EU PvP participation has collapsed, likely leading to the severe deflation many people are experiencing in 2v2 and 3v3. In 2v2, the major issue appears to be deflated rating at higher rating (i.e. the rating distribution has an unusually thin tail). In 3v3, EU realms are seeing a rating distribution that is incredibly narrow and with the centre of the distribution 300 rating points lower than in the US. On EU realms, 2100 rating corresponds to the top 0.5%. On US realms, both 2v2 and 3v3 rating appears relatively typical, though here too we may be seeing thinner tails (i.e. fewer people at high rating) than usual.
In 2v2, the melee meta is dominated by Fury Warriors and Subtlety and Outlaw Rogues. The ranged meta has seen the relative demise of Fire Mages, with Demonology Warlocks cementing their role as the dominant caster. In 3v3, the picture is a bit more complicated, with Fury Warriors, Unholy DKs, and Retribution Paladins being the most popular, but not necessarily the most performant. Demonology Warlocks, Affliction Warlocks, and Elemental Shamans are some of the key players in the ranged meta.
The healer meta has transformed from a Holy Priest-only regime to one where the different specs all appear relatively balanced. Discipline Priests are now clearly in the very top tier, but we no longer have a situation where there is one spec that is quite obviously the worst in all cases. Instead, most specs seem fairly competitive.
The ladder data is obtained from characters pseudorandomly sampled from EU and US realms. We only included characters that had logged on this season. This yielded 66,899 characters for EU realms and 64,597 characters from US realms. In order to be included in the analysis, a character must have played at least 50 games of 2v2 this season (N=5907) or 35 games of 3v3 (N=3585). The leaderboard data is a Blizzard-defined list, consisting of the top 5000 characters in each region/bracket.
Participation and rating distribution
PvP participation has taken a big hit this season. So far, out of characters that have been active this season, 18.4% have played at least one game of 2v2. This is far higher than for EU, where only 13.0% have played at least one game of 2v2. Of those who have played at least one game, the average number of games played is 41 for both regions.
In 3v3, 12% of US characters and 7% of EU characters have played at least one 3v3 game. This mirrors what we see in 2v2, where US PvP participation is generally far higher. However, compounding the issue of low participation is generally lower volume among EU participants: while US 3v3 players have played on average 36 games, EU 3v3 players have only played on average 27 games (25% fewer games).
Relative to last season, participation is very low, where we saw 17% and 21% for EU and US realms, respectively, roughly at this point in the season (compare that to 7% and 12% this season in 3v3 for EU and US realms, respectively). However, the 2v2 rating on both EU and US realms is roughly in line with that of Season 3. A notable exception is that the tails are far thinner, likely meaning that fewer people are able (or willing) to push to high rating. While 2v2 remains relatively normal, 3v3 on EU realms is seeing unusually extreme levels of rating deflation. Currently, the median rating on EU 3v3 is 1324, and only 5 characters in our EU sample of 1123 characters had a rating above 2100 (these are characters that have played at least 35 3v3 games this season). Thus, currently 2100 rating is similar to the old Gladiator cutoff of 99.5th percentile. For US realms, 3.7% of players have a rating above 2100, which is more in line with expectation.
While we should generally be careful in making causal statements, it seems plausible that this extreme rating deflation is related to the very low participation in EU 3v3. Note that while the sanctions against Russia likely explains why overall player numbers may be lower, they do not necessarily explain why PvP participation specifically is so low on EU realms compared to US realms, nor why EU players play far fewer 3v3 games than their US counterparts. Potentially, there is a self-reinforcing cycle here where people opt not to play because the ladder is deflated, in turn preventing the inflation of the ladder.
Kicking off this season as the main crushers are Fury Warriors. Fury Warriors are currently doing incredibly well in the upper-middle part of the ladder, with their representation falling off with increasing rating (comps above 2100 rating see relatively few Fury Warriors). Both Subtlety and Outlaw Rogues continue to be in a very healthy state, with Outlaw increasing subtantially in popularity over the past month.
The middling, but still very decent 2v2 melees are Havoc DHs, Windwalker Monks, Survival Hunters, Feral Druids, and Assassination Rogues, all being relatively well represented at all ratings.
Weaker performers this season are shaping out to be Unholy DKs, Arms Warriors, Retribution Paladins, Enhancement Shamans, and Frost DKs. However, it should be noted that many people may simply opt to play Fury instead of Arms, making it appear as though Arms is much weaker than its leaderboard representation would suggest. Indeed, if we use the ladder representation as an overall metric of how popular the spec is, we can see that Arms is not played much at any rating, and so from this alone, we would not expect to see it at high rating even if it was a pretty decent spec.
Unfortunately, the other specs at the bottom of the leaderboard lack such a favourable alternative interpretation to their poor showing, with most generally underperforming based on how often they are played.
Season 4 sees the continued march of Demonology Warlocks, with them now firmly overtaking Fire Mages as the dominant ranged spec. In fact, Fire Mages are now underrepresented on the leaderboard for the first time since records began (which admittedly is only since the start of the expansion). It seems that the Fire Mage survivability nerfs have had a significant effect.
