Open jung1411 opened 3 years ago
Can this be analyzed by us? or does this need input from users?
From the API but the role, lane data in the API are very inaccurate and there are 3 ways to resolve this according to riot api docs. https://riot-api-libraries.readthedocs.io/en/latest/roleid.html
Simple Mapping from docs like
{ (MID_LANE, SOLO): MIDDLE, (TOP_LANE, SOLO): TOP, (JUNGLE, NONE): JUNGLE, (BOT_LANE, DUO_CARRY): BOTTOM, (BOT_LANE, DUO_SUPPORT): UTILITY }
where (lane, role) = position
I found out that there are additional errors and I can try to look up more to make more mappings. One of the mapping I found was ("NONE", "DUO_SUPPORT") : "SUPPORT"
Playrates We can find out the champions played by the summoner and find where the champions are usually played (e.g. Lux plays mid 30% of the time and support 70%). This however conflicts with players who play champions off position like playing a mid champion in top lane.
Using timeline data We can find by getting where the players played based on the position of the map in the early stage of the games. Other factors may include summoner spells, runes and items etc. This method is very time consuming and may take awhile during runtime, which may decease the performance of the bot.
When doing make command, each team is divided according to its position. For example, one team should not have 3 junglers.