Build Battle Royale Matchmaking Is Broken

Right, I have reviewed every battle royale on this site and I have one consistent complaint that applies to all of them. The matchmaking is broken in specific ways and nobody seems to be fixing it. This post is about why and what could be done.

The TL;DR is that browser battle royale matchmaking suffers from these problems: skill spread is too wide, lobby fill times are too long for the small-player-pool reality, the bot-padding solutions everyone uses are obviously bots and reduce game satisfaction, and the leaderboard ranking systems are gameable in ways that punish honest play. None of these is hard to address. None of them are getting fixed.

What battle royale matchmaking is supposed to do

A battle royale match starts with maybe 50 or 100 players. The match runs for 10 to 20 minutes. By the end, one player wins. For this to be interesting, the 50 to 100 players should be roughly similar in skill level. If they are not, the match becomes a stomp by the most skilled player and an unfair grind for the rest.

Matchmaking systems job is to assemble lobbies of similar-skill players from the available pool. This requires knowing the players skill levels (which the system tracks through previous match results) and finding enough similar-skill players to fill a lobby within a reasonable wait time.

On a big platform like Fortnite or PUBG, this works because the player pool is huge. There are always similar-skill players available. On a browser game with maybe 200 to 2000 concurrent players, the math gets harder. There are not always enough similar-skill players to fill a lobby quickly.

What browser battle royales actually do

Three patterns dominate.

First, accept the wide skill spread. The matchmaking fills lobbies with whoever is available regardless of skill. Lobby fill time is good. Match quality is bad. New players get destroyed by experienced players. Experienced players get bored stomping new players.

Second, force narrow skill spread at the cost of fill time. The matchmaking waits for similar-skill players to be available. Match quality is good when matches start. But fill times can stretch to several minutes, and players abandon the lobby before it fills.

Third, pad lobbies with bots. The matchmaking takes the available human players (maybe 20 of them) and adds 30 to 80 bots to bring the lobby to its target size. Lobby fill time is excellent. Match quality is bad in a different way: bots are easy to kill, which gives early-game players a misleading sense of success. The mid-game where bots run out and real players remain is where the difficulty spikes obviously.

All three patterns produce bad matches in their own way. Browser battle royales pick one of the three and ship it as if there is no alternative.

Why this is structurally hard

Math is the problem. Battle royale wants 50 to 100 players. Browser games have small concurrent-player pools. The two requirements collide.

A console battle royale can assume 100,000 concurrent players globally and fill a 100-player lobby in seconds with similar-skill matchmaking. A browser battle royale might have 500 concurrent players globally. Filling a 50-player lobby with similar-skill players from a pool of 500 is mathematically harder. You might not have 50 similar-skill players in the active pool at any given moment.

This is why bot-padding is so common. It is the only solution that produces fast matches with the available player pool. The fact that it produces bad matches is treated as acceptable.

What could be done differently

A few options.

Reduce the lobby size. A 20-player battle royale is dramatically easier to fill with similar-skill players than a 100-player one. The genre has been pushing lobby sizes up for marketing reasons (100 players sounds more impressive than 20) but smaller is better for browser-game matchmaking realities.

Use cross-region matchmaking. Most browser battle royales fill from a single regions player pool. Cross-region adds latency but expands the available player pool by an order of magnitude. The trade-off (similar-skill matchmaking with 80ms more lag versus same-region matchmaking with bots) is one I would take.

Make bots obviously bots. If bots are unavoidable, label them. Show them with different visual indicators. Let the player know which kills were against humans and which were against bots. This breaks the lie that current systems try to sell. The lie damages player trust when discovered.

Implement skill-based difficulty curves for bots. If bots have to be there, scale their skill to match the player. New player gets weak bots. Experienced player gets strong bots. This adapts the bot-padding solution to skill matchmaking, even if the human-versus-human matches are still skill-mismatched.

What the browser battle royales on this site do

Build Battle Royale on this site at 4 stars uses approach number three (bot-padding) and the review notes this. Build Master IO at 3 stars also uses bot-padding but worse, with bots so easy to spot that the early-game feels like target practice. The 3-star rating partly reflects this.

Best of a bad set is Snake Royale at 4.5 stars, which uses 50-player lobbies but has enough concurrent players that the matches are usually human-only. The smaller lobby size (50 not 100) and the focused player base produce better-quality matches than the bigger battle royales.

This is the model I would like to see more battle royales adopt. Smaller lobbies. Human-only when possible. Honest about bots when they are required. And matchmaking that pays attention to skill rather than just availability. Snake Royale gets it right within the structural constraints. Most others do not.

What this means for you

When you start a browser battle royale, the first match tells you a lot. If you cleaned up early-game players easily and the late-game suddenly got hard, you were probably playing against bots in the early game. If you struggled from the start, you were probably against humans across the board.

Knowing which kind of match you are in changes the play strategy. Against bot-heavy early games, you can take risks because bots fail to capitalise on your mistakes. Against human-only matches, you need to play more conservatively from the start because everyone in your bracket plays at your level.

Matchmaking transparency that I would like to see is unlikely to come from the aggregators or the developers. Editorial sites like this one are stuck doing the analysis in prose, the same way we are stuck doing genre-disambiguation in prose. Read the reviews. Look for the matchmaking notes. They are there because they matter, even if the games would rather you not think about them.

Battle royale on a browser is a structurally compromised format. The compromises are real and not the developers fault entirely. But the compromises are also worse than they have to be in most current implementations. The genre could be better and isnt.