Every few months, the same argument erupts across gaming Twitter, Reddit, and YouTube. A streamer complains about “sweaty lobbies.” A casual player complains about getting stomped. A game developer tries to explain their matchmaking system and gets ratio’d into oblivion.
The debate is always framed the same way: SBMM is either ruining games or saving them. The pro-SBMM side says it protects new players. The anti-SBMM side says it punishes good players. Both sides are right. Both sides are missing the point.
SBMM isn’t a good or bad system. It’s an impossible tradeoff — and understanding why it’s impossible is more useful than picking a side.
What SBMM Actually Does
Skill-Based Matchmaking means the game tries to put you against players of similar skill. That’s it. The implementation varies wildly — some games use visible ratings (Valorant’s RR, League’s LP), some use hidden MMR, some use recent performance metrics, some use a combination — but the core idea is the same.
The goal is simple: create matches where both sides have a roughly equal chance of winning.
This sounds obviously good. In practice, it creates problems that aren’t obvious at all.
The Fundamental Problem
Here’s the core tension that nobody wants to acknowledge:
For a match to feel fair, both sides need to have a ~50% chance of winning. For a match to feel rewarding, you need to win significantly more than 50% of the time.
These two goals are mathematically incompatible.
Strong SBMM gives you fair matches. Fair matches mean you win about half the time. Winning half the time feels like a treadmill — you’re working hard and going nowhere. No matter how much you improve, the system adjusts and your opponents get harder. You never feel the improvement.
Weak SBMM gives you varied matches. You stomp sometimes, get stomped sometimes, and have close games sometimes. The stomps feel great (for the stomper) and terrible (for the stompee). The variety feels more “natural” but is objectively less fair.
No SBMM means pure random matchmaking. This is what most games had before ~2015. Lobbies were a mix of skill levels. New players got destroyed. Experienced players got easy kills. It “felt” better for above-average players because they won more than they lost. It felt terrible for everyone else.
Every game developer is choosing a position on this spectrum. Every position makes someone unhappy.
Why Good Players Hate SBMM
The anti-SBMM argument from skilled players isn’t actually about fairness. It’s about feel.
When you’ve spent hundreds of hours getting good at a game, you want to feel the difference between you and someone who hasn’t. You want to see your improvement reflected in your results. In a strong SBMM system, improvement doesn’t feel like anything — your lobbies just get harder, and you’re still winning 50% of the time.
This is especially brutal in “casual” modes with hidden SBMM. You queue up to relax, try a new weapon, mess around with friends — and the system puts you against players who are also very good, playing their main loadout, trying to win. The “casual” playlist doesn’t feel casual because the matchmaking doesn’t distinguish between your sweaty self and your goofy self.
There’s also the social problem. When a Diamond player queues with their Silver friend, SBMM has to decide: match at the Diamond level (Silver friend gets destroyed), match at the Silver level (Diamond player stomps), or split the difference (both have a bad time). None of these are good options. Most games choose “match near the higher player,” which means your friends don’t want to play with you.
These are legitimate complaints. SBMM genuinely makes the game feel worse for skilled players in casual contexts. That’s not a bug in SBMM — it’s a feature working exactly as designed, protecting worse players at the cost of the experience of better ones.
Why New Players Need SBMM
Here’s the other side: without SBMM, most new players quit.
This isn’t theory. It’s measurable data that every studio has and most don’t share. When new players consistently face experienced players in their first few sessions, retention craters. They lose every fight, don’t understand why, and uninstall. No amount of tutorials or onboarding fixes the experience of being annihilated by someone with 2,000 hours.
Before SBMM was standard, games relied on a constant influx of new players to replace the ones who bounced off. This worked when gaming was growing rapidly (2000-2015). It doesn’t work in a mature market where every game is competing for the same players.
The math is simple: if your game bleeds 80% of new players in the first week because they’re getting destroyed by veterans, you need five times more acquisition to maintain population. SBMM cuts that bleed rate dramatically. It’s not a philosophical choice — it’s a survival mechanism.
Games without SBMM don’t die because SBMM is inherently good. They die because the alternative — a shrinking pool of increasingly skilled players farming an ever-smaller group of newcomers — is a death spiral.
