- Apr 28
- 7 min read
Feedback Loops, Win Streaks & What's Next
By Aleco, Lead Game Designer at Leyline
Devblog | | | 28 April 2026 |
Introduction
Greetings! My name is Aleco and I'm the Lead Game Designer here at Leyline. In this dev blog I’m going to be walking through some of the thinking behind our most recent patch, as well as share some exciting plans for what's coming next.
The most recent patch shook up the game in a big way by making it harder and less powerful to assemble feedback loops, a gameplay pattern which had been the driving force behind all of the top scores on RR’s Endless Mode Leaderboards. Whenever I end up making a big or disruptive gameplay change to the game like this, I can’t help but feel like I owe an explanation of my thoughts and reasoning behind the changes to all of our dedicated players. If changes I make to the game are going to ask our players to throw out old strategies they’d spent time and effort to master, I think they deserve to understand why it was important to make those changes in the first place.
The second thing I'd like to discuss is player motivations. Our internal data and community feedback are both telling the same story: the most popular reason that players return to RR each week is to chase high scores on the Endless Mode leaderboards. That's something we genuinely celebrate, but at the same time we are planning to introduce new systems for players who are motivated by entirely different kinds of challenges, and I want to be as transparent as possible about what that means for RR going forward.
With that, let's dive in!
Feedback Loops
What is a feedback loop?
A feedback loop occurs when two or more effects interact with each other in a loop, each one continuously amplifying the other. The clearest example of this in recent patches was Yuuna's Nightshade specialization, which contained two components - let's call them A and B - that fed directly into each other:
Yuuna gains 1 Crit per 3 Poison on her target.
Whenever an Assassin Crits, they inflict 1 Poison per 20 Crit they have.
The result: every Crit inflicted more Poison (B), which granted more Crit (A), which inflicted even more Poison (B) and so on indefinitely. The vast majority of feedback loops in the game followed this same structure: one effect inflicts a Debuff on an enemy based on a Hero's stat, and a second effect increases that Hero's stat based on the number of Debuffs on enemies.
Are Feedback Loops Bad?
Not intrinsically, no. Feedback loops are an intentional and exciting part of RR's design, and I personally love the feeling of discovering and assembling them. The problem was never their existence; it was that they were far too easy to assemble and too powerful relative to other strategies once they got going.
Looking back at the old Nightshade design: both the A and B components of the loop existed within a single specialization. Players didn't need creativity, luck, or particularly deep game knowledge to set up a Poison/Crit loop with Yuuna, they simply had to select that one spec and the loop assembled itself. That frictionless path to a dominant strategy ran directly against the experience we want players to have of exploring and building synergies on their own.
This brings us to our first new standard for feedback loops in RR: no single piece of content should contain all the necessary components for a feedback loop. Assembling a loop should feel creative, clever, and perhaps a little fortunate. When it's too easy, that sense of discovery evaporates and in the long-term, the health of the game suffers for it.
Beyond ease of assembly, the other major issue was the speed and scale of power growth once a loop got started. Our top players were regularly assembling feedback loops that generated millions of stats within seconds of a fight beginning, numbers that are simply unreachable for any build not built such loops. When one particular strategy can reach numbers thousands of times larger than any other, it ceases to be one viable strategy among many and becomes the only viable strategy. For Endless Mode players in particular, this meant that runs increasingly felt like they were asking a single binary question: did you find the loop components quickly enough? If not, the run was effectively over before it had begun. A healthy metagame requires a diverse set of builds capable of competing at the top, otherwise runs stop feeling skill-testing and start feeling luck-dependent.
This brings us to our second new standard for feedback loops in RR: once assembled, feedback loops need to take an appropriate amount of time to ramp up. I still want feedback loops to produce impressive in-fight scaling and end-game power. But that power needs to be assembled and earned over time. When a loop can instantly and massively outpace every other scaling source in the game, it doesn't just make loops the best strategy; it makes everything else irrelevant.
