Dissecting the online meta: Who falls and who rises on Wi-Fi?

By Mathew "EazyFreezie" Aliotta | 05/04/21

In 2020, the pandemic made holding offline events for Smash Ultimate impossible. This led to the entire scene shifting to online tournaments. Today we will be looking at the meta of WWRv5, the time period of online tournaments between January 2020 and June 2020. We will also compare the differences between the beginning of the online metagame to the previous offline season, PGRU v2. If you have not read the breakdowns of previous seasons, please be sure to check out PGRU v1’s and PGRU v2’s here.

Breakdown

Similar to the last breakdown, we will be using the same established rules set for the previous season. The WWR uses the same tiering system as PGRU, so we will be utilizing the tiers already established to break down results. C-Tier tournaments, the lowest ranked tournament, will be represented through the top 8 placements. Next up the scale would be B-Tiers, where we will take the top 16, and then for A-Tiers, we will take the top 32. Finally for S-Tiers, the tournaments with the highest accolades and tournament entrants, we will be taking the results of the top 64. 
A player must win a single game with a character for the result to count. If a player pulls out a character in a set but does not win a single game with that character, then the character is not counted. Also consider that character data is dependent on data recording from tournament and event runners and may be incomplete. Let's take a look at the results.

Placement Percentages

click to expand this and all other images in this post
The image above showcases a characters’ “placement percentage.” A placement percentage signifies a percentage out of all counted tournament results for the PGRU v2 season in which a character successfully placed, based on the rules laid down earlier. As you can see, of the 76 characters in the game through WWRv5, 38 accounted for at least 1.0% of results, exactly half of the roster.
Sonic ended up as the most prominent character in the metagame with a total of 4.93% of results and a sum of 103 results (101 main, 2 secondary). During the previous offline season, Sonic was #28, with 1.22% of results, a 3.71 point increase. Wolf was the #1 last season, and has dropped out of the top 5 for the first time, ending up at 6th. Wolf has 3.16% of results this season, and saw a 1.96 point decrease. The only new addition to Smash Ultimate’s roster this season was Byleth, who failed to make a significant impact on the meta their debut season.

Biggest Changes from PGRU v2

Here we see the biggest changes between the last offline and first online season. One of the most important changes we see is how many results were gathered in each tier. We had more tournaments and results to count than the previous season, but we see that the different tiers are not as balanced as before. With a total of only 4 tournaments, A Tier’s result count drops significantly. S Tiers and C Tiers on the other hand have much more weight on the meta’s results.
In terms of characters, the changes for both increases and decreases are much more drastic across the board compared to last time. Characters in the middle of the pack such as Cloud, Luigi, Mr. Game & Watch, ROB and Sonic saw massive growth. But even more shocking are the characters who did not even make over 1% in the offline meta who went on to become prominent characters in this meta. Donkey Kong and Ike were characters who did not break top 50 in PGRU v2 yet found significant enough gains in WWRv5 to crack the top 20. Mii Gunner and Piranha Plant were not even in the top 60 last PGRU season but were top 30 in the online meta. 
For overall losses, there are a significant number of characters who previously made up over 1% in results that failed to do so in the initial online metagame. Some of the most notable characters were Olimar, Shulk, and Wario, who were all top 20 last season. We also see a significant drop in percentage for characters who were in the top 10 last season. Pokemon Trainer falls just below the threshold to make it back into the top 10 while Mario and Peach see a significant decrease to fall out of the top 10 entirely. Lucina sees the harshest drop off, barely making it over the 1% threshold after being in the top 10 for 2 seasons straight. The only characters in the top 10 last season who saw an increase in percentage were Palutena, ROB and Snake. 

Monthly Top 10 Breakdown

This chart showcases the top 10 of WWR v5’s placement percentages throughout the seasons. Pre-quarantine, we see that Snake had a dominating presence in the metagame. We also see that Roy had 0% of the results in January. This is a first for any character who made the top 10, and is quite interesting to see. In March, where quarantine began, we see that no character in the top really stands too high above the rest, and Snake has a large 2% drop. From April-June, we see a bit of a pattern. The top 5 characters remain in the top 5 for the rest of the season. While Snake continues to decrease in results and remains dangerously close to falling out of the top 5, no one usurps him. 

