当前位置:网站首页>Why is it that all the games are pseudorandom and can't make true random?
Why is it that all the games are pseudorandom and can't make true random?
2022-07-25 05:40:00 【Dafang teacher talks about SCM】
Why are all in the game pseudorandom , Can't make it really random ? I hope to have a simple explanation ?
I found it like LOL These games , The crit mechanism is pseudorandom , There are no critical hits in front , The critical hit probability behind is very high . The glory of Kings is more obvious , Can card critical hit .
Because this can ensure that the overall probability is completely unchanged ,「 Extreme good luck 」 and 「 Extreme bad luck 」 As few as possible , Reduce the luck of each player , Ensure the experience of both sides .
“ Robin Hood ” Of 「 Pseudorandom 」 The European emperor can speak more about the basic law , Let Fei Chieh have the most basic game experience .
If you play the game according to the real probability , The experience will only be worse , That's really disgusting. Remember one year at a time .( The answer is only for the analysis of the same probability value , It does not include the shipment of card drawing mobile games …)
for instance , To compare 20% Probability , Distinguish use 「 True random 」 and 「 Pseudorandom 」 Under the circumstances , yes 2 There are different calculation modes .
It's really random 20% operation pattern : The probability of each occurrence is 20%, Calculate independently each time .
Pseudo random 20% operation pattern : The first 1 The probability of occurrence is 5.6%, If it happens, press... Next time 1 Next 5.6% count , If it doesn't happen, the next probability +5.6%, For example 2 for 11.2%, If you don't hit it again 3 for 16.8%, And so on , Until it happens and returns to the 1 Next 5.6%. If the cumulative probability exceeds 100%, It is considered that it must happen next time .
5.6% This number can be calculated by computer , It is enough to ensure that the probability of occurrence of the two modes is the same , Let's not go over this . Each probability has a numerical solution corresponding to the above values ( Use the computer to calculate directly ok 了 , The simplest dichotomy works ).

Let's see 20% Probability of two modes 「 Good luck 」 and 「 Bad luck 」 The distribution of .
Let's start with extreme luck :
20% Probability good luck situation
continuity 2 Trigger
continuity 3 Trigger
continuity 4 Trigger
True random
4.0%
0.80%
0.16%
Pseudorandom
0.31%
0.018%
0.00098%
You can find , In pseudo-random mode , The possibility of extreme luck is greatly reduced .
Let's look at the probability of relatively bad luck :
20% Probability of bad luck
continuity 6 Times do not trigger
continuity 12 Times do not trigger
continuity 18 Times do not trigger
True random
26.2%
6.87%
1.80%
Pseudorandom
25.9%
0.239%
It's completely impossible
Ruodi 6 It hasn't happened since next ,「 Pseudorandom 」 Will greatly improve the probability of occurrence , Ensure that players trigger once as soon as possible ; And the first 18 The guaranteed minimum guarantee completely guarantees that it is no longer possible 18 Don't hit the bottom .
「 True random 」 Is to continue to be independent . alike 20% probability , Yes 0.8% May trigger continuously 3 Time , And then there is 1.8% The probability will be continuous 18 It doesn't trigger twice …
in other words 20% Under the premise of extreme bad luck , If you try to 100 Round 「 Continuous attack 18 Time 」, There are mathematical expectations 1.8 A round will happen without triggering , This probability is really not very low …
You can find 「 Pseudorandom 」 Under the premise of preserving randomness , Than 「 True random 」 Ensure the game experience of both sides .
Use true random , Many game experiences will only be worse .
For example, for 《 tetris 》, If you use really random , As long as you play enough , You will definitely encounter continuous 30 It's impossible to stand up , Sometimes I encounter 10 Inside out 6 A vertical bar … Which one is very painful .
And the war games are even more , such as 《DOTA2》 Inside PA This kind of hero with critical hit , Critical chance 15% If you use really random , The psychological expectation of players may be 6-7 One critical hit with the knife , But actually continuous 32 There is also a probability that the knife will not critically hit 0.55%… Not to mention the hero of vanity , If you use really random , There will certainly be many more 「 The other party hits you with a hammer in the void , You are directly full of blood and killed by Mu Da 」 The situation of …
Any one of the above may be enough to make people sick enough to delete the game , Those with poor quality may want to connect with his mother, the game producer (x
Just a few simple examples , It may not be so intuitive , It also involves 「 Random 」 And people's psychological misunderstanding is 「 Average 」 wait , It's not going to unfold here .
边栏推荐
- Summary of common attributes of flex layout
- C编程 --“最大子数组的和” 的动态规划的解法
- QT qtextedit setting qscrollbar style sheet does not take effect solution
- Array programming problem of CSDN programming challenge
- sqlilabs less-28~less-8a
- Leetcode 204. count prime numbers (wonderful)
- LCP插件创建对等802.1ad接口
- LeetCode第302场周赛
- Necessary skills for mobile terminal test: ADB command and packet capturing
- Leetcode 237. 删除链表中的节点
猜你喜欢

同条网线电脑正常上网,手机连接wifi成功,但是无法访问互联网

微服务 - 网关Gateway组件

Idea commonly used 10 shortcut keys

CCID released the "Lake warehouse integrated technology research report", and Jushan database was selected as a typical representative of domestic enterprises

QT qtextedit setting qscrollbar style sheet does not take effect solution

Basset: learning the regulatory code of the accessible genome with deep convolutional neural network

C100: smallest hevc visual IOT MCU

What are the ways to realize web digital visualization?

HTB-Optimum

Leetcode 15: sum of three numbers
随机推荐
Differences and application directions of GPS, base station and IP positioning
Leetcode 0122. the best time to buy and sell stocks II
npx和npm区别
LeetCode第302场周赛
Obj file format and.Mtl file format
easyrecovery免费数据恢复工具操作简单一键恢复数据
How to carry out the function test with simple appearance and complex interior?
The u-collapse component of uniapp mobile uview is highly init
The price is 17300! Why does Huawei mate x face Samsung fold?
Programming hodgepodge (I)
ZTE's first 5g mobile phone, axon 10 pro, was launched in the first half of this year
HTB-Arctic
JWT(json web token)
Sword finger offer 05. replace spaces
VIM search and replacement and the use of regular expressions
PHP warehouse inventory management system source code WMS source code
Unity接入ChartAndGraph图表插件
Please stop using system The currenttimemillis() statistical code is time-consuming, which is really too low!
OpenFegin远程调用丢失请求头问题
50 places are limited to open | with the news of oceanbase's annual press conference coming!