The “Just Add a Ranked Mode” Fallacy
The most common suggestion is: “Put SBMM in ranked, take it out of casual.”
This sounds reasonable. It doesn’t work.
When casual mode has no SBMM, good players go to casual because it’s easier. This makes casual harder for bad players, who then go to… ranked, because at least there they’ll face fair opponents. Now ranked is the casual mode and casual is the tryhard mode. You’ve just swapped the labels.
Call of Duty tried this repeatedly. Every time they weakened casual SBMM, the same cycle played out. Casual lobbies became just as sweaty as ranked because the best players naturally gravitated toward the easier competition.
The only way “ranked for SBMM, casual for random” works is if good players voluntarily choose the harder mode. Some will. Most won’t. Humans optimize for the path of least resistance.
The Connection Quality Tradeoff
There’s a technical dimension that gets overlooked: SBMM competes with connection quality.
In a perfect world, you’d match with players at your exact skill level who also happen to live near you. In reality, strict skill matching might pair you with someone across the continent. Loose skill matching finds someone next door who’s much better or worse than you.
This tradeoff hits hardest in games with smaller populations or non-standard regions. If you’re a high-rated player in OCE at 3 AM, strict SBMM means you don’t get a match. Loose SBMM means you play on 200 ping. No SBMM means you stomp — which feels great for you and terrible for the three people in your lobby who were just trying to play.
Battle royales struggle with this the most. Filling a 60-150 player lobby with similar-skill players requires a massive concurrent population. Most BRs quietly loosen SBMM during off-peak hours and hope nobody notices.
What Actually Works (Sort Of)
No system is perfect, but some approaches are less bad than others:
Engagement-optimized matchmaking (what most games actually use) doesn’t target 50/50 win rates. It targets a rhythm — you win a few, lose a few, have a close game, repeat. The system tries to keep you in a flow state where you feel challenged but not crushed. This works better for retention than strict skill-matching, but it feels manipulative because, well, it is.
Input-based matchmaking (Fortnite, some CoD modes) separates controller and mouse/keyboard players. This doesn’t solve the skill problem but removes the most egregious input advantage complaints.
Role-queue adjustments (Overwatch, Marvel Rivals) can match slightly less strictly for under-populated roles. This reduces queue times without massively impacting fairness.
Declining protection — some games give new accounts strict SBMM that gradually loosens as you play more. This protects the critical first hours without permanently constraining experienced players.
Seasonal resets — hard or soft MMR resets each season create periods of chaotic, varied matches that feel refreshing before the system re-sorts everyone.
The Uncomfortable Truth
Here’s what nobody on either side of the SBMM debate wants to hear:
The system isn’t the problem. The expectation is.
If you expect to win most of your games, you need opponents who are worse than you. For you to consistently face worse opponents, those opponents need to consistently face better opponents — you. Your fun requires their suffering.
That’s not a matchmaking problem. That’s a math problem. And math doesn’t care about your feelings.
The real question isn’t “should games have SBMM.” It’s “who should bear the cost of competitive gaming.” Right now, SBMM spreads that cost roughly evenly — everyone wins about half, everyone loses about half. The alternative is concentrating the cost on weaker players so stronger players can feel better.
Every time you complain about SBMM, you’re really saying: “I want to win more, and I don’t care who has to lose more for that to happen.” That’s a valid preference! But own it. Don’t pretend it’s about game design philosophy when it’s actually about who gets to have fun at whose expense.
Where It’s Heading
The next frontier is AI-driven dynamic matchmaking that considers dozens of variables beyond simple skill rating: playstyle, social behavior, play session length, party composition, recent win/loss streaks, time since last session, and more.
EA has patents on this. Activision has patents on this. Epic definitely does this already.
Whether this is better matchmaking or creepy behavioral manipulation is a debate for another article. But it’s where the industry is heading, and “just remove SBMM” isn’t a realistic option when player retention directly determines revenue.
The SBMM problem isn’t getting solved. It’s getting more sophisticated. The best thing you can do as a player is understand the tradeoffs, pick games whose matchmaking philosophy aligns with what you value, and stop expecting a system that makes everyone happy.
That system doesn’t exist. It can’t. And the sooner we accept that, the sooner we can have honest conversations about what we actually want from competitive games.
Discussion