THE META REPORT NAME IS TOO LONG, TOO DAMN LONG (n°24)

Patch 2.15 - Week 2 - At the end of the (S1) World

Valentino (Legna) Vazzoler
09-22-2021

Data

Number of (Ranked) matches analyzed 49005 or 98010 games.

Last Update: 2021-09-22 20:07

Note : following my recent analysis the following decks:

Have been aggregated into a single archetype: “ASZ - Sivir Ionia”

In addition a quick fix to account for the Rubin-baits (burn decks that pretend to be other archetypes) decks have been applied and they are called:

by the Numbers
Patch 2.15 second week / Master players1,2
Characteristic All Games Ranked
N = 83,0663 N = 49,0054
Status
Ranked 49,005 (59%)
Other 32,124 (39%)
Friendly 1,937 (2.3%)
Server
americas 34,745 (42%) 20,997 (43%)
asia 18,862 (23%) 11,564 (24%)
europe 29,459 (35%) 16,444 (34%)

1 Max datetime recovered: 2021-09-22 15:50:59 UTC from 2021-09-15 21:00:00 to 2021-09-22 21:00:00 UTC

2 EU Master players in the ladder: 558 while number of possible Master players recovered is: 552

NA Master players in the ladder: 611 while number of possible Master players recovered is: 609

ASIA Master players in the ladder: 268 while number of possible Master players recovered is: 268

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

4 n (%)

Regions

Play Rate

Plot

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.

Table

Region Play Rate
Relative Frequencies by Inclusion Rate of a Region
Region Freq Shard
America Asia Europe
BandleCity 24.48% 24.50% 22.82% 25.63%
Noxus 16.63% 16.79% 16.94% 16.23%
PnZ 12.72% 12.85% 14.06% 11.61%
Bilgewater 11.80% 12.26% 11.42% 11.46%
ShadowIsles 8.37% 7.44% 8.02% 9.81%
MtTargon 6.50% 6.66% 5.90% 6.71%
Demacia 6.00% 5.94% 7.13% 5.29%
Shurima 5.19% 5.42% 4.31% 5.51%
Freljord 4.40% 4.05% 4.62% 4.69%
Ionia 3.91% 4.10% 4.79% 3.05%

Play Rate by number of Cards

Note : currently all dual region cards have only their main region as possible value assigned. The same problem also apply to the card’s inclusion rates, with Lulu being the most notable case affected by this.

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
BandleCity 28.70% 29.03% 26.65% 29.71%
Noxus 13.14% 13.10% 13.58% 12.87%
PnZ 12.76% 12.79% 14.33% 11.62%
Bilgewater 12.64% 12.97% 12.34% 12.43%
MtTargon 7.06% 7.35% 6.37% 7.16%
ShadowIsles 6.58% 5.74% 6.78% 7.51%
Shurima 5.81% 5.89% 4.79% 6.41%
Demacia 5.56% 5.50% 6.61% 4.92%
Ionia 4.11% 4.22% 5.07% 3.30%
Freljord 3.65% 3.41% 3.49% 4.07%

Champions Combinations

Play Rates

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%)1

Match Ups

Note : Regarding MU, this is not the most accurate estimation you can get from my data. If you want a better picture of the current meta it would be better to look at the dedicated MU-page where I use all “Ranked” games with the current sets of buffs and nerfs. While one may object I don’t account for optimizations and differences in skills acquired during the weeks, the overall number of games / sample size makes them a better source of information. So, in case, pls refer to the MU - page for a better “meta-investigation”.

Match-up Grid

The win rates on the grid are among the 15 most played champion combination.

Match-up Table

Deck (builder) of the week - by ImpetousPanda

As I added a section dedicated to decks’ information which includes Meta and Underdog decks this section can be dedicated to more obscure decks are peculiar cases. And this weeks it is impossible not to consider the new meta-impact from RubinZoo with his bait-decks.

Sunfolk Invoke

This week I wanted to highlight the power of social-media / influencers on play-rates.

This is just an example but I’m planning on doing an analysis that tries to follows the release of videos/tweets from peoples like Swim, Mogwai, Grapplr, Silverfuse (so more the content creator, so, no top players like MajiinBae, that’s for another analysis )

To do this, let’s start with the deck chosen for this week: the Sunfolk Invoke by by ImpetousPanda.

He posted this tweet at 2021-09-20 15:33, how many games were played during this week but before the tweet? not many, just 8.

What is the number AFTER the tweet? 20

No matter how it looks, as the creator himself is not included in the counting there is no deny that there seems to be a correlation between the number of games played following the tweet, inspiring people to try it, even if it was only one who spammed it (just partially, this case) and it’s safe to assume to say it’s not just a correlation but a causal effect.

Again, this is more playing with the data, what follows are the usual deck structures and most played deckcodes (by server)

Deckcodes

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)

New: added a Tier classification of decks based on their LMI value. This is more experimental and not as strict as a result of one of my analysis (in this case it’s a first glimpse to the distribution of the LMI values). The classification is as follows:

This method will be improved in the next weeks, maybe adding the ban rate which can be extracted from Gauntlet and Bo3 Friendly Matches.

The LMI 2 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 BW DE FR IO MT NX PZ SH SI
1 Hunting Fleet Redoubled Valor Iceborn Legacy Nimble Poro Sneaky Zeebles Shiraza the Blade Academy Prodigy Arise! Sinister Poro
2 Scrapshot Vanguard Lookout Sown Seeds Crystal Ibex Trifarian Shieldbreaker Back Alley Barkeep Destined Poro Tortured Prodigy
3 Bayou Brunch En Garde Shadow Flare Fledgling Stellacorn Crimson Aristocrat Midenstokke Henchmen Waking Sands The Etherfiend
4 Vanguard Cavalry Zephyr Sage Porofly Savage Reckoner Unlicensed Innovation Sandstone Chimera
5 Silverwing Diver Ritual of Renewal Sumpsnipe Scavenger
6 Kadregrin the Infernal Insight of Ages Amateur Aeronaut
7 Yusari Vault Breaker
8 The Empyrean The University of Piltover
9 Solitary Monk
10 Shadowshift
11 Scales of the Dragon
12 Swole Squirrel
13 Zinneia, Steel Crescendo
14 Coastal Defender

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

  2. LMI - Early Theory↩︎

  3. LMI - Adding a Ban Index↩︎