Here’s Proof Why 3v3 Is Better for Competitive Battlerite

Posted by Zanetski 6 comments

I know I’m beating a dead horse here.

The debate between 2v2 and 3v3 has been going on for years. Even before Battlerite was a thing, we debated it constantly during the days of Bloodline Champions. Can’t I just let it go? Is it really that important? I think so. There are many reasons why picking a main format is necessary.

But let’s set aside opinions and look at data. Regardless of what you think is more fun or convenient, there is one true reason why 3v3 is objectively better than 2v2 for competitive Battlerite. The numbers prove it, and numbers don’t lie.

The Key Ingredient in Competition

There is only one thing that truly matters in any competitive activity. Sure we can talk about overall balance, fun and evolving metagames, a rewarding tournament system, a proper spectator experience, and all the other good stuff that goes into making an esport that people will adopt. But one thing reigns over all of that:

Consistency.

When two players go head to head, we should expect the better player to win most of the time. The same holds true when two teams compete. The converse is true too: we can judge which team is better than another by seeing how often they win against each other. If Team A beats Team Z over 90 percent of the time, the result is clear. If they share a 50 percent win rate, then we might say they’re on equal ground.

This is why RNG (random number generation) is so frowned upon in games aspiring to be esports. Randomness injects uncertainty, and uncertainty impacts consistency. When consistency drops too low, it becomes harder to take results seriously. Did Team A beat Team Z because they’re better or because they rolled a game-winning critical hit? Fluke wins are okay but they should be rare, not the norm.

Any mechanic or design decision that skews consistency should be frowned upon. Likewise, mechanics or designs that promote consistency should be encouraged.

Hence why the Grand Finals of any serious esports event increases the number of maps/matches played to determine a winner. Thousands (or millions) of dollars should never hang on a fluke win. When declaring a winner, you really want to make sure they earned it and deserve it.

For an esport to be taken seriously — by players and fans — it must hold consistency of results as its highest priority.

Why 3v3 Is Objectively Better

Most players intuitively know that 3v3 is more balanced and consistent than 2v2 in Battlerite, and we can point at all kinds of reasons why this might be true. Others disagree and think 2v2 is consistent enough for competitive play.

But what does the data say?

I went through battlerite-stats.ru’s leaderboards and combed through dozens of player profiles, looking specifically at win rate differences between 3v3 and 2v2. These are all top- and mid-tier players with significantly positive records, so we know they’re more skilled than the majority of players they encounter in queues.

Here’s what I found. These numbers are from the Season 5 Solo Leaderboards, only using League stats because Casual is itself inconsistent:

Averse (Global #1)

ProsteR18 (Global #2)

l0l (Global #3)

HotBiscuit (Global #23)

Lanofrose (Global #24)

Alboniks (Global #25)

Bo4 (Global #48)

TrinityInfinity (Global #49)

FireBlaze (Global #50)

What we see is a significant difference when top players queue for 3v3 versus 2v2: their win rates are regularly higher when playing 3v3.

This shows that something about 3v3 makes the mode inherently more consistent than 2v2. Seeing as these are some of the best players in the game, we can assume that they know what they’re doing. Why are they losing more often in 2v2? Is it because comps matter more? Is cooldown trading more prominent? Are matches frequently coming down to sudden death and turning results into coin flips?

We can’t say for sure. The data doesn’t speak to any of that. It only shows us that something is indeed unstable in 2v2 gameplay.

Can we trust this data? What if this player always queues solo while another always queues as teams? Wouldn’t this skew the data? If it does, there’s no indication of it. The trend still holds for nearly every player in the Top 100. Whether these people play alone or with friends, they clearly perform better in 3v3.

What Do We Do With This Info?

First of all, if you’re fed up with solo queue because you feel like you lose more often than you should, reconsider playing 2v2. People aren’t lying when they say that 3v3 is a different game — it pushes the skillcap higher by lessening the emphasis on 2v2’s cooldown trading.

The consistency makes 3v3 more fun and competitive.

More than that, we should keep pushing Stunlock Studios to decide on an official competitive format, even if it means picking 2v2. While I’d personally prefer the game to be designed for 3v3, it’s certainly possible to tweak and balance the game for 2v2 in a way that makes it more consistent and skillful.

But not while 3v3 hangs over its head.

Deciding on an official competitive format would mean removing the other format from League play (ranked). There should only be one ranked queue. If they want to keep a Casual queue for the non-competitive format, so be it.

This is a crucial step because it primes the playerbase for one proper format.

As long as both formats exist for ranked play, there will never be a consensus on balance. Some champs are good in 2v2, others are good in 3v3. If one format is designated as official, we can get that much closer to becoming a better balanced game.

And at the end of the day, that’s all we really want, isn’t it?

Zanetski

Written by Zanetski

He is the lead writer at Battlecrank. You can find him on Discord at Zanetski#5578.

Discuss This Article

avatar
Sort by:   newest | oldest
Econael
Econael
1 month 18 days ago

The only upside of 2v2 is that it’s so much easier to have 1 friend whom I can play with consistently than 2.
We don’t like having to rely on randoms.

