Is this a good way to fudge percentages of things happening?

Frostorm

[]D[][]V[][]D aka "Staf00"
Veteran
Joined
Feb 22, 2016
Messages
1,116
Reaction score
807
First Language
English
Primarily Uses
RMMV
Apologies for straying off-topic earlier. Thanks, @Wavelength for bringing us back on track!:kaophew:
The method used in LoL seems pretty sound. If it weren't for the fact that it's a PvP game, there would be no qualms about it. Am I correct to assume that in that system, if a player gets a lucky string of crits, then the following attacks will have a reduced chance to crit?
 

Wavelength

Edge of Eternity
Global Mod
Joined
Jul 22, 2014
Messages
5,404
Reaction score
4,799
First Language
English
Primarily Uses
RMVXA
The method used in LoL seems pretty sound. If it weren't for the fact that it's a PvP game, there would be no qualms about it. Am I correct to assume that in that system, if a player gets a lucky string of crits, then the following attacks will have a reduced chance to crit?
I believe that is correct - if you've gotten a lucky string of crits lately your next couple attacks will have reduced chance to crit (versus your actual Crit stat) until you fail to crit once or twice (or you spend an amount of time out of combat, at which point I think the Crit Bias resets).

No one seems to know the exact way that they adjust it, but here are a few things I found while searching to suggest this is correct:
LeagueWiki said:
  • Critical strike chance will adjust dynamically based on the number of basic attack that didn't critically strike. At 30% critical strike chance you're sure to have 30 or more attacks critical strike per 100 attacks. If you didn't critical strike for a long time your next attacks will have a higher chance.
  • When averaged over a large number of autoattacks and at base critical damage (100% bonus damage), critical strike chance is essentially a damage multiplier where each 1% of critical chance adds 1% bonus damage.
Reddit Post said:
it's not linearly stacking but logarithmically, nobody knows the exact algorithm but it probably averages the amount of crit you got in your last few autos and if it's higher than your crit chance it will lower your actual crit chance and if it's lower it will increase it.
jjcm.org said:
PRNG stands for Pseudo-Random Number Generation. Essentially it's a method of random number generation that more evenly distributes the chances of success. In the context of DOTA or LEAGUE, it means that there's a far smaller chance of being "streaky" (having a ton of crits in a row, or none at all for a long period of time). If you have the time, take a look at the gamepedia wiki entry on PRNG (beware its dota2 gamepedia XD). The basics though work like this: PA ADC's actual chance to crit on her YOUR first attack is 3.222% (not 15% if you happen to get ghostblade as your first item Kappa). However, each subsequent attack she you do after that increases her your chance to crit by another 3.222%. So her your 2nd attack will have a 6.444% chance to crit, her 3rd attack will have a 9.666% chance to crit, and so on until the crit procs. Once the crit procs, the percent is reset back to 3.222%. This averages out to her you critting 15% of the time over many hits, but each individual attack will have a far different chance to crit on its own. The quick ones of you will notice that this system can be gamed. If you count your attacks since your last crit, you can compute the probability of a crit on your next attack. The returns aren't that great though - a 3% increase each attack isnt much.
Essentially I think this kind of system is fair, especially since fights move so fast that you would have to have the brain of a supercomputer to game it, and it's probably better than letting skirmishes get decided by a lucky/unlucky string of crits (or non-crits) rather than player skill. Since three or four basic attacks (which can come out in less than two seconds) are often enough to kill an opponent late-game, and crit chances often hover around 50% for most Attack Damage Carries, those lucky/unlucky strings of crits are a real possibility and that's probably why they felt the need to bias the RNG toward moderate results.
 

Frostorm

[]D[][]V[][]D aka "Staf00"
Veteran
Joined
Feb 22, 2016
Messages
1,116
Reaction score
807
First Language
English
Primarily Uses
RMMV
What are your thoughts on Pseudo-RNG vs Multi-Roll RNG? I guess technically they're not mutually exclusive and could both be employed if the dev really wanted to...
 

