sparkle sparkle

Posted August 26, 2020 at 10:19pm

I'm gonna break down how shiny odds are calculated in every single generation and how different methods of shiny hunting affect those odds because this is what I find fun for some reason.

Gen 1 has no shinies.

Gen 2 does have shinies. Your odds of a Pokemon you encounter being shiny are based on DVs. DVs are basically the Gen 1 and 2 equivalent of IVs, except instead of going from 0-31, they go from 0-15 (or in binary, which the game uses to calculate DVs, 0000-1111). This is the case for every stat, except Special Defense (which is always identical to Special Attack) and HP (which is calculated by taking the last binary digit of each other stat and making a new four digit binary string out of them). DVs dictate a lot of things in Gen 1 and 2, including gender, Hidden Power type, and Unown forms, but for right now, we only care that it also determines shininess.

In order for a Pokemon to be shiny in Gen 2, it must have a very specific set of DVs. Defense, Speed, and Special Atk/Def (same DV, remember?) must all be 10, while the Attack stat must be 2, 3, 6, 7, 10, 11, 14, or 15. You have a 1/8192 chance of randomly encountering a Pokemon with this stat spread in the wild.

Fun fact, a shiny Pokemon in Gen 2 can only have an HP DV of 0 or 8. That's not a rule for the shiny calculations, it's just a mathematical certainty based on what stats are necessary for shinies to be generated, and how HP is calculated. Also, it's impossible for females of any species with a 7-1 male to female gender ratio to ever be shiny. This includes all the starters, fossils, and Eevee.

So you might already see how we can increase our odds, since breeding for specific IVs is a huge part of competitive Pokemon. The easiest way is to start with a shiny Pokemon, and breed it, since it'll always pass down at least a few of its own DVs. If at least one parent is shiny, the odds of the baby also being shiny jump from 1/8192 up to a staggering 1/64! Luckily, there are a few easy methods to guarantee a couple very specific shiny Pokemon in Gen 2, so you can get a head start on breeding.

First method should be obvious. The shiny Gyarados at the Lake of Rage. It's hardcoded to always have the same set of DVs ensuring its shininess, so it still counts. Second method is the Odd Egg in Crystal. It seems that in English versions of Crystal, the egg has a 14% chance of hatching into a shiny Pokemon. Keep in mind that the Pokemon in the egg is set as soon as you receive it, so if you intend to soft reset for a shiny Odd Egg Pokemon, save before you get the egg, not before it hatches.

The third method is a bit involved, and does require some minor glitches, but is absolutely the most valuable if you want to breed for shinies in Gen 2. To start, catch the shiny Gyarados at the Lake of Rage, and then transfer it to a Gen 1 game. Although shinies don't exist in Gen 1, the way the game calculates DVs is the same, so a Pokemon that would be shiny in Gen 1 will gain shininess upon being transferred to Gen 2. In Gen 1, Gyarados can learn Mimic through TM. You'll want to teach it Mimic, and then deplete its PP for all other moves. Then, encounter a Ditto.

For reasons still unknown to us, if a Ditto transforms into a Pokemon that it's already transformed into in Gen 1, it copies the target's DVs. Not just for the duration of the battle either, it completely overwrites its own DVs with those of the opposing Pokemon permanently. If you use Mimic on the first turn of the battle, your Gyarados will temporarily learn Transform, and when the Ditto then transforms into your Gyarados, the only move available to it will also be Transform. The Ditto transforms again on the next turn, and permanently copies the DVs of a shiny Pokemon. Catch the Ditto and bring it back to Gen 2, and hey presto, it's a shiny. You now have a near universal breeding partner that can give you 1/64 odds on any non-legendary shiny from the first two generations.

Whew, that was all just Gen 2. We have another six generations ahead of us, each one more interesting than the last... kinda.

Gen 3 is not, in fact, more interesting than the last. While the way shinies were calculated did change dramatically, there really aren't any methods to increase your odds in these games. So instead of going over shiny hunting methods, let's use this section to talk about what changed since the last Generation in terms of calculating shinies.

