Skip to content

dokjahyung/Naraka-AI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Naraka-AI

Purpose

Naraka Bladepoint is a game that is not big in th western world but has become a central hub for competitive gamers alike most notably in China and Vietnam. boasting a 40 million player base and quite possible been called "the most difficult game ever". Pro play has become a spectacle for many to gaze in awe. Many are trying to build their own strategies for winning competitive events and be able to say they won the "Herald's Trial". Because of this many veteran players have tried to test and rationalize developing metas. The purpose of this code is to be both an informative experience for basic low level linear regressions and also possibly providing an edge to driven players in deducing the most important objectives in competitive play. Is it fighting for the 3v3 Yang fights? winning off spawn? Or is it simply ratting? Hopefully this code will offer some assistance in users testing various versions and combinations of stats to find a good fit for them

How to install

preparations

 Thanks to pyinstaller you only need to download the 'dist' folder

Step 1

go to the 'dist/app/app.exe'

Step 2

run app exe and enjoy

Features

VERSION 1.0.1 is out. This version shows the basic backbone of project however lacks in the much needed data from website. Will either need to create a better webscraping program and/or compile a better list of other stat fields as well i.e yang participations, yang wins, off spawn wins, etc.

This only uses a least squares model built with fewer dependencies than basic AI models out there like scikit however may be lacking in performance as well escpecially when data is either outside specified ranges or in a "unique" arrangement of values.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published