THE META REPORT NAME IS TOO LONG, TOO DAMN LONG (n°18)

First week of patch 2.13 / Ruination Event 3° week

Valentino (Legna) Vazzoler
08-04-2021

Data

Number of (Ranked) games analyzed: 72085

Last Update: 2021-08-11 17:32 1

by the Numbers
Patch 2.13 first week / Master players1,2
Characteristic N = 185,1363
Status
Other 83,954 (45%)
Ranked 72,084 (39%)
to Scrap 21,951 (12%)
Friendly 7,147 (3.9%)
Server
americas 89,807 (49%)
asia 27,491 (15%)
europe 67,838 (37%)

1 Max datetime recovered: 2021-08-04 20:59:56 UTC from 2021-07-28 21:00:00 to 2021-08-04 21:00:00 UTC

2 EU Master players in the ladder: 1215 while number of possible Master players recovered is: 1215

NA Master players in the ladder: 1378 while number of possible Master players recovered is: 1386

ASIA Master players in the ladder: 497 while number of possible Master players recovered is: 499

3 Metadata from Friendly Matches (that aren't Bo3) is not recoverable, the value may not be perfect since I lack the starting time of the game. The amount of Games to still scrap is also an estimation based on the 'position' of the game

Regions

Gini Index

I decided to add an additional information to help interpret the state of the Regions’ Play Rates: the Gini Index.

The Gini Index is a measure of heterogeneity so, in this case and in simpler terms, how much the play rates are similar. The Index goes (when normalized like here) \(in\) [0,1] and it’s equal to 1 when there’s a single value with 100% play rate or 0 when all play rates are equal. Of course a Gini Index of 1 needs to be avoided but it’s not like the aim should be 0. As said, it’s just to add some additional tools.

Also, the HHI is used by Dr.LoR is just another way to measure the Gini Index. So let’s also add some reproducibility / comparable numbers.

Before showing this weeks results I thought it would be better to have a generalized idea of past data. The following plot display the Gini Index over the latest 17 weeks for region frequencies by inclusion rate.

Play Rate

Plot

Table

Region Play Rate
Relative Frequencies by Inclusion Rate of a Region
Region Freq Shard
America Asia Europe
Ionia 20.09% 19.46% 20.72% 20.50%
Shurima 18.05% 17.33% 19.25% 18.30%
Bilgewater 14.98% 16.20% 13.75% 14.20%
Noxus 10.45% 9.88% 10.59% 11.01%
PnZ 9.69% 9.78% 9.52% 9.66%
Freljord 9.32% 10.27% 7.99% 8.88%
ShadowIsles 8.36% 7.96% 10.09% 8.04%
Demacia 6.12% 6.42% 5.17% 6.23%
MtTargon 2.94% 2.71% 2.94% 3.18%

Play Rate by number of Cards

Plot

Table

Region Play Rate
Relative Frequencies by number of times a Card within a Region is included in a Deck
Region Freq Shard
America Asia Europe
Ionia 20.26% 19.61% 20.29% 20.96%
Bilgewater 18.38% 20.83% 15.04% 17.21%
Shurima 17.30% 16.68% 18.34% 17.52%
PnZ 10.78% 10.79% 11.01% 10.67%
Noxus 9.63% 9.13% 9.93% 10.04%
ShadowIsles 8.93% 8.20% 11.33% 8.67%
Demacia 6.26% 6.54% 5.49% 6.31%
Freljord 5.54% 5.48% 5.72% 5.53%
MtTargon 2.90% 2.73% 2.85% 3.10%

Champions Combinations

Play Rates

Note: I’m currently not sure if I shuld add the Gini Index for Champions too. There are currently too many cases with many decks that are bound so have near 0 play rate.

Plot

Table

Day by Day

Highlisting the top10 most played decks (at the moment of the last game played).

Win Rates

Tie games are excluded

Meta Decks

Win rates of the most played combination of champions. Play Rate >= 1% in at least one of the servers.

Underdog

Top Win rates of the top10 best performing least played combination of champions. Play rate \(\in\) [0.1%,1%)2

Since I added the server-specific values it would have been wrong to ignore this also in this ‘Underdog’ section. I decided to leave the rule untouched, without taking in account the different servers. While for the “Meta Decks” it was more a natural expansion, here the results can widely differ by small changed and as mentioned in the past I would like to leave the current rule as much as possible unless it’s too broken. Of course one could complain as there decks whose results are mostly carried by a single server and/or not even played in the others.

All true, but there are limitation like the facrs that the asian server is very low on number during all the season so even the condition of a mininum of 50 games with playrate \(\in\) [0.1%,1%) is a lot screw the results. Also this section in not as the Meta Deck one where players are interested in seeing the performance of the most played decks but it’s meant to be a “look around there may be some dark tech that’s worth investing some time”.

A better solution would be ‘expanding’ this section or providing some tools to allows players to search the result. But there’s a problem, not providing immediate readable results and leave the research to the user is a bad choice in a report from my point of view at least and for sections like this one in particular. It could be worth exploring as a standalone option like the recently uploaded page for the MU table (MU-Table)

Looking at this week results it’s clear most of these decks can be grouped as ‘Plunder’ decks. I’m working on it, but I’m still far from reaching a quality of the results that satisfy me.

Match Ups

Match-up Grid

The win rates on the grid are among the 10 most played champion combination.

Match-up Grid v2

Just the OG way I used to display the grid. Aside for the different color palette it doesn’t bring more information compared to the other version that actually use more shades and which is also easier to the eyes. I’ll probably remove this version the next week.

Match-up Table

Filtered cases with less than 100 games

Deck Structure of the week

I guess it’s fine showing a meta deck once in a while, so of course the obvious choice is the rising star: Lulu/Zed

Lulu / Zed (DE/IO)

How to read the table:
- Play rate: How often a card is included in this class of decks / the table is order by this column.
- 3/2/1 is the relative and absolute frequency of the number of copies in the decks that plays them
- Frequencies from 50% to 100% are colored from shades of green to white to identify more easily the highest values

LoR-Meta Index (LMI)

The LMI 3 is an Index I developed to measure the performance of decks in the metagame. For those who are familiar with basic statistical concept I wrote a document to explain the theory behind it: , it’s very similar to vicioussyndicate (vS) Meta Score from their data reaper report. The score of each deck is not just their “strength”, it takes in consideration both play rates and win rates that’s why I prefer to say it measure the “performance”. The values range from 0 to 100 and the higher the value, the higher is the performance.

Cards Presence

Play Rate

It seems that not even Twin Disciple can beat Sharsight

Top 3 Play Rates by Region

Forgotten Cards

Cards that couldn’t find place even in a meme deck.

Region Demacia Freljord Ionia MtTargon Noxus PnZ Shurima ShadowIsles
1 Plucky Poro Augur of the Old Ones Nimble Poro Resplendent Stellacorn Crimson Aristocrat Academy Prodigy Destined Poro Stirred Spirits
2 Vanguard Lookout Sown Seeds Wish Wrathful Rider Unstable Voltician Waking Sands
3 For Demacia! Paddle Star

Legal bla bla

This Meta Report was created under Riot Games’ “Legal Jibber Jabber” policy using assets owned by Riot Games. Riot Games does not endorse or sponsor this project.


  1. Reminding that while I may collect the ‘list of all games-id’ played in a certain timeframe I can only collect all their metadata several hours later. So after the first release the numbers may be updated later during the week, be it for the sake of updating or modifying some code.↩︎

  2. Min number of games 50, during the times a meta/ladder just changed.↩︎

  3. LMI - Early Theory↩︎