Gen 2's shiny calculations kinda sucked. It was easy to manipulate, didn't allow certain genders or forms of certain Pokemon to be shiny, made shinies competitively nonviable, and worst of all, involved math that humans can actually understand. Gen 3 fixed all of these by introducing a completely new system, seperated from DVs (now IVs) that no sane human would ever attempt to comprehend. I'm going to explain to you how it works.

If you go to your trainer card in any main series game, you will notice a Trainer ID number. This is nothing new, it's been there since the first games, but starting in Gen 3, you will now have a second, Secret ID number. Where is it? Completely inaccessible without cheating devices. That would be why it's called "Secret." Your Secret ID ensures that even if you share a name and ID number with another trainer, there's still a check to ensure that you're two different people. The odds of any two trainers having the same name, Trainer ID, and Secret ID are approximately 1 in 4,294,967,296. None of this matters for shiny hunting. Just know that Secret IDs are a thing, you have one, and you're not allowed to see it.

For calculations and RNG routines that are supposed to go on behind the scenes, this is good! It means there's now a value that the average player using only official hardware can't possibly access, and therefore manipulate to their own advantage. From this point on, shinies will be calculated using Secret IDs, and a little something else called Personality Values. Personality Values are like a Secret ID assigned to each wild Pokemon as soon as an encounter is generated. It's an unsigned 32-bit integer that blah blah blah...

Look, if you're interested in the specifics of the computer math, it's on Bulbapedia. You're probably more interested in what the Personality Value does. Quite a lot, as it turns out. Almost every random aspect not dictated by IVs uses the Personality Value. It determines a Pokemon's gender, ability, nature, Pokeathlon stats, Unown's letter, Spinda's spots, which Pokemon Wurmple will evolve into, and of course, whether or not the Pokemon is shiny. Most of these calculations are simple. The shiny calculations are not. I'll try to simplify to the best of my abilities, but feel free to skip the entire next paragraph if you want. You won't be missing much.

So first, the first 5 digits of the Pokemon's Personality value and the second 5 digits of the Pokemon's Personality Value are separated, and XOR'd together. XOR stands for "Exclusive or" and it basically means that if there are two of the same value in a column, that column becomes a 0, and if not, it's a 1. Then your trainer ID and Secret ID are converted to binary and XOR'd together, and finally the result of both XORs are XOR'd together one last time, and if the end result is the decimal equivalent of 8 or less, you have a shiny.

Congratulations, you made it through! Or you skipped it, in which case, congratulations anyway! This has all probably been a lot to take in, and if you're still with me, I'm glad. What's important isn't the math anyway, it's the probability, and even with this drastically different way of calculating shininess, the odds are actually still the same, 1/8192. Only difference is, now a Pokemon being shiny is completely independent from its ability, gender, stats, or anything else, and it can't be manipulated by breeding (yet). In Gen 3, there was no way to increase your odds, so let's move right along to Gen 4.

We still have the same algorithm and base odds in Gen 4 (and for the most part, we will from now on) but we did get a number of new shiny hunting options. First is the PokeRadar. This is available in all three Sinnoh games, and you get it after obtaining the National Dex. How the PokeRadar works is based on chaining, and get ready to hear that word a lot because it won't be the last you've heard from it. Chaining involves repeating the same process over and over, usually encountering the same Pokemon under a specific set of circumstances, to increase your shiny odds. The way your odds increase is by increasing the tolerance of the algorithm.

Reminder from the boring paragraph, or for anyone who skipped it, the end result of the calculations must be 8 or lower. With the PokeRadar, that number goes up, to 9 after a chain length of four, 10 after a chain length of eight, 11 after a chain length of ten, and so on at an exponential rate. This increases the tolerance of the shiny algorithm, making the range of Personality Values that can result in a shiny much wider. Your odds max out at 328/65536, or 1/200 per patch of grass (of which up to four can generate at once, for an effective maximum of 1/50) after a chain length of 40. So that's how the odds work, now for the actual technique.

