The Copycat Meta
Does pro Apex really copy TSM’s homework on the meta?
This is a guest post by Singh Labs, whose visualizations of tournament pick rates are an amazing resource for the community. Go follow them on Twitter. And if you like the post, support The Final Circle by subscribing below so that I can continue to pay writers for excellent work like this.
After a long off-season, the competitive Apex machine is beginning to spin its gears once again. The inaugural Apex Legends Global Series Pro League kicks off this Saturday, October 16th. Invited teams are playing in minor tournaments to prepare, and aspiring pro league squads have finished grinding through the preseason qualifiers to earn their slot.
Now that competitive action has picked up again, the issue of copycatting in the metagame has cropped up, centered around the latest wave of Valkyrie-Gibraltar-Caustic team compositions. A tweet from former TSM team member Albralelie sparked a large part of the discussion:
Now this is an intriguing claim. Is TSM really so admired and watched so closely that the team can essentially dictate large portions of the competitive meta as a whole? And just how much does copycatting occur in the game overall?
Luckily, we can look into the team composition data from the first ALGS season to get some answers.
Choose your fighter
Before we get into the findings, let me begin by mentioning that all data being looked at here is from the Grand Finals of major tournaments, and therefore relevant to the highest level of competition – Tier 1 competition, if you will. I think it’s fair to assume that copycatting does occur with a certain regularity at a Tier 2 level, and especially beyond, where players are less likely to consider the intricacies of their team comp decisions and more likely to adopt a certain comp if they see it being played successfully by a top team. What intrigues me is whether this line of thinking is apparent at the top level of the game as well.
To investigate this, I used the following methodology:
I gathered team composition pick rates from major event Grand Finals in the NA region, beginning with Summer Circuit Online Tournament 1 (a good place to start, given that the variety in competitive Legend picks increased substantially following this event, with prior events being near-exclusively comprised of Wraith, Wattson, Pathfinder & Gibraltar)
I noted the compositions used by the winning team, and by TSM, for each event
I then noted the growth or decline in usage of those same compositions in the following event
The idea here is twofold: if there is indeed copycatting at the top level, we would expect it to be reflected by a growth in usage for tournament-winning compositions. And we can compare that rate to the results for team compositions used by TSM to judge their relative influence.
TSM doesn’t rule the roost
The data shows that TSM and other top teams don’t have much influence at the highest tier of the competitive meta.
Since SCOT1, NA tournament-winning team compositions have, on average, declined in usage by 1.2% for the following event. TSM’s compositions? A 1.3% decline. In both cases, basically a wash as far as real influence goes.
What does this tell us? A couple things, I think. First, rote copying of prominent or successful team compositions does not seem to have occurred with any regularity at the highest level of competitive play last season. Second, and more importantly, if even major tournament wins were not enough to consistently influence team composition usage, that means Legend balance (plus the cadence of game updates – more on this below) has been the driving factor. Top level teams and players have been far more likely to drop their established compositions and try something new in response to a major patch than in response to a win from an off-meta squad.
Let’s not be too hasty, though. Maybe it’s rare to see teams lifting full compositions from one another. But what about individual Legends? After all, if teams are in fact keeping more than a close eye on TSM’s latest use of Valkyrie-Gibraltar-Caustic, we wouldn’t necessarily expect all of them to run exactly that composition. Their influence could still be there even with the swapping of a Legend or two.
I took a look at this as well. I tracked the growth or decline of individual Legends with the same method, and the data again showed that there wasn’t much copying going on. With the exception of Octane (who I’ll address in a minute), Gibraltar was the Legend with the most positive swing in pick rate after being used in a winning composition or by TSM, but the level of influence in both cases was pretty minor: an overall average increase of 3.3% when part of the winning squad, and 5.6% when used by TSM.
This isn’t to say that everything about the copycatting narrative is meritless. As mentioned, it does seem to occur every once in a while in lower-tier lobbies. Albralelie mentioned one of these times himself:
I certainly remember noticing an increase in Revenant-Crypto teams being played in scrims and minor tournaments towards the tail end of the Summer Circuit. It just never made its way en masse into Tier 1 lobbies (Revenant and Crypto’s usage in Grand Finals actually fell slightly after they were played by TSM). If TSM does partly control the NA meta, that control doesn’t quite seem to have extended to the very top of the region’s food chain thus far.
There is one case, though, that deserves particular scrutiny. If you were to single out one example that proves TSM’s ability to dictate the meta at the highest level, it’s Octane. Octane’s pick rate jumped an average of 26.8% by the next event when used by TSM – but he was only played in the final three major Grand Finals last season. It was a very noticeable change to competitive play. Octane’s rise halted the longtime pick rate dominance of Wraith, and in the biggest tournaments of the season to boot. But we are still talking about a pretty small sample size. Octane’s quick rise in pick rate also follows a trajectory similar to other Legends that have received game-changing buffs just before major tournaments, such as Bloodhound and Crypto. Would Octane have risen in popularity as sharply as he did had TSM not picked him up? Perhaps not. But I think it’s fair to say he was destined to make a big impact all on his own.
A look ahead
What I find interesting going forward will be keeping this whole thing in mind as the pro league plays out. As I hinted earlier, I think part of the reason we haven’t seen any clear pick rate trends based on winning compositions has a lot to do with the irregularity and timing of balance patches versus major tournaments.
During the first ALGS season, major tournaments were regularly 3-4 weeks apart, and sometimes more. Balance updates to the game came around even more infrequently. More than that, the competitive scene seemed to be on a peculiar cadence compared to the production cycle of the game. There were several instances of competitively-significant balance updates dropping immediately before some of the biggest tournaments. You may recall, for example, Crypto and Bloodhound appearing out of nowhere in the Summer Circuit Playoffs, or more recently Valkyrie being a pivotal factor right out of the gate for the ALGS 2021 Championships.
Pro League represents a golden opportunity from a meta and trends tracking perspective. We’re going to get high-level and high-stakes gameplay at regular weekly intervals, providing not just a lot of entertainment, but a lot of data points over shorter periods of time, and under the same set of game balance circumstances.
I know I’m looking forward to it.
Liked the post? Want to see more original work like this during the ALGS season? Support The Final Circle by subscribing below so that I can continue to pay writers for their excellent work.