Boards›Gameplay

ARAM picks, a statistical studying

Sukrat··2 votes·841 views

Hello there, some of you maybe have wondered how often each champion appears in a serie of "several random" games. Currently i considered my 39 last games as I don't have much time to spend on lol... Of course I intend to increment it.

I'be begun to complete an excel file containing so far :

  • How many times each champion was played ( goes from 0 times to 7=> Ziggs for instance)
  • the apparition frequency distribution compared to a normal one using the same parameters (mean + sigma).

It's aim is to help me to ask several questions :

  • How does the frequency distribution look like at N > 100 with a mean >> 0 ? Can it be considered as "normal" => fair
  • How could the random selection system be improved (even chance for each champ), even if we can pass more or less through thanks our dices etc...

Eventually, I would like to know what you think about it + any ideas/remarks ? corrections in particular ? I would actually need help with the probability computations when you use a dice ^^ (I'm a dumb physicist).

5 Comments

Thorn Fierce Red2 votes

Aram is busted

Sukrat1 votes

Can you be more specific in terms of champions maybe ?

Thorn Fierce Red1 votes

Double Stun Q Lux with low cdr on his R and abilities and Zyra low cdr annoying passive plant everywhere.

Baka Red1 votes

Your research sounds interesting, but please remember these couple of factors while you do it:

  • You do not know the champion pool structure of the other players.
  • Rerolls change the team compositions towards "better" ARAM team compositions.
  • Dodges change the team compositions towards "better" ARAM team compositions.
  • Please note: "better" above is subjective term and varies from person to person.

The first factor tells that you cannot use the champion data from other players to get any reliable data; they can have widely different champion pools and that will affect your research.

The last two factors tells that you have to use the initial champion you get for your research, if you wish to see if some champion comes from random more often than other. If you used the final result, your results will be skewed towards the better ARAM champions. You will need to consider changes to own champion pool and the free to play champions of the week in your research to get good results. And finally you will need thousands of games to get reliable results, so good luck with that.

I much rather play and have fun, than start doing work like research without pay, but people are different; I hear some people even love the work they do. I hope you have fun doing your research for as long as you happen to do that. :)

Sukrat1 votes

Hi Baka, thanks for your reply. You are absolutely right, there are just to many unknown dynamic parameters. Yeah the best I can do is to reach N=1000, when the time has come I'll send a new post containing my graphs.

"- You do not know the champion pool structure of the other players. "

As I played more than 1000 games with mates (who got more or less all champions) I assumed that the ennemy team had an analogue champion pool at their disposal which is totally inaccurate. I don't even know if how our "aram elo" is computed.

"Rerolls change the team compositions towards "better" ARAM team compositions.

  • Dodges change the team compositions towards "better" ARAM team compositions."

That's what I want to assess eventually by comparing it to several empirical ditributions.

""better" above is subjective term and varies from person to person."

I could not agree more, however getting Zyra Ziggs or Sivir seems so be always quite satisfying... ( all 3 appeared 7 times in 39 games, there was also Nautilus btw). But yeah, better define that betterness in the first place, i'll add 2 columns with the amount of V + D.

Actually i got this idea for 1 week, in 2 hours the file was created and since i just report thanks to ctrl+f each champions while the game loads. Takes no more than 1min. I'll keep you in touch then ;)

More from Gameplay

3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay764 comments
7,167
votes
3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay156 comments
3,510
votes
UPVOTE FOR URF
Gameplay606 comments
2,440
votes
Yes plz.
Gameplay456 comments
1,721
votes
3 Refunds per season, up vote if you want 3 refunds per season.
Gameplay114 comments
1,716
votes
When is ANYONE at riot going to address the awful new player experience?
Gameplay634 comments
1,383
votes

More by Sukrat

How about a gardener skin serie ?
Concepts & Creations2 comments
5
votes