Bulbapedia only uses common, rare, ect. Everything else on the official Routes/Named Routes always equals 100. The only time they use percentages is when it's on their Hidden Grotto page.
Also doesn't account for when you get no items, so it doesn't really make sense. This is also why their Rock Smash encounters do not include percentages for items/nothing, and instead equals 100. Xbser (talk) 05:08, 26 January 2015 (CET)
Perhaps the "Nothing" was the glitch, haven't tested it just talking to people on IRC, where just the sprites weren't appearing, nonetheless this is how bulbapedia does it. Xbser (talk) 05:08, 26 January 2015 (CET)
I actually can't understand why bulbapedia didn't simply exclude the chance of not finding anything when doing the grotto listings. Like, there's a chance that the trainer doesn't find anything whenever they walk in a cave, for example, but that's never taken into account when listing normal encounter percentages. It seems sloppy to simply list them without exact percentages, but then again I guess it's less confusing than having to consider what percentage of the total is comprised by wild items/Pokémon. I honestly don't know if it's possible to get Nothing from the grottoes in Insurgence, but regardless it just seems like the same principle of overall encounter rate when walking through the grass.
Their Rock Smash example confuses me because I can't find any source of information that indicates that the item and Pokémon percentages are dependent on each other (i.e. if it's possible to have a wild Pokémon encounter and also find an item beneath the rock with 1 Rock Smash). If they are dependent then having them use separate methods for listing grotto encounter and Rock Smash encounter percentages doesn't make sense to me. I'd personally prefer to see the uncommon/common, etc. listing used for Rock Smash here as well because it's effectively using the same system.
The encounters system is this by the way, http://pokemonessentials.wikia.com/wiki/Encounters
Except, for Rock Smash which Suze is using his own code, which was leaked and I used that as a base to calculate the percentages.
It works like:
It picks a random number 0-2, if it's not 0 it returns nothing. Else it does another random which picks another 0-2.
If it's not 0, it picks from the Pokémon list which uses a list of 12 Pokémon. Which had Geodude 4 times, Roggenrola 3 times, Nosepass 2 times, Carbink 1 time, Shuckle 1 time, and Dwebble 1 time. Else it picks from a item lists that has 28 items.
So it's 66.6% of nothing, 22.2% of a Pokémon, and 11.1% for an item.
What's strange about that is that I manually tested encounters for Rock Smash and found that after 2022 attempts, the number of Pokémon/items combined comprised 1005 "hits" whereas the number of times where nothing was found comprised 1017 "misses", which would seem to indicate that the chance of finding nothing is approximately 50%. I'd pass that off as too low of a sample size if the sample size were 200 or so, but with ~2000 I find that a bit hard to believe especially considering that the numbers were so even throughout the entire trial. Another oddity is that my list has 15 items, not 28, unless the item list in the code was supposed to have multiple entries for the same item? Another fun fact is that 501 of the "hits" were Pokémon encounters while the other 504 were items, which makes it seem that the chances of finding a Pokémon or item are also the same.
However, ignoring that for the moment, I still believe that it would be more ideal to have the Rock Smash and Hidden Grotto lists use the same system when listing the chances of finding items/Pokémon because even if the code works differently with Rock Smash versus Hidden Grotto encounters, the reality is that the two feature items or Pokémon that have a certain chance of appearing. I personally don't mind if the consensus is to use the Common, Uncommon, etc. listing or listing the Pokémon in terms of percentages (i.e. HG/SS Rock Smash listing) so long as they're consistent and logical. (1ofthe4rocketbros (talk) 06:34, 26 January 2015 (CET))
- I was discussing what the code meant with Xbser last night, and now that I look at it again I believe I made a mistake and you're in fact correct. Each time a random number is called it uses the function rand(2), which I now remember picks a random number from 0 and 1, rather than 0, 1 and 2 as I had assumed earlier. So it is indeed 50% each time. Berlyda (talk) 11:14, 26 January 2015 (CET)
- Yeah that makes a lot more sense. I should have noticed this yesterday when looking at the level generated for Pokémon encounters considering that the levels ranged from 39 to 43 when using a level 42 Pokémon, which would mean that level - rand(5) + 1 would have to produce a maximum of 4 in rand(5). In retrospect I probably shouldn't examine code at 2 AM.
Well I can send you the code through reddit or someplace else, and yeah the 28 items had duplicate items like the encounter list. Or you could ask Suze if that's how the Rock Smash system works. In the end you can do the lists however you want, it doesn't really matter to me I just want it to no longer say ? or Unknown for the pages. I also was just following how bulbapedia is. Xbser (talk) 06:52, 26 January 2015 (CET)
Yeah, go ahead and send it to me on reddit, I'd be very interested in taking a look at it; don't want to distract suze from whatever it is he's doing for this just yet. Also, thanks for that Essentials Encounters link; I had no idea what to make of those lines of text before. (1ofthe4rocketbros (talk) 07:10, 26 January 2015 (CET))