Marvel Rivals and 106 Team-Ups: The Balance Problem of a Forced-Synergy System

The LCK Summer 2026 final at Jamsil Arena was packed. I was twenty, an intern at an esports broadcast station, sitting in the top row of the technical area. Longzhu Gaming picked Jayce for Khan on the very first pick. A male commentator turned to me and laughed: "What does a girl know about lane dominance?" I pointed at the pick-and-ban board and said quietly: "Jayce with Kalista will force early fights. SKT will lose their way around minute twenty." I was right. Longzhu attacked relentlessly, controlled the game completely, and won three-one. That night, back in the dorm, I wrote the first line in my notebook: a single match can be a poem. It took me several more years to understand why I had been right. I was not reading the future. I was reading a pair. Top-tier esports, in the end, is the study of pairings: top with bottom, jungle with mid, the standard-bearer with the follower. Viewers remember individual plays, but analysts remember the pair that produced the play. Seven years later, in a different game, they no longer let players guess which pair is strong. They write the pair into the rules. Marvel Rivals is a six-versus-six hero shooter developed by NetEase, built on Marvel character licensing. Its roles split into three basics: vanguard, duelist, strategist. Its identity lies in a system rivals in the same genre do not have: the Team-Up ability. Season Ten just added a new character named The Hood along with his Team-Up kit, expanding one more node onto an already-vast synergy web. The way the system works is described clearly enough in the current guide. Every character in the game has Team-Up abilities, with no exceptions. Every character has exactly two Team-Ups. No character is launched without a Team-Up — that is the developer's design commitment. Team-Ups run on a two-layer mechanic: the base effect is always available to the host character, while the enhanced effect only activates when a teammate picks the correct partner character. At the time the guide was updated, the game had a total of one hundred and six Team-Ups. New heroes arrive roughly every month, a little faster or slower depending on the period. And the developer states that new heroes will team up with old ones, so that previously released characters do not lose value over time. Reading those numbers, the first thing I thought of was not a specific character but a structure. In an ordinary hero shooter, the question a player answers each match is: which hero is strongest right now. Marvel Rivals turns that question into another one: which pairing web is strongest under this patch. That is a shift from the problem of picking a single item to the problem of a graph. Each Team-Up is an edge connecting two characters. One hundred and six edges, with at least two more added every month when a new character arrives with its two Team-Ups. That is what caught my attention more than whether The Hood is strong or weak. Why does this structure matter? Because the balance burden does not grow in a straight line as the character count rises — it grows with the number of pairs. With more than a hundred Team-Up edges, every time the developer tunes a single Team-Up, the effect ripples across many compositions rather than stopping at one character. If any pair becomes too strong, it does not merely lift one character — it drags along the partner character, the role that pair replaces, and the other pairs pushed out of the optimal composition. Balance here is a combinatorial problem, not a linear one. That is why I rate this as the most meta-defining lever in the game. The two-layer mechanic is notable for another reason. Because the base effect is always available and the enhanced effect only activates with the partner character, the power of each pair is split in two: one half free, one half conditional. If this description matches the actual in-game mechanic, then heroes retain baseline value even when a teammate does not pick the partner character. This is a design choice that softens forced-synergy pressure but does not remove it. The pressure remains fully present at the enhanced layer. A player without a partner can still swim, but far more slowly. I noted this point and flagged it: it needs independent verification against official patch notes, because a guide's description may not perfectly match the real mechanic. The backward-compatible design intent is another signal worth weighing. The commitment that new heroes will team up with old ones protects the value of released characters and avoids power-creep-by-obsoletion. But it has a downside. Each new hero does not only strengthen itself; it silently raises the ceiling of an older character through a new Team-Up. In other words, a character released three years ago can be pushed to the top of the meta simply because a newly released character happens to team up with it. In a professional environment, this kind of indirect change is extremely hard to predict and further compresses the adaptation cycle for teams. Here I see a parallel with my own LCK memories. There were seasons when a small patch lifted a bottom-lane pair, and for the next two weeks every team had to relearn how to play the bottom lane. Marvel Rivals does this on a monthly cadence. A new-hero release every month or so keeps the meta almost never in a solved state. If I were still in the booth calling this discipline, I would have to rewrite my entire Team-Up analysis every thirty days. For someone in my trade, that is both an opportunity and a torment. One further consequence is rarely mentioned: encyclopedic knowledge becomes a competitive asset. One hundred and six Team-Ups sit beyond the range of reflexive human memory. No one can hold them all in their head entering a fast-paced match. The guide itself recommends bookmarking the list for reference. This means the game sets a high and rising knowledge barrier. Newcomers and returning players are pushed far away. Veterans and teams with analyst coaches hold the advantage. The knowledge map tilts toward those who prepare. But hold on. Before I nod along with the story that "strong pairings win," I want to argue against myself. All the meta-direction judgments I have just offered are structural, not data-backed. The source provides an inventory — which Team-Ups exist — but not a single performance metric: no win rate, no ban rate, no pick rate. That is the biggest blind spot in the entire analysis. I can say the structure creates risk, but I cannot say which pair is dominating, because I have no data. And anyone who claims otherwise