Genshin Impact: Banner Architecture, Pity Mechanics and the Recurring Revenue Engine Ahead of Version 7.1

Genshin Impact: Banner Architecture, Pity Mechanics and the Recurring Revenue Engine Ahead of Version 7.1

A Wednesday evening in Seoul

On a Wednesday evening, the corner cafe on Teheran-ro was as loud as ever. Two people at the next table were arguing about something I understood completely after half a sentence. One said: "I'm not pulling anything in 7.0 phase two, I'm saving everything for 7.1." The other replied: "But are you sure Vesna and Vodyanitsa are even real?"

I took off my headphones. Over more than a decade covering sport and entertainment, I have heard this story in many forms. At Incheon airport, a supporter told me he would skip a Champions League return-leg ticket because he was "saving for the quarter-final". At an open training session at a K League club, a scout said his side would sign no short-term deals in the winter window because "better players will be available in summer". Both times, belief in an unconfirmed future beat a confirmed present.

The difference this time is that players are weighing a decision on information with no clear source.

The central thesis of this piece: Genshin Impact's banner system operates like a financial contract in which the rule-maker, the seller and the information publisher are the same entity. That characteristic matters more than any pity threshold the community passes around.

Before the analysis, I have to state one thing about this article itself, because I read my drafts aloud before publishing, and this time, reading it aloud, I heard a very large gap.

Context: a content cycle packaged into two purchase windows

Genshin Impact has no official professional tournament circuit, no franchised club system, no transfer market in the esports sense, and no competitive balance patch system in the league sense. That needs saying plainly, because a great deal of content about this game is mechanically labelled "esports" when its nature is an open-world action role-playing game running on a gacha model.

The game's operating rhythm is divided into versions. Each version splits into two phases, each lasting roughly 21 days. Each phase has its own banner. That is the entire "format" of this game — a content-release cadence, not a competition format.

According to the material currently in circulation, version 7.0 is in its second phase with reruns for Flins and Ineffa. Version 7.1 is said to open with two simultaneous debuts, Vesna and Vodyanitsa, followed by a phase two rerun for Skirk and Escoffier. In the wider picture, names such as Odette, Aino, Iansan and Lan Yan also appear scattered across lists passed hand to hand by the community.

Exactly one item in this entire information set is tied to an official source: the publisher's own announcement about upcoming content. Everything else — character lists, phase order, version names — carries no source or is explicitly the author's opinion.

And even the sourced item concedes that the exact banner schedule is still to be confirmed.

The emotional context also needs putting on the table. The phrasing "an exciting adventure in Snezhnaya" carries clear promotional weight. For a game running on gacha, opening a large region typically coincides with elevated spending pressure and inflated banner value. That is an observation about revenue architecture.

Where the player sits in the banner

To analyse a system with no players, no coaches and no rosters, the unit of analysis must change. The unit here is the player and the premium currency they are holding.

Based on my experience watching matches, I always start by identifying who holds the advantage and who holds the risk. In this frame, the advantaged group is players who have accumulated premium currency before entering 7.1, plus those waiting on reruns for Skirk or Escoffier.

The losing group is players who poured currency into 7.0 phase two for Flins and Ineffa and then realised they lack resources for two simultaneous debuts in 7.1 phase one.

The point to stress: there is no data at all on strength, kit or optimality for any of these characters in the entire information set. Only schedule data exists.

That is why I call this a schedule explainer, not a decision-support document. Readers get an answer to "when", and no answer to "should I".

Pity: a contract with no third-party auditor

The most analytically interesting mechanism in the whole system sits here.

Players are guaranteed a five-star character within 90 pulls. On an event banner, the first five-star has a 50% chance of being the featured character and a 50% chance of a standard character. If the first lands on standard, the next five-star is guaranteed to be featured.

The community calls this pity. Technically, it is a guarantee threshold combined with a two-tier probability distribution.

What matters is not the number 90, but that this guarantee threshold and probability distribution are defined, published and operated by the publisher itself. There is no third-party audit, no independent arbitration mechanism, and no way for players to verify actual rates beyond trusting the disclosure.

