teamsGenerator

This repo is about genberator of teams in soccer (or other sports) with a multiple open source algorithms.

Algorithms Explanation

Back And Forth Algorithm:

Set randomly an order of the teams. The first team pick a player, the second team pick the next player, the third team pick two players. After this round, it goes back. After the third team pick two players, then the second team pick a player, and then the first team pick two players. and so on. The team picks a player according to the rank order. Here is an example of the algorithm selection:

Skill Wise Algorithm:

I believe that variety of teams should be picked by uniform distrubtion according to skills. Here we define player by 5 skills: Leadership (field domination, increase the moral), Attack (finishing, key-pass), Defence (close to the goalkeeper, physics, aggression), Stamina (fit, pace, speed) and Passing (passing accuracy, playmaker).

The teams are ordered randomly, and then the algo picks randomly one of the skills above. Now we pick top 3 players with the highest rank in the specific skill, and spread them in each team.

Note: While picking 3 players ordered by skill rank, it goes like this:

1. We pick the highest skill rank.
2. If there are at least two players with the same skill rank, we pick the one with the higher avarage rank (avarage rank means avarage of all other skills rank)
3. If there are at least two players still, we pick randomly.

After this step. we have 3 teams, and each team contain one player according to the random skill. Now we pick a new skill (of the rest skills), and put the players in the same way. But now, the order of the picking teams is not random, but ordered by the team total rank (calculated by the summary of the players rank in the team).

Development

Add your own algorithm

First, think of an algorithm which makes sense and could be helpful and disribute it!

In Algos folder you could find AlgoType enum which declares the existing algo types. Add you new enum value that suitable to your algorithm name. Now, Create your own folder for your algorithm with the name of your algo with a postfix of Algo. For instace, if you create an algorithm with the name TheBestAlgoEver you should add enum with the name TheBestAlgoEver and folder with the name TheBestAlgoEverAlgo.

Here are 3 interfaces you should implement:

1. IAlgoManager is the one who implements teams generation.
2. IPlayerReader is the one who implements the players reading.
3. IPlayer is an interface which reflects the player model.

IPlayerReader receives the path string as parameter and returns collection of IPlayer with the relevant data (according to the implementation).

In case your algorithm engine works with local files, you should create a file with the name players with any extension you desire. The file path is provided to your IPlayersReader implementor and reading the players as you wish.

Add a CLI command

Under CLI folder you can find all of the commands and the printer classes. Creation of a new command for CLI requires these steps:

1. Create an implementation of IPrinterOptionCallback.
2. Fill the Description and DoCommand.
3. Add your command to the optionsCallback dictionary in Program.cs file.

Run

Endpoint (Standalone)

This runner is an endpoint application which runs on Windows machine only. In order to run the user should maintain a config.json file. The file contains an array of shirtColors which is an array of objects based on name which is the color name, and whatsappSymbol which is the color symbol that integrated with whastapp app.

After this file is created, the user should edit the player file which is stored in the algo folder.

Then all the user need is to run the app, choose the relevant algorithm and receive its teams.

Web Service

In order that any one at the world could use your algorithm, you should maintain the web service. In WebAppAPI we should add our algorithm to the both dictionary (follow the existing items), and in addition we need to add a new AlgoType enum.