Key takeaways
- Marvel Rivals runs split seasons, with each season divided into two halves that each carry their own content and rewards.
- A planned mid-season rank reset in Season 1 was cancelled after community pushback, and players kept their ranks into the second half.
- The split-season structure solves a real engagement problem but puts the competitive ladder and the content calendar in direct conflict.
Live service games have settled on a season as the unit of content. Three months, a battle pass, a reset, repeat. Marvel Rivals does something slightly different – it splits each season into halves, each with its own content drop and its own set of rewards, which means a meaningful update arrives roughly every six weeks instead of every twelve.
That sounds like strictly more content. It also creates a problem the developers ran into immediately in Season 1, and how they resolved it says more about live service design than the structure itself.
The split-season structure
A season is divided into two halves. Each half brings its own content and its own rewards, with the second half typically arriving with new characters, maps or balance passes rather than simply continuing the first.
The reasoning is straightforward. A three-month season has a well-documented engagement curve – a spike at launch, a long decline through the middle, and a small bump at the end as people finish the pass. Splitting it in two puts a second launch spike in the middle of what would otherwise be the trough.
For rewards it also means two completion targets instead of one. In Season 1, earning the second half’s rewards required completing ten competitive matches and meeting the relevant conditions before the season ended – a second, smaller commitment layered on the first.
This is a sensible response to a real pattern, and several games have arrived at it independently. Our piece on battle pass design covers the underlying engagement machinery that split seasons are tuning.
The rank reset that was not
Here is where the structure hit its problem.
Alongside the mid-season content drop, the developers planned a mid-season rank reset – competitive ranks would be adjusted downward at the halfway point, in the same way they are at a season boundary. The logic follows from the structure: if the second half is a new content cycle with its own rewards, it should have its own competitive arc too.
The community response was strongly negative, and the plan was scrapped before it shipped. Players kept their ranks and scores going into the second half, and the second-half rewards were tied to playing ten competitive matches rather than to re-climbing.
The announcement was unambiguous – there would be no rank reset.
Reversing a planned system before launch is unusual enough to be worth noting. It is also, in this case, a reasonable reversal, because the objection was substantive rather than reflexive.
Why a mid-season reset is worse than a seasonal one
The argument against it comes down to time, and it is the same argument in every game that has tried this.
Climbing takes most of a season. For a player who queues a few evenings a week, reaching a target rank is a six-to-ten-week project. A reset at the halfway point means the climb never finishes – you reset at the moment you arrive, every time.
Resets compress the ladder. Immediately after a reset, players of very different skill levels are at the same displayed rank, and matchmaking is at its least accurate. Doubling the number of resets doubles the amount of time spent in that bad window.
The reward structure does not need it. A second set of rewards can be gated behind playing matches, which is what Marvel Rivals ended up doing. Tying them to re-climbing adds nothing except the climb.
Street Fighter 6 reached a similar conclusion from a different direction – its phase resets were changed from a flat return to baseline into a scaled adjustment, precisely because resetting everyone to the same number produced bad matchmaking at the start of each phase. Our breakdown of that system covers how it works now.
The underlying conflict
Strip away the specifics and this is a structural tension that every live service game with a competitive mode has to resolve.
The content calendar wants short cycles. More frequent drops, more frequent reward resets, more frequent reasons to come back. Six weeks is better than twelve.
The competitive ladder wants long cycles. Enough time for ratings to converge on actual skill, enough time for a climb to feel like a project, enough time for matchmaking to settle.
These pull in opposite directions and there is no setting that satisfies both. Games resolve it in one of three ways:
- Decouple them. Content runs on its own calendar, ranked runs on a longer one. Cleanest solution, least used, because it means the ranked reset is not an engagement event.
- Soften the reset. Scale ranks towards the middle rather than resetting them, so a strong player starts strong. This is what Street Fighter 6 moved to.
- Gate rewards on participation, not rank. Play ten matches, get the reward. This is what Marvel Rivals ended up doing, and it is the simplest of the three.
The third option is underrated. It gives the content calendar its mid-season hook without touching the ladder at all, and the cost is that the reward is less prestigious. For a game whose competitive population is a minority of its total players, that is a cheap price.
What the reversal signals
Two things worth reading from it.
First, the competitive community has disproportionate influence on a hero shooter’s direction even when it is not the majority of the playerbase. That is consistent across the genre – our notes on hero shooter design and our Overwatch 2 system breakdown cover the same dynamic in the game that defined the category.
Second, a developer willing to cancel an announced system before it ships is usually a developer that has not yet committed engineering to making it permanent. The window for that kind of reversal closes fast – once a system is live and other systems depend on it, the cost of removing it stops being about community sentiment.
Does the split season work
On the content side, probably yes. A six-week drop cadence addresses a real engagement problem and the industry keeps converging on it independently, which is usually a sign that the underlying pressure is genuine rather than fashionable.
On the competitive side, the honest answer is that the split season does nothing for it and the right call is to leave the ladder out of the structure entirely. The Season 1 reversal landed in roughly the right place – content splits in two, the ladder does not, and the second half’s rewards are earned by playing rather than by re-climbing.
What remains unresolved is whether that stays true as the game ages. A ladder that resets once per season in a game that ships content twice per season will eventually look out of step with everything around it, and the pressure to synchronise them will not go away. Our notes on rotation and rework design cover the broader version of that problem, which is what happens when a live game’s systems are built on different clocks.
Images: Marvel Rivals project gallery.