Standing in a patch of grass, activating the PokeRadar generates between 0-4 shaking grass tiles around you. Walking into one will begin an encounter. Once in that encounter, defeat or capture the Pokemon, but DO NOT run away. Once the battle ends, another 0-4 patches should begin shaking automatically. If they do, you've started a chain. To continue the chain, you must encounter the same species of Pokemon repeatedly from the shaking patches of grass, and catching/KOing them. If you run away, the chain will break. If you encounter a different species than the one you encountered previously, the chain will break. If the PokeRadar generates zero shaking patches, the chain will break. If you quit the game, the chain will break. If you get on your bike, the chain will break. If you encounter a Pokemon outside of the shaking patches, the chain will break.

To preserve your chain as long as possible, there are four rules you should always follow. The first rule is to use Repels,as this will prevent Pokemon from appearing outside of shaking patches. The second rule is to never go into patches along the edge of the grass, as that increases the odds of generating zero new shaking patches. The third rule is to only go towards patches that look the same as the first one you entered, as going into a patch shaking harder or softer than the first one may cause a different Pokemon to appear. The fourth rule is to always go for the patch furthest away, as that increases the odds of the same Pokemon spawning. As a general guideline, never enter a patch closer than three tiles away. If no shaking patches generate that you feel comfortable with, run 50 steps and use the PokeRadar again to reset them. You can do this as many times as you like until you get a good patch, and it will not break your chain.

Once you reach a chain of 40 (a Poketch app can track your chain for you in these games), DO NOT enter any more shaking patches. Just keep resetting the shaking patches with the PokeRadar until you see one sparkle instead of shaking. That means there's a shiny in there. Walk into the patch, being very careful not to accidentally walk into any others on your way there, and claim your reward.

Another method introduced in Gen 4, and available in all subsequent games, is commonly referred to as the Masuda Method. Intended as a way to incentivize use of the new WiFi functionality of the DS, the Masuda Method provides a flat 5x multiplier to the odds of a hatched Pokemon being shiny, provided one of the parents comes from outside your game's language region. It's hard to use this method in Gen 4 these days, since Nintendo WiFi has been shut down and the GTS is no longer available in DS games, but it's worth noting that Masuda Method began here.

(Actually the way the odds increase in this method are actually quite interesting, since instead of increasing the tolerance for a successful shiny like the PokeRadar does, it just generates five different Personality Values for each egg, and if any one of them would be shiny, it keeps it. We'll call this increasing rolls, as opposed to increasing tolerance, and while it doesn't make any difference yet, it'll be important when we get to Gen 8)

Gen 5 introduced no new shiny hunting methods, and actually removed PokeRadar. Aside from the Masuda Method, the only way to increase your shiny odds in Black and White is by encountering wild Pokemon in double battles, which doesn't actually affect the algorithm but just generates two Pokemon with two different Personality Values. Technically this is an option in Gen 4 as well, but seeing as PokeRadar is far more effective, it's not used very often.

In Black and White 2 however, a very significant change was introduced: the Shiny Charm. A reward for completing the National Dex, the Shiny Charm is a key item that passively adds two rolls to every wild encounter or hatched Pokemon, and this stacks with other methods (bringing Masuda Method up to 8x normal odds, or 1/1024). While it isn't very noticeable in Gen 5, the Shiny Charm is present in every subsequent game, and its effects on other methods can be dramatic.

Okay real talk for a second, it's like 1 AM and I'm fairly confident nobody will ever read this. If they start to read it, they'll probably skim, and if they do start reading in depth, they likely won't make it this far. If you are reading this, thank you very much, this honestly took a lot of time and knowing somebody read it does kinda validate the effort that went into it. Since I know pretty much everyone is on the Discord now, could I ask a small favor? Just DM me the word "lemon." I know you don't really want to have a whole conversation about this topic after reading all this, and I don't want you to feel pressured to start one, but just giving me that little indication that you're still here and care what I have to say is more than enough. It'll honestly make my whole day knowing that someone was paying so much attention to my 5000 word rant for so long that they managed to make it far enough to read this. Okay, back to the math.

