Does the RNG really even work?

General feedback and discussion of the game.

Moderators: Developers, Forum Moderators

Does the RNG really even work?

Postby Mithrus » May 21st, 2010, 8:54 pm

I have been playing Wesnoth on and off for the last year or so, and one thing REALLY bothers me: the RNG just stinks. I love playing the game, except when several turns in a row my unit with a 70% hit chance and 4 attacks misses EVERY time, while the AI has a 30% hit chance and doesn't miss once. If this happened just once or twice in a game, I'd just blame my bad luck, but this happens about once out of every 5-6 units I attack with. It doesn't matter which Era I use, the results are the same. Are there other invisible factors in determining hit?
Mithrus
 
Posts: 2
Joined: May 21st, 2010, 8:46 pm

Re: Does the RNG really even work?

Postby Dave » May 21st, 2010, 9:08 pm

Thank you for reporting this problem. Please post a replay of a game where you are experiencing this problem so that we can analyze and resolve it.
“At Gambling, the deadly sin is to mistake bad play for bad luck.” -- Ian Fleming
Dave
Founding Developer
 
Posts: 7055
Joined: August 17th, 2003, 5:07 am
Location: Seattle

Re: Does the RNG really even work?

Postby Maeglin Dubh » May 21st, 2010, 9:10 pm

This is an often-claimed but rarely demonstrated problem, which is why we'd like to see a replay of it occuring. If it's actually happening that way to you, in multiple games, there may actually be an error that can be corrected.
Cuyo Quiz wrote:I really should push for Temuchin's brainstorming with all my might someday, when the skies are cloudy, the winds dance and the light is free to roam over the soil along the fog.
User avatar
Maeglin Dubh
Moderator Emeritus
 
Posts: 1154
Joined: November 16th, 2005, 8:38 pm
Location: Valley of the Shadow of Death

Re: Does the RNG really even work?

Postby Christheturtle » May 21st, 2010, 9:30 pm

It's happened to me a few times actually, but it's something I thought nothing of really. It's not commonly occurring, so it's not that bad an issue. Just every blue moon I'll get a game and I'll barely hit at all. If It happens again, I'll remember to save a replay.
www.turtlenetworking.co.uk - 25% completed. Adding more servers. Looking for volunteers.
Christheturtle
 
Posts: 46
Joined: May 19th, 2010, 1:08 pm
Location: United Kingdom

Re: Does the RNG really even work?

Postby monochromatic » May 21st, 2010, 9:42 pm

:roll:

1, 2, 3, 4, 5, 6, 7, 8. You've got some readings to do.
Two things in common: 1) they're all about luck/abnormal strike sequences 2) they're all locked threads. Please, I beg you not to do the same. If you really want to prove that the code sucks, post a replay, or point out the part in the code.
monochromatic
Forum Regular
 
Posts: 1542
Joined: June 18th, 2009, 1:45 am

Re: Does the RNG really even work?

Postby Spell » May 21st, 2010, 10:26 pm

Maybe you should make some Testrow. You do onehundret strikes and look if about 70% of them hit. (or thousand)
User avatar
Spell
 
Posts: 48
Joined: January 23rd, 2009, 9:04 pm

Re: Does the RNG really even work?

Postby Christheturtle » May 21st, 2010, 10:41 pm

elvish_sovereign wrote::roll:

1, 2, 3, 4, 5, 6, 7, 8. You've got some readings to do.
Two things in common: 1) they're all about luck/abnormal strike sequences 2) they're all locked threads. Please, I beg you not to do the same. If you really want to prove that the code sucks, post a replay, or point out the part in the code.


I do understand the system, I'm a competent Java/PHP/Lua coder and I had a shot at C++ for about a year, I just havn't yet had a look at the Wesnoth code. I wasn't saying that there was a bug, or fault in the code, or slagging off the continuous efforts of the developers, I was simply saying that if somewhere down the line something is wrong, then I may have also noticed it. You're only going to find out by people bringing it up surely. :)

Anyway, regardless of the outcome it's a rare occurrence and it doesn't affect the game play too bad so I don't really care.
www.turtlenetworking.co.uk - 25% completed. Adding more servers. Looking for volunteers.
Christheturtle
 
Posts: 46
Joined: May 19th, 2010, 1:08 pm
Location: United Kingdom

Re: Does the RNG really even work?

Postby Ken_Oh » May 21st, 2010, 11:21 pm

See the link in my sig.
User avatar
Ken_Oh
Moderator Emeritus
 
Posts: 2176
Joined: February 6th, 2006, 4:03 am
Location: Baltimore, Maryland, USA

Re: Does the RNG really even work?

Postby boru » May 21st, 2010, 11:24 pm

Mithrus wrote:I have been playing Wesnoth on and off for the last year or so, and one thing REALLY bothers me: the RNG just stinks. I love playing the game, except when several turns in a row my unit with a 70% hit chance and 4 attacks misses EVERY time, while the AI has a 30% hit chance and doesn't miss once. If this happened just once or twice in a game, I'd just blame my bad luck, but this happens about once out of every 5-6 units I attack with. It doesn't matter which Era I use, the results are the same. Are there other invisible factors in determining hit?