To address this, I've moved away from effects which continuously update to instantly grant Heroes stats based on the current number of Debuffs on enemies. Take the old version of Funke's Stoked spec as an example:
Your Mages gain 1 Magic per 5 Burn and 1 Mana Regen per 15 Burn on enemies in their range.
With that design, a Magic → Burn → Magic loop could fire the instant any amount of Burn appeared on the enemy team. The updated Stoked spec instead grants your Mages a small, ever-growing amount of Magic each time an enemy takes Burn damage, an event that can trigger at most once per second, per enemy. The ramp-up is gradual and predictable rather than instant and exponential, and it eliminates the frustrating situation where your Heroes lose stats when an enemy dies. Crucially, it's still powerful and fun: as of writing, The Stoked is Funke's top-performing and most-played spec in our internal data.
I hope the reasoning here makes sense. If you're playing the game and discover something that seems to contradict these principles, I'd genuinely love to hear about it in our Discord. That kind of feedback goes a long way toward making each update better than the last.
Win Streaks & What's Coming Next
As I mentioned at the top, the data is unambiguous: the primary driver bringing players back to RR each week is the pursuit of high scores on the Endless Mode leaderboard. It's been genuinely inspiring to watch that community grow: the passionate Discord discussions, the enormous Steam playtime screenshots, the creative team-building that happens all across the board. Endless Mode has well exceeded our original expectations and has become a cornerstone of how we approach design decisions patch over patch.
Because of that, a large proportion of player feedback and community focus has naturally centered on Endless Mode. As RR continues to mature and our public demo release draws closer, we're also approaching a point where we'll be introducing new features and motivations to play beyond the Endless Mode leaderboards. I think it's important to be upfront about these upcoming additions, because some of the changes required to support them may, on the surface, feel at odds with how Endless Mode-focused players experience the game.
Introducing “Challenge Mode” and Win Streaks
The new feature I'm excited to announce today is Challenge Mode and a brand new Win Streak Leaderboard.
The current plan (subject to change, as always) is to introduce an optional set of challenges that players can opt into before a run begins on SSS difficulty. If all challenges are completed before the run ends, that run counts as a "win" for a new leaderboard which tracks the longest win streaks each patch. A loss, or a failure to complete all challenges, resets a player’s streak back to zero.
I’m incredibly excited for our players to get their hands on this new game mode and hope that it will serve a new audience of players who might prefer to play RR in a new way, where the core challenge of each run shifts from assembling the highest power cap team for Endless Mode to the challenge of pure survival. If you’re the kind of player who's motivated by a challenge, then this mode is for you! My initial goal for the level of difficulty is to have it be on par with the challenge level of Slay the Spire 1’s A20 + Heart rotating runs, where the current world record win streak is in the 20s.
What This Means for Endless Mode
Supporting Win Streaks will very likely require some adjustments to RR's overall difficulty, and I want to be candid with our Endless Mode community about that. Players who approach RR primarily through the lens of Endless Mode have, understandably, tended to react negatively to difficulty increases - something I can totally sympathize with. When your goal in every run is to assemble the highest-power team possible, additional difficulty works against that goal in direct and tangible ways.
Tuning a game to feel great for both Endless-focused players and Win-Streak-focused players is going to be far more art than science. Having spent time balancing games like Hearthstone in the past, I've learned more than once that changes which delight one audience can frustrate another. Every decision in that space requires a subjective judgment call aimed at maximizing overall enjoyment and minimizing overall frustration, and the relative popularity of each mode will be an important input into those calls. Given the current popularity of Endless Mode, I promise that it will continue to remain a top priority for us going forward and that I'll be working hard to ensure the Endless experience remains as fun and compelling as possible. In the near term, Endless Mode enthusiasts may find the game asks them to play a little less greedily and permanently-stat-scalingly than they're used to. But there's plenty of exciting content on the horizon for Endless Mode too, and I'm looking forward to sharing more of it soon.
As always, thank you for playing, for the feedback, and for your continued passion in Rift Reborn.
See you on the leaderboards,
Aleco