Meta Relevancy

As stated earlier, 38 characters were able to reach up to 1.0% for their season placement percentage. Exactly half the roster was able to remain somewhat relevant in the metagame. But, there are some stronger drop offs in percentages between placements. The top 5 characters who are labeled as “Meta Defining” all had over 4% of the game's results, making more than 20% of the total metagame. 
As stated earlier, the top 5 held steady every month after quarantine started, resulting in a large percentage gap between them and the rest. In fact, the difference between #5 ROB, and #6 Wolf, is 1 whole percent. The lower percentages are a lot less spread out, creating a steeper graph where the percentages are hoarded towards the top characters. When compared to last season, you can see this is true: Even though fewer characters hit 1% during WWRv5, the PGRUv2 had a larger percentage of characters who were considered not to have a large presence in the meta, which included characters who hit over 1%.

Tier Breakdown

Here we see a breakdown of the results in each tier, which includes which characters have the highest amount of results, and their peak performances in each tier. Cloud was the only character who achieved 1st in every tier, achieved by players Sparg0 and Lights. Mr. Game & Watch, Ness, Snake, Sonic, and Young Link were the only ones who achieved 1st for 3 of the 4 tiers.
In S-Tiers, we see 4 of the top 5 make-up the top spots, with Snake just falling short behind Wolf. Palutena and ROB showcased the strongest results, with a 1% lead over 3rd, Cloud. Pokemon Trainer (#11) and Pac-Man (#14) stand out as the non-top 10 members who succeeded in the major format. 
With there only being a total of 4 A-Tiers, the percentage makeup for this tier is very different than elsewhere. Snake dominates this tier with a 1.5% lead over 2nd, Sonic. The limited results also allows many non top 10 characters to have high percentages, including three—Mii Gunner (#25), Bayonetta (#26), and Bowser (#28)—outside the top 20. The-B Tiers also have 4 of the top 5 at the top, with ROB missing the cut and Joker having quite a high percentage. This tier is the only tier that features all of the overall top 10 in its top 10.
Finally, Sonic and Snake have a very strong presence in C Tiers, with Sonic breaking 6.0%. Notable non top 10 members that garnered a large portion of results in this tier are Luigi (#13) and Ken (#23). This tier is where a majority of Ken’s results come from.

Player Base Size Analysis

This chart showcases the correlation between a character's placement percentage, and the amount of players who have contributed to that character's results. Compared to PGRUv1 and PGRUv2, the graph has a lot more outliers from the norm.
The cyan diagonal line showcases the scattered plots average linear growth. Characters above the line are characters who had a diverse range of players bring in results, while characters under the line had fewer players than average who gathered a majority of the results for said character. 
As just mentioned, there are a lot of characters who stem far from the line, both at high and low percent. Above the line, we notably have Palutena, who has an incredibly high player base amount compared to any other character by a large margin. Wolf, Joker, Pokemon Trainer, and Mario also trend above the line, with various players from across the world representing their characters. On the smaller percents, Hero is a very interesting case. This character almost has 20 players who successfully placed with him, but failed to break over 1%. 
On the opposite end, Sheik only had 2 players who successfully placed with him (Sharp and VoiD), but was just able to produce over 1%. Peach and Zero Suit Samus were also characters who had smaller player bases but consisted of players who consistently placed and entered events. Sonic is a very extreme case similar to Palutena. Sonic, who ended up with the highest placement percentages, had a very small player base  of 16 players. The most notable of the list are Sonix, who placed 26 times, SuperStriker, who placed 19 times and BlueJay, who placed 11 times.
There you have it, folks. WWRv5 began Smash Ultimate’s first online competitive year, showcasing a large shift in the meta from offline. Next time we will be analyzing July 2020 to December 2020, WWRv6.
Mathew "EazyFreezie" Aliotta has been involved in Super Smash Bros statistics since the release of Super Smash Bros for Wii U. His most notable work is his co-ran algorithm based ranking, OrionRank. Follow him on Twitter at @EazyFreezie 
Do you have a data/analytics-driven Smash Ultimate or Melee idea you'd like to turn into an article, or possibly a video for the PGstats YouTube? Pitch your idea with a 1-2 paragraph description of your idea to jack@panda.gg.
Meta Analysis