Super Smash Bros. Ultimate has always been seen as a game with a ton of character diversity in its tournament meta. From the beginning, whether looking at a top 8 at a local or top 64 at a supermajor, the characters showcased were of a wide variety.
Super Smash Bros. Ultimate has always been seen as a game with a ton of character diversity in its tournament meta. From the beginning, whether looking at a top 8 at a local or top 64 at a supermajor, the characters showcased were of a wide variety.
Smash Ultimate, however, has a roster that has grown to feature over 80 characters. Although a single tournament may showcase numerous characters, when looking at results across multiple tournaments in a more holistic way, you might not find the diversity you initially expected. This leads us to an obvious yet still unanswered question: Is Ultimate's apparent diversity real, or just a reflection of its unprecedented roster size? This project is going to take a deep dive into Smash Ultimate’s meta, starting with PGRU v1, covering Spring 2019. The objective is to analyze how much character diversity truly exists in the metagame.
Breakdown
If we are going to look into the results of PGRU tournaments and try to showcase a metagame, some definitions need to be established. Firstly, what is the cutoff of what is seen as “meta” and viable? Our results are coming from a PGRU season, 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.
Secondly, how do we classify a player who has used multiple characters in a tournament run? The rule being applied is that as long as the player was able to win a single game with a character, then the character shall be counted for a result. If a player pulls out a character in a set but does not win a single game with a 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.
With this set of rules in place, let's take a look at the results.
Placement Percentages
The image above showcases what I am going to call characters’ “placement percentages.” A placement percentage signifies the number of times (out of 100) that a character was successfully played in all of the PGRU Spring 2019 season, based on the rules laid down earlier. As you can see, of the 72 characters in the game through PGRU v1, 30 accounted for at least 1.0% of results, combining to make up 81.15% of all placements. Palutena ended up with the highest percentage of 5.28%, with 78 total placements (43 as a main, 35 as a secondary), making her the most prevalent character in the first competitive season of Smash Ultimate.
Monthly Top 10 Breakdown
Even though Palutena was number one for the first season, she did not dominate the number one spot for the whole season. It was not even until May that she had the highest placement percentage, as she was out-placed by Snake and Wolf for the months prior. Wolf, Pichu, and Fox were the most inconsistent.
Within the last month of the season, Pichu drops below 1%, reflecting the nerfs Pichu received, heavily reducing the power of his previously devastating forward tilt and turning him into an even more brittle glass cannon through increases to his self-damage when using electric attacks. Snake, Mega Man, and Peach—despite nerfs to the latter two—wound up as the most consistent members of the top 10. After the first month, Snake never even fell out of the top 3.
Meta Relevancy
When you look at an overview of all the percentages together, you can see how condensed the results can be throughout the roster. The top 4 (Palutena, Snake, Wolf, and Peach) make up an entire 20% of all placings in the first season despite comprising less than 6.0% of the entire roster at the time. On the lower end, we see that more than half of the roster takes just 20% of tournament placements, and almost all of these characters were individually below 1.0% in total placements.
When you look at this graph, it does look like Smash Ultimate’s diverse nature comes from having a large roster rather than character viability, but that is not entirely fair. Some of the characters in the bottom 20% have better peak performances than others in higher percentiles. On top of that, every character on the roster had at least one placement, which is relatively incredible considering there are more characters than measured placements at even the biggest tournaments.
Breakdown By Tournament Tier
As mentioned previously, the PGR season has multiple tiers for tournaments. If you break down each category and see the placement percentages for each, as well as highest placement, you uncover some interesting correlations. The top 10 placement percentages of S-Tiers and the top 10 for the entire season are the same, but the characters are in a different order. This is interesting because all the S-Tier tournaments put together contributed the second most results.
C-Tiers put together actually contributed the highest amount of results, even though we only counted the top 8 placements. Another odd correlation regarding C-Tiers: We see two characters who do not show up in any other tier’s top 10 in Roy and Ness. Roy is the most interesting here due to being number 11 in the total placement percentage amount. Roy didn't break top 20 in any other individual tier in terms of percentage, showcasing that some characters' results are not consistent when comparing tiers.
Player Base Analysis
The last part of this breakdown is a chart showcasing the correlation between a character's placement percentage, and the amount of players who have contributed to that character's results. The blue diagonal line showcases the scattered plots average linear growth. In the bottom left are all the characters who make up less than 1.0% of results, and create a solid basis of the line's 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 gather results for the character than expected.
Wolf breaks far from the average, having an incredibly large number of different players who contributed results. As a main or secondary separately, Wolf has more than a 10-player gap over the second highest player base, Palutena. Palutena's player-base, though, despite being the second largest, is right in line with what we would expect given her copious results in the first season.
Joker is also another character who had a large player base compared to the average, which may have been due to the hype around being a DLC character. On the other side of the coin, Snake, Peach, and especially Olimar have much smaller player bases, at least at the high level, than would be expected from their placements. Olimar in particular barely features more than 10 different players, but those players contributed enough to make Olimar a top 10 character in placement percentage.
PGRU v1 was Smash Ultimate’s competitive debut, showcasing a plethora of characters with each character earning at least one placement in a PGR tiered tournament. Certain characters dominated a large portion of the meta, but every character showcased that they can be somewhat viable. Next time, I will be going over PGRU v2, covering Fall 2019, the next season of Smash Ultimate’s competitive life, where we will see which characters rose from this first season, and which fell off from the metagame.
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 This post and Tuesday's post on Melee's 2020 LAN tournaments were both paid posts written by freelance writers. 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