Gen 6 was a good time for shiny hunting. To start off, the tolerance of the shiny algorithm was straight-up doubled, from 8 to 16, with no strings attached. This brought the base shiny odds from 1/8192 to 1/4096, a change that has persisted in every subsequent game. The Shiny Charm was brought back as well, meaning that by simply walking around, you could expect base odds of around 1/1365, which was better than Masuda Method in previous games. Masuda Method itself now has base odds of around 1/819, which increases to 1/512 with the Shiny Charm, and this is still a readily accessible method for anyone with a Gen 6 game, since online services for the 3DS are still active.

As for new methods, the Friend Safari is quite possibly the easiest shiny hunting method released up to this point. While in a Safari, all encounters are given seven additional rolls, with or without the Shiny Charm. This gives Friend Safari the same odds as the Masuda Method has with the Shiny Charm, but with a much higher rate of encounters. The only significant limitation is the limited selection of available Pokemon, and the fact that you need friends to make it work :(

If you have no friends, Chain Fishing is another fairly easy, if limited, method of acquiring shinies in Gen 6. Simply fish up Pokemon over and over until one is shiny. You don't have to fish up the same one, but you do need to get consistent bites, as failing to fish up a Pokemon will break the chain. To make bites as consistent as possible, I would recommend keeping a Pokemon with the ability Suction Cups at the front of your party, and fishing near the edge of a body of water. Your odds can go as high as approximately 1/100 after a chain of 20, and this stacks with the Shiny Charm, though it’s not certain exactly how. The main drawback is, like the Friend Safari, the small number of available Pokemon.

In X and Y, the PokeRadar makes a return, with mechanics largely unchanged from Gen 4. Keep the rules from then in mind, and you should see similar success with the PokeRadar in Gen 6. The max odds are still 1/200 after a chain of 40, and the Shiny Charm doesn’t appear to affect the max odds. One major difference is that when you’re chaining, you may notice a faster paced music begin to play. When this happens, it means your odds have just been temporarily increased to the theoretical maximum. Ignore the chain and just reset your patches until one sparkles or the music slows back down to normal.

In ORAS, the PokeRadar is gone, replaced by the DexNav. How this works is by chaining hidden Pokemon encounters, and increasing your Search Level for that Pokemon. As the Search Level increases, your odds steadily grow, until reaching 0.08% at 200, after which they increase by an additional 0.01% for each 100 search levels, up to a maximum of 0.167% or 1/625 at Search Level 999. This appears to stack with Shiny Charm, which provides its usual two extra rolls per encounter, resulting in a flat 3x multiplier to the odds at every search level, up to a potential maximum of about 1/208.

Alright, remember in Gen 4 and 5 how a pretty brute force way of increasing your odds was to encounter wild Pokemon in double battles, basically giving you two rolls? Well you might’ve seen this one coming then, because this method just takes that, and turns it up to… five. Horde battles allow you to encounter five Pokemon simultaneously, effectively giving you five rolls per encounter, or equivalent odds of about 1/819. This stacks with the Shiny Charm as well, giving you an incredible 15 rolls at once, increasing your odds per encounter to about 1/273. Your base odds don’t actually change, but your rolls are dramatically increased, and for Pokemon available in hordes, this can be a very fast way to find a shiny one. If you weren’t aware, you can trigger a horde 100% of the time where available if you use the field move Sweet Scent or the item Honey.

I think that’s everything in Gen 6 covered, on to Gen 7. With how much the last generation gave us, this has to be even better right? Well, they removed PokeRadar, DexNav, Chain Fishing, Friend Safari, and Horde Battles, so we’re not really off to the greatest of starts. The Masuda Method is retained as always, and the new method introduced to us in Sun and Moon was SOS Chaining. How this works is that if a Pokemon is low on health (and especially if you use an Adrenaline Orb) it has a chance every turn to call for help, and there’s a chance that another Pokemon of its species will come to its aid. If you can get a Pokemon to do this many times consecutively, a chain will begin, and as the chain grows, you’ll be rewarded with more rolls on each new encounter, up to 13 rolls after a chain of 31, giving you effective odds of about 1/315.