without numbers is telling a story, not analyzing. There are nights I call out a match's name, and the stadium only echoes my own voice. This mechanics guide is the same. It says that mastering Team-Ups helps players climb the ranks, but offers no data sample to verify that advice. That is a claim reasonable in design terms and unproven in outcome terms. These two things are different, and a careful reader needs to distinguish them. The second point I want to dissect: the tendency to romanticize teamwork. In every team discipline, there is a beautiful story that unity beats individual talent. That story is usually told without data. The forced-synergy design in Marvel Rivals does not merely praise teammates; it makes teammates a mandatory condition to unlock half of a character's power. That is a fundamentally different mechanic from voluntary teamwork. When a character loses its enhanced effect for lack of a partner, the player is not making a free choice. They are being forced to pick. On a professional draft board, this kind of coercion can produce must-pick pairs that both teams must hold, and both teams know the opponent will hold. Game-design history shows that forced-synergy systems tend to produce such mandatory pairs. That is a structural risk, not a certain curse, but it exists. The third point: the nature of the source material itself. This article is evergreen content, written for long-term reference and search, not breaking news. It has value as an inventory, not as a performance analysis. The headline frame presenting it as a complete list encourages readers to trust it absolutely, while the number one hundred and six can drift from reality between updates. A continuously updated list can silently go stale. That risk is modest but real, and it differs from the game's balance risk. I say all this not to diminish the game. On the contrary. Marvel Rivals dares to bet on a system its competitors avoid because it is expensive to balance. That is design ambition. But design ambition and the ability to operate design are different things. A synergy web of one hundred and six edges, growing every month, puts the developer in a position of having to balance a surface wider than any tuning team can handle comfortably. This is the largest risk, and it is systemic, not located in any single Team-Up. I wonder what will happen to a professional circuit for this discipline. The patch says nothing about tournaments. The guide mentions no team, no coach, no player. So I cannot speak about a professional scene with data. I can only speak about preconditions. A discipline can only become sustainable esports when its balance surface is stable enough for fair play and followable for viewers. When the number of variables to track exceeds the memory capacity of both viewers and commentators, the value of the stage is tested. People cannot cheer for something they cannot see in time. Back to my memory. The first summer I believed I would live forever in that recording booth, I learned that a great match is not measured by the number of beautiful plays but by the number of layers of meaning a viewer can peel away in ten seconds. A teamfight only has value when someone understands in time why it happened. If a game pushes the number of layers too high — one hundred and six pairings, each with a base effect and an enhanced effect — it strips away the public's ability to understand. It becomes a subject for specialists. I am not saying that is wrong. I am only saying it needs to be recognized. Now, look at that structure from the perspective of someone in media. When a game has a Team-Up system this dense, the commentary profession changes in nature. The commentator no longer narrates who hits whom, but which pairing is unlocking which power. That is heavier work, but also richer. A good commentator can turn a teamfight into a story of two characters finding each other. A poor commentator just reads out ability names. The gap between these two people will widen, just as the gap between a knowledgeable player and an unknowledgeable one. On the players' side, I have a personal take. The forced-synergy mechanic implicitly rewards those with broad hero pools and punishes one-trick specialists. This is a talent-evaluation principle worth bringing into scouting, if this discipline moves toward professionalization. A player who can play many heroes will be a flexible node in the team's composition web, because they can pair with many different teammates. A one-hero player becomes a bottleneck. The source says nothing about players, but the structure speaks on their behalf. Three times I mispronounced a name, to learn that a title does not forgive carelessness. I still keep the habit of re-reading every number before going on air. One hundred and six Team-Ups is a beautiful, memorable, quotable number. That is exactly why it is a dangerous number. It becomes a symbol of the game's scale, and in the process of becoming a symbol, it can be repeated without anyone verifying it. A professional must separate symbol from fact. If that number goes wrong after a few patches, thousands of articles will go wrong together. I want to close the rebuttal with one positive thing that deserves acknowledgment. The idea of having new heroes team up with old ones is a way to keep the community attached along the vertical axis of time. It gives old players a reason to return to their favorite characters, rather than only chasing the newest hero. This is a kind of designed memory, and it is rare. Most competitive games survive by letting old characters slowly die. Marvel Rivals chooses otherwise, at least on paper. If executed well, this is a considerable humanist touch in an industry that often forgets old characters. But executing well is hard. Because every added Team-Up commitment adds a new balancing responsibility. The wider a system, the harder it is to keep a promise of fairness. This is the central paradox of Marvel Rivals at Season Ten. The game bets on richness as its identity, and that very richness is its heaviest burden. The match does not end when the stadium lights go out — it only changes who is listening. The question I leave is not which pairing is strong. The question is whether a web of one hundred and six edges, growing every month, can be kept stable enough to someday become a sport an outsider can understand in ten seconds. If the answer is yes, Marvel Rivals will not merely be a good game. It will be a new definition of how to tell a team story in esports. If the answer is no, it will forever be the playground of those who remember well — not of those who create.