Ahoy Folks, I could use your help Datamining LP slants

Kai Guy·Oct 21, 2019·3 votes·3,893 views

Look up the Match average of a game after you win or lose. Compare your LP gains to that average.

Example, If your G3 and the Match average was G1. What LP did you get for the win or loss?

I want to see if the Slant has a predictable behavior. If your playing 1-2 divisions different then your Rank I think the slant might look like 20 LP and if your at 2-3 I think it might be closer to 30 lp.

This could be helpful in trying to figure out what folks MMR values if there is a trend.

6 Comments

VerxintOct 21, 20191 votes

This is actually a really good idea if we could get enough data. Maybe make a Google docs people can input into?

Kai GuyOct 21, 20192 votes

Don't know if I trust the community. Probably run it as a smaller group of folks who are curious and use data collected with screen caps from strangers

VerxintOct 21, 20191 votes

Good point. Could still have a separate open data set as well as the tightly controlled one.

Kai GuyOct 21, 20191 votes

Im gonna head off to work but i assume a prettier version of something like this? is what you had in mind?

VerxintOct 21, 20191 votes

Pretty much, but probably want the persons winrate prior to the match.

If I was good at programming the information is probably all in the API but alas I am not.

Kai GuyOct 22, 20191 votes

If I had riots API I really wanna do champion data breakdowns. Like Wr by # games cuz having 98% of a champs tracked data be from folks who play less then 10 matches on them provides skewed #s.