A good strategy is to use False Swipe on the first Pokemon to bring it down to 1 HP, then use an Adrenaline Orb. Keep using Adrenaline Orbs (they will fail after the first and not be consumed, but still advance your turn) until the wild Pokemon calls for help, and when help arrives, kill the help. Continue killing the help until the help is shiny, then kill the original and catch the help. If shiny help doesn’t show up for a while, kill the original to keep it from struggling itself to death, then repeat the process on the help. Help doesn’t sound like a real word anymore.

Due to a glitch, your chain will automatically reset after 255 encounters. There’s really no workaround to this, it just kinda sucks. This method also stacks with the Shiny Charm, giving you an extra two rolls to each encounter like usual, up to 15 total rolls at a chain of 31, or effective odds of about 1/273. The requirements on obtaining the Shiny Charm were slightly relaxed this generation, requiring only a completed Regional Dex to obtain, rather than the National Dex. This is good.

Ultra Sun and Moon brought about a couple of changes. Firstly, they fixed the glitch that resets your SOS Chain after 255 encounters, which is nice. Secondly, they added a new, absolutely broken shiny hunting method, called the Ultra Wormhole, and unlike other shiny hunting methods that make you long for the cold embrace of death, this one is actually slightly fun! After completing the game, you gain access to a minigame where you ride Solgaleo or Lunala through Ultra Wormholes for as long as you can before entering portals with rings around them. I could explain what the different colours of portals and numbers of rings mean, but I’m tired and Bulbapedia exists. Basically all you need to know is that the more lightyears you travel, and the more rings around the portal you eventually fall into, the more your shiny odds increase, up to…

...Am I reading this right? After travelling 5000 light years and dropping into a three-ring portal, your shiny odds max out at 36%??? Sure, you can only catch about a dozen and a half Pokemon this way, but wow, 36%. If you have a copy of USUM and have never caught a shiny, you really have no excuse at this point.

Oh, and it may or may not be affected by Shiny Charm. I don’t know, but it doesn’t really matter, 36% is still absurdly high.

And finally, Gen 8. Masuda is back, SOS and Wormhole are gone, pretty standard stuff by now. The new chaining method in Sword and Shield is pretty much just straight up genocide. KO the same Pokemon consecutively to increase the chain and add more rolls. You can encounter different Pokemon, as long as you run away and don’t KO/catch them, and your chain won’t break. This one is a bit interesting, because you actually kind of have two counters working in tandem here. The maximum chain is 25, which increases your odds of getting extra rolls per encounter up to about 33%, but how many extra rolls you get is based on how many of that Pokemon you’ve ever battled. You only have a chance for one extra roll when you start, but after knocking out 500 of the same Pokemon at a chain of 25, you’ll start getting six extra rolls on 33% of your encounters. Shiny Charm still adds two extra rolls per encounter, meaning that on average, every third encounter will be rolled nine times. Shiny Charm is still obtained by completing the Regional Dex by the way. I was a bit worried you’d have to complete the National Dex to get it, but good ol’ Game Freak put my mind at ease and didn’t even put the National Dex in the game to begin with.

There’s one more thing to discuss with Sword and Shield, and that’s the different types of sparkles that shinies can have. Some have the normal star-shaped sparkles, while some have square-shaped sparkles. Initially, it was thought that squares were more rare, but that doesn’t seem to be the case. Pokemon hatched from eggs seem to almost always have stars, but Pokemon found through chaining seem to usually have squares. The best explanation I’ve heard for this is that you get stars when a shiny was found on the first roll, and squares when it was found on a reroll, and at some point in the last four generations, Game Freak must’ve switched Masuda Method from giving four extra rolls to just multiplying the base odds by five. Since eggs should then never get rerolled, they would always* yield stars, but since chaining can give out a lot of rolls, you’d be more likely to get squares. I have nothing to back this explanation up, but it’s the most plausible one I’ve heard

*eggs do very rarely get extra rolls, probably due to a glitch

And that’s how shiny hunting works in all the main series games up to Gen 8! I may or may not update this if/when another game comes out (I probably won’t), but at least now nobody can call me out for dropping shiny hunting essays in the Discord because this one is WAY longer and it’s on the main site, not the Discord.

I’m going to bed.

August 28, 2020 at 7:34pm
i have never shiny hunted in my life. i have no idea why i read this