• Check out the relaunch of our general collection, with classic designs and new ones by our very own Pissog!

Project Ladder Achievements Project 2 - Gen 5 Pokebilities - WEEK 2

For probabilistic context, Just checked the code and there's like 7.75 Gem users (Acro users / Hitmonlee, subtracting sets without them) out of 356 possible species (not counting alternate forms), that would be a 2.18% Gem rate. If my math is right, that means every team has a 12.37% to contain at least one gem user, or about 1/8 (which cannot possibly be right unless I got really unlucky too).


1760543506758.png


Sources:
https://github.com/smogon/pokemon-showdown/blob/master/data/random-battles/gen5/sets.json
https://github.com/smogon/pokemon-showdown/blob/master/data/random-battles/gen5/teams.ts (lines 609-610)

In case you were wondering, I like probability.
 

Attachments

  • 1760543347749.png
    1760543347749.png
    24.7 KB · Views: 6
Shynii21 said:
For probabilistic context, Just checked the code and there's like 7.75 Gem users (Acro users / Hitmonlee, subtracting sets without them) out of 356 possible species (not counting alternate forms), that would be a 2.18% Gem rate. If my math is right, that means every team has a 12.37% to contain at least one gem user, or about 1/8 (which cannot possibly be right unless I got really unlucky too).
Click to expand...
These five on the right total up to 1.75 gem mons, and the acro mons might not have guaranteed gem? All 6 of them would surprise me
Like who is it? ledian, vespiqueen, emolga, can't think of the rest honestly
 
Inazuma-Fulgur said:
These five on the right total up to 1.75 gem mons, and the acro mons might not have guaranteed gem? All 6 of them would surprise me
Like who is it? ledian, vespiqueen, emolga, can't think of the rest honestly
Click to expand...
I checked the code and it said that if the Mon runs Acro, then the item is always F Gem, which makes sense as it otherwise would have to run no item or another item, thus missing the full boost. The mons with guaranteed Acro and by extension Gem are Ledian, Drifblim, Archeops, Emolga, Jumpluff and Vespiquen.

Some mons like Charizard and Tornadus have multiple sets, some of them with no Acro, so we cannot count them the same as Drifblim or Archeops, which always run it.

The other condition is being Unburden Hitmonlee with Fake Out for Normal Gem or Fighting Gem otherwise. The other set is choiced Reckless.
 
As for the total number of species, I just took the whole amount from the JSON and individually looked for the ones with -'s (such as Wormadam-Trash and Arceus-Fire) to delete the alt forms (since Ho-Oh and P-Z just happen to have dashes without being alt forms). Iirc, every species has an equal chance of being genned, so those with multiple forms will have a proportional, lower chance for each form and so we cannot weight them the same.

Anyway, math.
 
True, but I ain't doing that. However, I did find a small mistake in my calcs.

I took every single chance of getting a Mon with no Gem as 97.82%, which should actually increase over time as your team fills up since you cannot repeat species. So it would start at 7.75/356, then 7.75/355, 7.75/354, and so on until 7.75/351 for the last mon. We should not decrease the numerator at all since we are calculating the chance that we get no Gem mons, so the amount of Gem mons in the pool remains constant. Then, we just take the product of them all and a more accurate result would be 12.45% (it's bigger, so yay?).

1760559737750.png


And I'm still not accounting for the way the team gen works with typings or the chance to gen the right species with non-Gem sets, but at that point we could just do a couple thousand simulations and call it a day.
 