Affliction Warlocks have become very popular with significant buffs to their damage output. However, despite this popularity, they are actually underrepresented at higher rating, meaning that they don't actually do as well as one would expect given how many people play the spec in 2v2. Beast Mastery Hunters on the other hand have been solid performers for a while, and continue to be in a pretty decent spot, despite losing ground on the leaderboard relative to last season.
The middling ranged specs are Shadow Priests, Elemental Shamans, Balance Druids, and Destruction Warlocks. Marksmanship Hunters continue to be a popular spec at low rating, likely due to the enormous burst potential of this spec, but are generally poor performers at higher rating. Destruction Warlocks are very unpopular as a spec, and are also underrepresented. However, they likely see a similar effect to Arms Warriors: there is another very strong alternative spec available, so the lack of popularity is not necessarily purely indicative of of the spec being weak.
It seems clear, however, that both Frost and Arcane Mages remain in a very dire state in 2v2, and while both have gained some representation on the leaderboard since last season, it is hardly enough to get excited about.
At long last, the dominion of Holy Priests has ended and we return to a more balanced healer meta. Discipline Priests are now top dogs (marginally), but Restoration Druids, Holy Priests, and Restoration Shamans are all incredibly performant healers in 2v2. Holy Paladins are arguably the worst healer of the lot, being the only healer that is underrepresented on the leaderboard. Mistweavers have been making inroads after seeing a number of small buffs, and now appear to be approaching a healthy state. It is worth noting that Mistweaver Monk is the least popular spec in the game, and so the fact that they are lower down on the leaderboard than Holy Paladins does not necessarily mean that they are in a worse spot. Indeed, once we adjust for the popularity of the spec, we see that Mistweavers outperform Holy Paladins in 2v2. Overall, we currently appear be witnessing a fairly well-balanced 2v2 healer meta.
Spec performance in 3v3
As in 2v2, Fury Warriors appear to be the dominant force. However, a worthwhile observation is that Fury Warriors have actually decreased in their leaderboard representation, while simultaneously increasing in overall popularity. Thus, Fury Warriors are actually more common at lower rating 3v3 than they are at higher rating, which is typically not the pattern we would see for an overperforming spec. Nevertheless, their popularity alone makes them a defining feature of the current 3v3 meta.
Unholy DKs remain in the same strong position in which we left them last season (with DK/DH being a particularly abundant comp this season), and while Retribution Paladins have taken a hit to their representation, they continue to be a strong performer in 3v3 while running (almost exclusively) Ret/Warrior/Healer. Havoc DHs are in a decent spot, but are clearly underrepresented given their high popularity.
Now for some controversy. The top four specs in terms of leaderboard representation are not necessarily the strongest specs. Spec strength is hard to quantify due to the number of factors that can influence it (such as whether that class has another viable spec, whether play rates/popularity of specs is uniform across the rating range, etc.). However, in this case, we see a fairly clear example for why naively interpreting leaderboard representation is problematic. While Feral Druids, Subtlety and Outlaw Rogues, and Enhancement Shamans all see lower leaderboard representation (i.e. representation at high rating), they are also just played by far fewer people. These specs, however, are all overrepresented given how commonly they are played by the general player base (i.e. their ladder representation), and this pattern is typically what we see when a spec is strong. So a reasonable (though, of course, by no means uncontroversial) supposition is that these four specs are at least as strong as the four specs at the top of the leaderboard.
As in 2v2, Arms Warriors aren't necessarily doing incredibly poorly; they are just deeply unpopular given how very strong Fury is as a spec. Windwalker Monks, Survival Hunters, and Assassination Rogues are fairly well represented given how commonly these specs are played. Indeed, the only melee spec that is quite clearly underperforming appears to be Frost DKs. All in all, the melee 3v3 meta appears to be fairly well balanced.
Demonology Warlocks have consolidated their position as the top ranged in 3v3, with Fire Mages having gone from a leaderboard representation of 5.4% at the end of last season to 2.8% now. Fire Mages continue to be a strong spec, but are now likely an A-tier spec rather than S-tier. With Affliction Warlock, we see the same pattern in 3v3 as in 2v2, where Affliction Warlock popularity has exploded over the past month, yet their performance at high rating is lagging behind. Shadow Priests, Elemental Shamans, and Balance Druids are all performant.
Both Marksmanship and Beast Mastery Hunters are hugely popular in 3v3, and Marksmanship in particular appears to have a decent overall representation, especially given that the overall pattern in the past several seasons is that they tend to be poorly represented at high rating while being well represented at low rating. Frost Mage and Destruction Warlock appear to be more niche specs (and still often paired with each other). Arcane Mage remains uncommon, but Arcane Mage RMP variants appear to be on the rise.
As with 2v2, Holy Priests have been dethroned and are now very much an average healer. In fact, the healer meta seems pretty well balanced, with no healer being either completely dominant or completely absent. Indeed, from the data, it is hard to single out a "worst" healer, as even Mistweavers are doing decently these days. It does seem clear, however, that Discipline Priests are now the leading healer.
If you have any comments or question, please pop by the Ludus Labs Discord. If you like the work I do on the site, please consider supporting the site on Patreon. Patrons get early release access to news on the site, insights into new tools and projects we're working on, and also enjoy special access to a Patron-only channel on our Discord.