Sword_of_Dusk

Keyblade Master
Veteran
Joined
Sep 13, 2015
Messages
177
Reaction score
199
First Language
English
Primarily Uses
RMMV
Pokemon is also an offline game that has many hidden mechanics. However, players still end up going on Bulbapedia to look that stuff up. There will always be a subset of the population that will try to maximize their EV/IV/Natures/etc... Sure their design choice in not telling the player was probably the right one in the early days when it came out since its players were generally kids. But all those kids playing Pokemon in their youth are now adults w/ minds capable of comprehending complex mathematical formulae. Hell, it doesn't even need to be an in your face way of letting the player know of these mechanics. They could just include an NPC, lets use Blue (Gary who ended up becoming a Pokemon Prof in canon) for example and place him somewhere in the game. When you talk to him, he could explain all the hidden mechanics of the game, since he's been there, done that, aka a true OG Pokemon veteran. Plus he's a retired Pokemon Master and now a Pokemon professor/researcher so he should be expected to know stuff like that.

I'm aware of its common practice, but lots of practices that are common aren't always the best. Just take Tai_MT's rants on the flaws of 99% the RPGs he's played as an example.
I have to point out that people only care about the hidden stats in Pokémon because of competitive battles. As in, battles against real players. Far as in game versus the A.I., no one cares.

Aside from that, I can assure you that all of the players who care already know about the various mechanics. The younger crew who don't care do not need all the extra fluff told to them.
 

Frostorm

[]D[][]V[][]D aka "Staf00"
Veteran
Joined
Feb 22, 2016
Messages
1,116
Reaction score
807
First Language
English
Primarily Uses
RMMV
I guess I'm a weirdo then, cuz I like having "perfect" Pokemon just to have. I don't participate in PvP in Pokemon at all, but it irks me to no end if I have a sub-optimal Pokemon in my party lol.
 

Sword_of_Dusk

Keyblade Master
Veteran
Joined
Sep 13, 2015
Messages
177
Reaction score
199
First Language
English
Primarily Uses
RMMV
I wouldn't say you're a weirdo.

For me though, doing that would dilute my story experience. Corny as it may sound, I bond with the Pokémon I use for the adventure, and part of that comes from the act of catching them and rotating them into the team. If I went for perfect stats, I'd have to specifically breed my mons, then EV train them correctly, and that would just bring the whole game to a halt.
 

theartofme

Villager
Member
Joined
Feb 21, 2019
Messages
14
Reaction score
11
First Language
English
Primarily Uses
RMMV
Nit: although PRNG does stand for pseudorandom number generator, it does not mean the system described above for League of Legends. PRNG refers to any one of many methods for generating a number that appears random, although it isn't actually random (hence "pseudo").

Good ones have a good spread across all known numbers (coverage), do not favour any particular numbers (bias), and do not have any obvious patterns.

All random numbers in RPG Maker come from a PRNG; in general the vast majority of random numbers on a computer come from a PRNG because true randomness is expensive.
 

Frostorm

[]D[][]V[][]D aka "Staf00"
Veteran
Joined
Feb 22, 2016
Messages
1,116
Reaction score
807
First Language
English
Primarily Uses
RMMV
I dare say true randomness only exists on the quantum level. Even rolling a dice IRL is really just a complex system of physical interactions that our minds can't comprehend, thus we summarize it as "random" lol.
 

gstv87

Veteran
Veteran
Joined
Oct 20, 2015
Messages
2,184
Reaction score
1,162
First Language
Spanish
Primarily Uses
RMVXA
So i was wondering how one can tweak the actuall percentages so it meets player expectation
by building a good battle system.
because, often the percentage shown in the description is the chance to *land*, but not the chance to *effect*, because there's a resistance on the other side that the player isn't seeing.
 

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Latest Threads

Latest Posts

Latest Profile Posts

Yanfly told me the new earth all animations in MZ look a bit like summoning potatoes and now I can't unsee it. halp
Despite time and all the other things I am working on, I decided that I am going to start a Fire Emblem-like project. Follow your dreams! I'll just have to balance hours efficiently. I'm also not making this as complex as my Might and Magic inspired project, so it will be a much easier project to work on.
So I've taken to doing a test of the game, and making a list of everything I think of during the test to add/fix. Over the course of the night I check them off as I do them, and by the end of the night I always feel like I've accomplished something. :LZSexcite:
Eventz MZ? Events MZ!

Forum statistics

Threads
100,586
Messages
977,521
Members
132,181
Latest member
Reich
Top