I still got two monkeys, two weather abilities and zero gems

The gem odds just can't be right, monkey is much harder and I got more.
Like, genuinely thrice as many monkeys as gems across all wins and losses, at less than half the chance to get one
 
Shynii21 said:
True, but I ain't doing that. However, I did find a small
I took every single chance of getting a Mon with no Gem as 97.82%, which should actually increase over time as your team fills up since you cannot repeat species. So it would start at 7.75/356, then 7.75/355, 7.75/354, and so on until 7.75/351 for the last mon. We should not decrease the numerator at all since we are calculating the chance that we get no Gem mons, so the amount of Gem mons in the pool remains constant. Then, we just take the product of them all and a more accurate result would be 12.45% (it's bigger, so yay?).]
Click to expand...
Worse news, unless pokebilities uses a different pool/generation method there are actually close to 8 possible gem users
0.3447 (Charizard) + 0.5341 (tornadus) + 0.3163 sceptile + 0.3143 mienshao + (0.2343 + 0.2482) (Hitmonlee) = 1.9919

But to make up for it there are 389 possible pokemon, not 356

https://play.pkmn.cc/data/random/gen5randombattle.json
 
I think the 389 is important because set generation rolls odds for all forms until one has been selected, it's not doing genning deoxys and then selecting one of four forms.
It's just genning one deoxys and removes all from the pool

Still quite decent odds of gaining at least one gem user in a true random environment
0,11794183
 
Shynii21 said:
Sleeping Tight as well
https://replay.pokemonshowdown.com/gen5pokebilitiesrandombattle-2462569601 (turns 26, 29, 32)

Now it's only a matter of around a 1.69% chance to battle a team with Bibarel (not counting weakness-related team gen), or like 60 games on average. And I still gotta KO it and win. Sounds fun!
Click to expand...

When I wrote your pokemon needs to be put asleep, I meant that it would be put asleep by a sleep inducing move and I didn't think about Rest-Talk (that's on me), but I guess it can count
 
Rules don't state that monkeys need to be used in battle, I realized just after I got the last one but never on the field (my mistake)
https://replay.pokemonshowdown.com/...-2462713607-6w8ak9b06px2mghqa9la25nqevjexktpw

https://replay.pokemonshowdown.com/gen5pokebilitiesrandombattle-2460876165

https://replay.pokemonshowdown.com/...-2462369174-otcscx1shg7me564uinan3shydgg2hwpw

But I also didn't realize that you didn't need to win with the monkey on your team, so I'll look for another Simipour replay (Edit: nevermind forgot the simisage game had simipour, so I had all three for a while already)

First two gems though
 
Last edited:
35Q71N said:
When I wrote your pokemon needs to be put asleep, I meant that it would be put asleep by a sleep inducing move and I didn't think about Rest-Talk (that's on me), but I guess it can count
Click to expand...
I literally asked you about this when I saw this achievement....

"Be put asleep? No-*you cut me off but I finished the sentence while you were talking* rest?"
"yeah be put asleep" "no no" *in your funny Italian accent* and then another 10 minutes of you saying why you think these achievement won't be too difficooolt ensued
 
Inazuma-Fulgur said:
Rules don't state that monkeys need to be used in battle, I realized just after I got the last one but never on the field (my mistake)
https://replay.pokemonshowdown.com/...-2462713607-6w8ak9b06px2mghqa9la25nqevjexktpw
https://replay.pokemonshowdown.com/gen5pokebilitiesrandombattle-2460876165
https://replay.pokemonshowdown.com/...-2462369174-otcscx1shg7me564uinan3shydgg2hwpw

But I also didn't realize that you didn't need to win with the monkey on your team, so I'll look for another Simipour replay

First two gems though
Click to expand...

Ok no, this is a mistake I made but all the challenges that involves speicific pokemon required them to touch the field. My bad, I forgot to specify it for this cycle -_-
 
Shynii21 said:
Sleeping Tight as well
https://replay.pokemonshowdown.com/gen5pokebilitiesrandombattle-2462569601 (turns 26, 29, 32)

Now it's only a matter of around a 1.69% chance to battle a team with Bibarel (not counting weakness-related team gen), or like 60 games on average. And I still gotta KO it and win. Sounds fun!
Click to expand...

Ok nevermind, I guess I wasn't specific enough so I'll mention that your pokemon needs to be put asleep by other pokemon (being Magic Bounce'd counts though). If you can't get this achievement in the next days, I'll still count your replay but please try to find another replay :x
 
You must log in or register to reply here.
Back
Top