当前位置:网站首页>How can Huawei online match improve the success rate of player matching
How can Huawei online match improve the success rate of player matching
2022-07-05 21:39:00 【Huawei Developer Forum】
background
When we match players , It is inevitable that the number of online players of the game is small or the difference of player attributes is large, which does not meet the matching conditions and other reasons , As a result, players cannot match the right opponent . At this time, how can we increase the success rate of player matching ?
Solution
Method 1 : Optimize matching rules
When configuring matching rules for Huawei online battle service , Allow configuration of relaxation properties . What does that mean ? When online players cannot be matched within the attribute range of a matching rule , At this time, online players with large attribute differences can match , You can configure the attribute relaxation interval and attribute relaxation duration . When the matching duration exceeds the relaxed condition , That is, trigger the relaxation of attribute interval , That is, players can match in a larger attribute range , To increase the probability of successful matching . This way can better take into account the fairness of players and the playability of the game . 
Method 2 : Robot filling
If the above method , The effect cannot be achieved , When players match, there is still a high probability that they cannot match online players , Consider opening the filling function . Opening method : When configuring matching rules , take “ Robot filling ” Options , Select on . After opening the service , When the player matches the overall timeout , Still can't match the real player , Then the system will calculate according to the player information , Return to the robot player who meets the player's ability range , Robot player information includes all player attribute information configured in the matching rules , Developers can construct robot capabilities according to attribute values .
Function opening is very simple , What matters is how the game is used .
Above screenshot , After opening robot filling , Players can always match successfully , Developers need to return according to the successful matching room Instance to get the list of players in the room , Analyze players isRobot attribute , If the value is 1 Indicates that the player is a robot player , If the value is 0, Expressed as a real player . The developer determines whether to trigger the game's local robot logic according to the above attributes .
How to synchronize robot operations between different clients ?
Huawei online battle service provides pseudo-random number generator , Developers can generate random numbers at different ends according to this method . Reference documents :https://developer.huawei.com/consumer/cn/doc/development/AppGallery-connect-Guides/gameobe-randomnumber-js-0000001250978031
边栏推荐
- Teach yourself to train pytorch model to Caffe (2)
- Emotional analysis of wechat chat records on Valentine's day based on Text Mining
- 第05章_存储引擎
- 从零开始实现lmax-Disruptor队列(四)多线程生产者MultiProducerSequencer原理解析
- Li Kou ----- the maximum profit of operating Ferris wheel
- 【日常训练】729. 我的日程安排表 I
- Robot operation mechanism
- SQL knowledge leak detection
- Robot framework setting variables
- 校招期间 准备面试算法岗位 该怎么做?
猜你喜欢

How to prepare for the algorithm interview and answer the algorithm interview questions

What should I do to prepare for the interview algorithm position during school recruitment?

面试官:并发编程实战会吗?(线程控制操作详解)

leetcode:1755. Sum of subsequences closest to the target value

Why can't Chinese software companies produce products? Abandon the Internet after 00; Open source high-performance API gateway component of station B | weekly email exclusive to VIP members of Menon w

Uni app Bluetooth communication

2.2 basic grammar of R language

使用Aspect制作全局异常处理类

Zhang Lijun: penetrating uncertainty depends on four "invariants"

Teach yourself to train pytorch model to Caffe (2)
随机推荐
XML modeling
MySQL deep paging optimization with tens of millions of data, and online failure is rejected!
冯唐“春风十里不如你”数字藏品,7月8日登录希壤!
Add ICO icon to clion MinGW compiled EXE file
Making global exception handling classes with aspect
Teach yourself to train pytorch model to Caffe (III)
Oracle HugePages没有被使用导致服务器很卡的解决方法
Kingbasees v8r3 data security case - audit record clearing case
深信服X计划-网络协议基础 DNS
Golang(1)|从环境准备到快速上手
How to send samples when applying for BS 476-7 display? Is it the same as the display??
vant 源码解析 event.ts 事件处理 全局函数 addEventListener详解
ArcGIS\QGIS无插件加载(无偏移)MapBox高清影像图
vant 源码解析 之深层 合并对象 深拷贝
leetcode:1139. The largest square bounded by 1
Selenium gets the verification code image in DOM
Modifiers of attributes of TS public, private, protect
递归查询多级菜单数据
vant 源码解析之 utils/index.ts 工具函数
Teach yourself to train pytorch model to Caffe (I)