Ronald
Ronald
1 month 13 days ago

I think this article has an interesting premise, though I’m not sure how compelling its conclusion really is. The premise is that by observing the win rates of the top players, you can deduce how consistently they can win and measure the competitiveness of the game. In a perfectly consistent game with perfectly consistent players, the most skilled player of the game would have a 100% win rate by definition. Since “competitiveness” is a measure of how much skill affects the outcome of the match, I agree that observing win consistency of the top players can serve to measure the competitiveness of the game.

However, Battlerite is hardly a perfectly consistent game played by perfectly consistent players. In fact, no game is, since people are terribly inconsistent things. Certainly, this doesn’t mean that all analysis is useless, but there are a number of factors that should be controlled before trying to draw any conclusions. For example, you did not take into consideration casual matches when observing win rates. This makes sense, as players will play casual matches for a variety of reasons beyond simply trying to win (trying experimental strategies, new heroes, playing with less competitive friends, playing while drunk, etc.) which will reduce the consistency. Unless I’ve misunderstood, you only look at the win rates of the top players in solo queue where that player only accounts for 33-50% of the total players on the team. A team where at least half of its members are chosen randomly is hardly consistent. I’m not sure to what extent this skews the overall results, but it seems to me that observing the top teams would lead to more reliable results than only observing the top players.

Similarly, you point out that “any serious esports event increases the number of maps/matches played to determine a winner.” Clearly larger sample sizes will yield results that are more reliable. Considering this, it seems a bit dubious to be drawing conclusions from a comparison of 2v2 and 3v3 win rates when the sample size for 2v2 matches of every player you’ve included in the article is considerably larger than the sample size of their 3v3 matches. In fact, all of the players you’ve included have played at least 4x the amount of 2v2 matches as 3v3 matches, with one playing as many as 20x more 2v2 matches. Again, perhaps even with the disparity in sample sizes the analysis is “good enough,” but it serves to undermine the overall conclusion.

Moreover, this says nothing about the particular methodology that was used to determine the “top 100” players. Perhaps whichever ranking algorithm battlerite-stats.ru is using to select the top 100 players happens to reward players that perform better in 3v3 than 2v2. Perhaps the players in the “top 100” are simply better at 3v3 than 2v2. Maybe 2v2 has a higher skill ceiling than 3v3, such that the top players are not skilled enough at it to play consistently. Personally, these things seem unlikely to me, but they serve to illustrate some more things that the article doesn’t acknowledge or account for before drawing its conclusion.

Furthermore, the article doesn’t only attempt to establish that 3v3 is more competitive, but also that 3v3 is more fun and also that 3v3 is the best “primary mode” because it is both more fun and competitive. I do not believe that you say anything to suggest that modes that are more competitive are inherently more fun, nor anything to suggest that 3v3 is somehow more fun for other reasons. Additionally, I don’t think that it’s a given that more competitive games are even better for the competitive scene, as contradictory as that may sound. Consistency and competitiveness go hand in hand, but on the other side of the coin is uncertainty and uncertainty brings excitement. Excitement brings viewers and viewers bring cash, which serves as an incentive for players and the cycle continues. I could go on from here, but I think that’s beyond the scope of this response.

Overall, I do think this was an interesting article. Similar to “Will F2P Save Battlerite? Maybe We’re Asking the Wrong Question,” I’m always glad to see articles that attempt to use actual factual data to support a point. I obviously have a few issues with it, but to be fair, you can only do so much to tackle such a large and involved topic in a single ~1000 word article. Hope to see more.

danl9rm
danl9rm
1 month 3 days ago

I think you’re missing a huge correlation that essentially explains away every conclusion you’ve made: # of games played in each queue.

Every one of your examples has more games played in 2v2 than 3v3. If you do some more digging through the numbers, you’ll see it’s very common, almost universal even, for a player’s winrate to go down with # of games played.

This makes sense because even if you can climb to GC with 100% winrate, you must then face many more other GC players than you did while climbing. The more you play after your climb, the less those initial “free wins” count in the statistics. Most, no, almost all, GC level 3v3 teams only climb to GC, if they ever make it past Champion due to queue times, and then stop playing for the season.

Moreover, many higher level 3v3 teams tend not to play as much until later in the season. This is harder to prove, but I’ve noticed the creme de la creme are busy climbing solo or even 2’s at the very start of the season, when there is 0 rank inflation and climbing is the most difficult.

Also of interest here is how many high level players have “mains” whose winrates are lower than that other multiple other champions, even with 100s of games played on these off-champs. I would be loathe to debate, however, that their skill level was higher with these other champions.

Lastly, it’s also possible that when high level players gather two of their buddies vs just one, that their aggregate skill level is much higher compared to other teams that also have to come up with three players. In short, high level players may have better winrates on their 3’s teams compared to their 2’s because everyone else’s 3’s team is worse. This doesn’t even take into account the tournament level experience high level 3v3 teams have over and above the general population.

wpDiscuz