106 Team-Ups: How Marvel Rivals Turned the Ranked Ladder Into a Combinatorial Problem

On a late September evening, I sat in front of two screens in a small apartment in Shanghai. The left screen showed a Diamond-ranked match in Marvel Rivals. The right screen showed a spreadsheet I had built myself, exactly 106 rows, each row a hero pairing capable of chaining their abilities together. The player in that match did not pick the character he liked most. He waited to see what the teammate beside him would pick, and only then clicked. Forty seconds of the draft phase passed just to answer one question: can these two lock together.

There was a moment in that match that made me pause and take notes. A player picked Scarlet Witch. His teammate immediately switched to Magneto, and the instant the two appeared on the field together, Scarlet Witch's ability bar lit up in a different colour. She was not stronger because of better skill. She was stronger because someone stood beside her in the right way.

That was when I understood Marvel Rivals had changed the core question of the hero-shooter genre. The old question was "which hero is strongest". The new question is "which pairing is strongest". And the answer to the new question runs 106 rows long.

Context: a self-replicating system

Marvel Rivals is a 6v6 hero shooter, launched as a direct competitor to Overwatch 2. Its biggest difference lies in the Team-Up system. Every hero in the game owns at least one chained ability, and these abilities only reach full power when the correct teammate appears beside them.

The mechanic runs on two layers. The first layer is a base effect, always present even when you play alone. The second layer is an enhanced effect, triggered only when the designated partner is also on the battlefield. In other words, half of each hero's power is free, and the other half is conditional. I have re-checked this point many times in live matches, and the way it operates reminded me of relay teams: each runner can run alone, but only when someone waits at the handoff line does the real speed burst out.

Season 10 has added The Hood, a new character, along with its own Team-Ups. Notably, the developer states it will release a new hero roughly every month, and every new hero must come with Team-Up abilities. No exceptions. This is a long-term design commitment, and that commitment creates the entire story behind it.

With two Team-Ups per hero, a full six-player composition produces a dense interaction web. As the hero count rises month by month, the edges in that web rise too. The count now sits at 106 pairings. Next season it will be 108, then 110, and so on.

The track and the pitch are not far apart; few people simply bother to run a full lap to see it.

Core analysis: a balance problem with no final answer

The Team-Up system is the single most meta-defining lever in Marvel Rivals. It turns the familiar question of the shooter genre into a combinatorial problem. When you change one pairing, you change not only two heroes, you change the entire composition chain revolving around that pair.

Based on my experience tracking matches across different ranks, from Silver to Grandmaster, one pattern emerges clearly. At low ranks, players pick characters by personal preference. At high ranks, players pick characters based on the pairings available in their composition. That shift is not driven by better aim, but by the awareness that half of the power sits with the person beside you.

Take the classic pairings. Hulk and Iron Man share gamma energy. Rocket Raccoon and Groot reinforce each other in both damage and survivability. Luna Snow and Namor create a layer of protective ice. Thor and Storm resonate with lightning. Hawkeye and Black Widow share battlefield vision. Each pairing is a small contract between two players, signed in silence before the match begins.

The most interesting thing I found was not the pairings themselves, but the pressure they place on the whole team's decisions. In a 6v6 match, if two players pick a strong pairing, the other four must orbit that pair. The composition becomes a solar system with two large planets at the centre. Anyone drifting off the orbit is accused of breaking the comp, even if they picked the character they are best at.

Here a problem appears that I call the combinatorial burden. With 106 pairings and a hero added each month, the game's balance surface grows faster than any tuning team can plausibly keep up with. Each new hero does not just add itself to the list, it retroactively adds two new edges to the web. A new hero can accidentally raise the power ceiling of a hero released in season one, simply through one unexpectedly strong pairing.

I spent time comparing how other titles in the genre handle a similar problem. Overwatch balanced by limiting heroes by role. League of Legends wrestled with hundreds of champions, but most of their interactions are linear, not a spiderweb. Marvel Rivals chose the hardest path: every new hero must pair with old heroes, and that commitment is publicly declared.

People do not run to leave others behind, but to see how far they can go together.

There is a small but important detail most players overlook. The base effect always existing means a lone player keeps baseline value. The enhanced effect only triggering with a partner means the real reward sits in coordination. This two-layer split softens the pairing pressure but does not remove it. You are not punished for playing alone, but you are left behind when others play as a pair.

And this is where I want to pause a little longer, because it touches what I have always believed in sport. Elite sport is never just the sum of good individuals. It is the product of individuals who know how to place themselves correctly. Novak Djokovic once said he does not train to hit the ball better, but to read the match better. In athletics, Jamaica's 4x100m relay team broke the world record not because its four men were the fastest in the world, but because its four men handed the baton most smoothly in the world. Marvel Rivals, through a Team-Up system, is simulating exactly the principle athletics and football have taught me across fifteen years of watching.