This has been explained more than once. It's a perceptual problem, not a problem with the code.

ATTACK ONE: Perception: "I'll probably hit this guy .. it says I have 70% chance." Code: if you repeat this 100 times, you will miss 30 times.
ATTACK TWO: Perception: "Hey I missed! I'll probably hit him this time, 'cause I missed the first time." Code: if you repeat this 100 times, you will miss 30 times.
ATTACK THREE: Perception: "I missed again? Impossible!" Code: if you repeat this 100 times, you will miss 30 times.
ATTACK FOUR: Perception: "Something's wrong with the code. I had 70% four times! That's, um, a 280% chance of hitting, right?" Code: if you repeat this 100 times, you will miss 30 times.

Frustrating? Yes, sometimes. But the truth is, a 70% chance of hitting is not as great as it sounds.
“It is written in my life-blood, such as that is, thick or thin; and I can no other.” - J.R.R. Tolkien

My campaign: Swamplings - Four centuries before the founding of Wesnoth, the first wolf rider emerges from a tribe of lowly swamp goblins.
User avatar
boru
 
Posts: 788
Joined: November 19th, 2009, 11:02 pm

Re: Does the RNG really even work?

Postby monochromatic » May 21st, 2010, 11:25 pm

@Christheturtle sorry if you took it wrong. My post was directed at the OP.
monochromatic
Forum Regular
 
Posts: 1542
Joined: June 18th, 2009, 1:45 am

Re: Does the RNG really even work?

Postby PeterPorty » May 21st, 2010, 11:26 pm

Uh.... this has also happened to me... a lot, but it has also happened that I have 30% CtH and I hit with all my 4 attacks, while the enmy has 70% CtH and misses all of them...so... I don't complain.
"The real world is for people who can't imagine anything better."
User avatar
PeterPorty
Translator
 
Posts: 310
Joined: January 12th, 2010, 2:25 am
Location: Chair, In-Front-Of-Computer

Re: Does the RNG really even work?

Postby Christheturtle » May 21st, 2010, 11:26 pm

Ken_Oh wrote:See the link in my sig.


I'd actually already read that from your signature, I'd just like to say that it was a very informative post. :) I must stress again, I'm not saying that there actually is an error somewhere in the code... But as I said before the only way errors can be found is by people reporting them and I have to say that it's really not very pleasing to see people disparaging others for bringing them up. Yes, he's probably wrong but regardless isn't it better to have excessive information than a lack of? :)

Just saying.

EDIT: Posts made while typing, no worries Elvish, I didn't take it as that. :)
www.turtlenetworking.co.uk - 25% completed. Adding more servers. Looking for volunteers.
Christheturtle
 
Posts: 46
Joined: May 19th, 2010, 1:08 pm
Location: United Kingdom

Re: Does the RNG really even work?

Postby Pentarctagon » May 22nd, 2010, 1:28 am

ok, how about this: i just had two units (unit A and unit B) with 1-1 attack do 100,000 rounds of combat. they were both on 40% dodge. unit A hit unit B 60,011 times. unit B hit unit A 60,026 times.
99 little bugs in the code, 99 little bugs
take one down, patch it around
-2,147,483,648 little bugs in the code
User avatar
Pentarctagon
Moderator
 
Posts: 2669
Joined: March 22nd, 2009, 10:50 pm
Location: Earth (occasionally)

Re: Does the RNG really even work?

Postby Sapient » May 22nd, 2010, 4:33 am

Until the original poster (or someone supporting him) provides his replay, or a mathematical proof of flaws in the code, I don't see any reason to continue this discussion. If on the other hand this is a false bug report filed in the User forum, as it certainly appears to be, then I guess we have a right to be annoyed. :annoyed:

Still yet, maybe I should just try to see the humor in the situation like Dave, and say "Thank you for the valuable Shrug Report!"
http://www.wesnoth.org/wiki/User:Sapient... "Looks like your skills saved us again. Uh, well at least, they saved Soarin's apple pie."
User avatar
Sapient
Developer
 
Posts: 4307
Joined: November 26th, 2005, 7:41 am

Re: Does the RNG really even work?

Postby Dave » May 22nd, 2010, 5:38 am

Christheturtle wrote:as I said before the only way errors can be found is by people reporting them and I have to say that it's really not very pleasing to see people disparaging others for bringing them up. Yes, he's probably wrong but regardless isn't it better to have excessive information than a lack of? :)


Well the problem is that this is something that we've gone over and over and over many times, and it's wasting a lot of developer's valuable time and clogs the forums.

If anyone has evidence of the RNG being broken we will be happy to look at the evidence, but there will be no evidence, because the RNG is not broken.

David
“At Gambling, the deadly sin is to mistake bad play for bad luck.” -- Ian Fleming
Dave
Founding Developer
 
Posts: 7055
Joined: August 17th, 2003, 5:07 am
Location: Seattle

Next

Return to Users’ Forum

Who is online

Users browsing this forum: No registered users and 1 guest