2
edits
Talk:Catch rate: Difference between revisions
→Are we sure the shake probabilities for Gen III-VI are correct?: new section
Dragonfree (talk | contribs) |
|||
| Line 468: | Line 468: | ||
::::So, in my example, for a would equal 2.5=(2+2048/4096), or if using the current catch rate, a=3*1.5/3=1.5=(1+2048/4096)[[User:Zombiedude347|Zombiedude347]] ([[User talk:Zombiedude347|talk]]) 15:46, 29 December 2014 (UTC) | ::::So, in my example, for a would equal 2.5=(2+2048/4096), or if using the current catch rate, a=3*1.5/3=1.5=(1+2048/4096)[[User:Zombiedude347|Zombiedude347]] ([[User talk:Zombiedude347|talk]]) 15:46, 29 December 2014 (UTC) | ||
:::::Yes, that's the gist of it. [[User:Dragonfree|Dragonfree]] ([[User talk:Dragonfree|talk]]) 19:06, 29 December 2014 (UTC) | :::::Yes, that's the gist of it. [[User:Dragonfree|Dragonfree]] ([[User talk:Dragonfree|talk]]) 19:06, 29 December 2014 (UTC) | ||
== Are we sure the shake probabilities for Gen III-VI are correct? == | |||
And I don't mean just numerically correct--I ask because, the formula written down for Gen III-IV is: | |||
'''''b'' = 1048560 / sqrt(sqrt(16711680/''a''))'''; | |||
which is weird, since it uses numbers that require 3-bytes to represent. It is made even weirder when you consider that the formula there is ''exactly'' equivalent to: | |||
'''''b'' = 65535 / sqrt(sqrt(255/''a''))'''; | |||
which takes numbers of no more than 2-bytes, which makes a lot more sense, and unless the former was ripped straight from the code, seems closer to what the function actually would be (since it takes up less memory and uses numbers on a similar scale found elsewhere in the mechanics). | |||
This also makes me want to ask if the shake probabilities on Gen V and Gen VI don't have a small typo, since both have similar formulas to the above (though Gen VI has a denominator of (255/''a'')^0.1875), but instead of 65535, they have 65536, which is 1 off, but is 17-bit to 65535's 16-bits and that would be...inelegant. It would break the symmetry, since if it's 65535, then Gen III-V would all have the same shake probability function, and it seems a little silly to tweak it by 1 in a way that would require an extra byte of memory for the input, and right before a more substantial tweak in the very next gen. | |||
But I don't know--for all I know, that could be faithfully ripped straight from the source code. | |||
[[User:Xovvo|Xovvo]] ([[User talk:Xovvo|talk]]) 20:46, 23 February 2015 (UTC) | |||