Contrarian angle: when knowledge becomes a weapon

The truth is, the most frightening thing about a 106-pairing system is not that some pairing is too strong. It is the sheer volume of knowledge a player must remember.

106 pairings far exceed the reactive memory of an ordinary player. No one can, within forty draft seconds, sift through all 106 pairings in memory to find the best one. Players are forced to use notes, spreadsheets, or external tools. The original guide I read recommends players bookmark the list for reference, advice that sounds harmless but in effect admits the game's knowledge barrier has risen beyond what pure memory can cross.

I once mispronounced Mbappe's name three times, but football has never been wrong about kindness. That story taught me that speed is the enemy of accuracy, and that unverified knowledge sooner or later becomes error. In Marvel Rivals, misremembering one pairing can collapse an entire composition.

This is where the biggest risk appears. It turns out Belgium's pressing did not just win the ball, it took the opponent's belief too. In Marvel Rivals, the 106-pairing system does not just shape the meta, it takes away the newcomer's chance as well. A player just entering the game must learn 106 pairings just to start playing with awareness. That entry barrier turns a knowledge advantage into a time advantage, and a time advantage, in sport, is the hardest thing to close.

Fighting games once solved this by limiting the number of characters in a match. Marvel Rivals chose the opposite: infinite expansion. I am not sure that is the right or wrong design choice. I am only sure it shifts the competitive focus from skill to knowledge, and that may be fine for a live-service game, but it is a problem for a sport that is broadcast. Viewers cannot feel why one click shines if they do not know the pairing behind it.

Second contrarian angle: a long-term commitment as a double-edged promise

There is a point I believe the analyst community has overlooked for too long. The developer's commitment never to release a new hero without Team-Up abilities is not a neutral rule. It is a contract with existing players. It guarantees that a hero you have invested hundreds of hours into will not be erased by a stronger new one. A new hero will pair with old heroes, not delete them.

That commitment protects the mental ownership of veteran players, and that is a decent thing. But it is also a design shackle. Every new hero must find a place in the existing web without breaking the power ceiling. The more pairings there are, the harder it is to find a neutral place. The developer has tied itself to a problem whose complexity grows non-linearly over time.

The quietest summer often hides the loudest transfers. The tenth season of Marvel Rivals looks on the surface like one new character, but underneath is a structure growing month by month. I have followed this game's content cadence for many months, and what I noticed is that the pace serves not only to retain old players, but to create a content economy: guides, videos, pairing tier lists, community debate. Each new pairing is a new topic to discuss, a new reason to return.

But there is a paradox I cannot ignore. If there is one new hero with two new pairings each month, the game's balance state will never stay stable long enough to be validated by real data. Players are living in a permanent adjustment period. No one knows which pairing is truly strongest, because by the time the community answers, the season has turned the page.

For a live-service game, that can be a plus. For an esport wanting to build a fair and watchable tournament system, it is an obstacle. I look at athletics and see something interesting: the rules there are extremely simple, which is exactly why spectators need to learn nothing to feel the race. Marvel Rivals is going the opposite way, and I am not sure that is the road to a universal sport.

I once mispronounced Mbappe's name three times, but football has never been wrong about kindness. I repeat that because it reminds me that complexity does not automatically mean depth. Thirty-two cards in a deck do not make a game better. What makes a game better is how those cards talk to each other. Marvel Rivals has created 106 conversations. The open question is how many of them are truly worth listening to.

Risk and opportunity intertwined

Based on what I observed, the biggest risk of this system is not one broken pairing, but the sprawl of the balance surface. As the number of pairings grows arithmetically, the chance of one accidentally overpowered pairing grows too. The history of live-service games shows that must-pick duos always appear when a pairing system is forced to expand quickly. That has not yet happened to Marvel Rivals at mass scale, but its structure is inviting that scenario to arrive.

Conversely, this system also creates a rare opportunity: it rewards flexible players. Someone with a wide hero pool will always find a suitable pairing during the draft phase, while someone who plays only one hero will be isolated when teammates refuse to pair. In talent scouting terms, this means professional teams will prefer players who know many heroes over players who are great with one. That is a fundamental shift in talent evaluation, and I believe it will shape how academies train young players over the next few years.

Every time I stumble on the track, I hear another heartbeat fall into step with mine.

A thought worth considering

There is something I learned after fifteen years standing in the quiet corners of major sporting events: the true power of a system lies not in the number of its components, but in how it forces people to seek each other out.

Marvel Rivals has 106 pairings. That number can feel overwhelming, can make newcomers quit, can exhaust developers in tuning. But it is also a statement. It says that no one wins alone, that half of your power is decided by whether you find the right person to stand beside you.

I do not know whether this system will become the foundation of a lasting esport, or merely a clever retention mechanic of a live-service game. But I know that, on one September evening, I sat before two screens and asked myself a question sport has never stopped asking: how far can we go together.

The track and the pitch are not far apart; few people simply bother to run a full lap to see it.