of course your Synchroniser gets paralyzed itself.

If you're prepared to get your Synchronisers by breeding, a Munna with Yawn or Hypnosis and Sleep Talk might work similarly.

 

[wekhter]

If a Pokemon with the abilities Magma Armor or Flame Body are at the front of your party, they will cut in half the amount of steps needed to hatch an egg.

Only the front? I thought it was anywhere in the party?

Also, it might be worth mentioning that eggs can now hatch one right after the other instead of having to wait a whole egg cycle.

 

[Elite_Battle_Master]

Didn't see this mentioned anywhere but is it true that you can't get the DW ability when breeding w/ a Ditto, even if the DW Pokemon is female?

Some guy I was chatting with earlier mentioned it, just wanted to make sure if it was in fact true.

 

[Kitsuner]

Anyway, in terms of what's changed between Gen IV and Gen V, the step counter for the eggs doesn't seem to reset when you talk to the day care man. I had an egg hatch when I walked about 20 steps away after taking an egg from him. Needs confirming though.

I noticed the same thing on several occasions. Sometimes he gives me a second egg before I've gotten down the stairs, and I've hatched eggs within two or three steps of each other a few times as well. I haven't actively tested the new mechanics, but something's definitely changed.