In sport, this model has a structural parallel: competitions run by an entity that organises, owns the rights and publishes the rules. That concentration typically brings transparency disputes. Here the concentration is even higher, because the paying party and the informed party are the same group.

From a design perspective, the 90-pull threshold does two things at once. It creates a sense of accessibility for a rare reward, while sustaining high spending variance. That is a classic design in randomised-reward revenue models.

50/50 and the design of variance

The 50/50 mechanism is the most sophisticated part of the system, and I think it is under-analysed.

With only a guarantee threshold, players would know exactly how much they need to reach a target. Cost becomes a deterministic function. The 50/50 mechanism breaks that determinism by inserting a coin flip before the guarantee applies.

The result is that the actual cost of acquiring a featured character becomes a variable that can roughly double between the favourable and unfavourable case. Players who win the first flip spend far less than those who must ride the guarantee twice.

In behavioural finance, this is a structure that increases suspense and engagement. In revenue design, it is a structure that increases income variance without reducing expected income.

Notably, the guarantee on the next five-star softens the sense of loss. Players who land on the bad branch still receive a firm promise for next time. That promise does not reduce current cost, but it keeps the player inside the system.

I once wrote that a transfer is not real until someone agrees to tell it as a fate. Here too: a character only becomes "worth pulling" when someone can tell the story of their coin flip.

Shared pity: less friction, more frequency

Another technical detail gets little attention: the pity counter is shared across banners of the same category.

The practical meaning is concrete. When a player moves from a debut banner to a rerun banner, their guarantee progress is not wiped. In other words, friction when switching between spending windows is reduced.

From a revenue standpoint, this is a cash-flow smoothing mechanism. It limits situations where a player is stranded between two targets and abandons spending altogether. Instead, it encourages spending in both debut and rerun windows.

The sporting parallel: this resembles a league allowing points to carry across stages rather than resetting, which encourages continuous rather than episodic following.

No fixed rerun schedule: scarcity as a tool

The system does not run on a fixed rerun schedule. Some characters are absent for over a year. Others return within a few versions. No public rule explains the order.

For players, this is the source of most anxiety. For design, it is a controlled scarcity mechanism.

When nobody knows when the next target will appear, the opportunity cost of skipping the current window rises. Players no longer compare two known options; they compare a known option against an undefined risk.

In competition management the same principle appears as partial fixture release. But in sport, schedules are usually published for format and stakeholder reasons. In the gacha model, no participating party needs that protection, so scarcity can be operated more freely.

Chronicled Wish: a second revenue lane

A separate mechanism exists alongside the main banners, running under its own rules and typically aimed at older characters.

Architecturally, this is a second revenue lane. It lets the publisher re-monetise long-released characters without placing them on the main rerun schedule.

This solves a specific problem: if every older character had to rerun on main banners, the calendar would jam and new characters would have to wait. The second lane keeps the debut cadence unbroken while sustaining earnings from the back catalogue.

For players, the consequence is more spending windows per year. For the publisher, the consequence is deeper exploitation of the catalogue.

Comparison with the esports ecosystem

This is the section I want to give the most room, because I believe mis-comparing revenue models is the source of much imprecise analysis.

A professional esports ecosystem runs on a multi-layer value chain. A publisher makes the game. Tournaments create the arena. Teams create competitive content. Broadcast platforms create distribution. Sponsors create cash flow. Fans create attention, and that attention converts to revenue across many touchpoints.

The gacha model bypasses nearly that entire chain. It runs on a two-layer loop: the publisher announces content, players spend directly in game. There is no middle layer.

Consequence one is very high marginal efficiency. No tournament operation cost, no team operation cost, no broadcast rights cost.

Consequence two is equally high dependence on a single entity. The publisher controls content supply, price and information.

Consequence three concerns resilience to calendar shocks. An esports ecosystem depends on schedules, international travel and audience presence. The gacha model depends on none of these. It runs when every stadium is shut.

That is why I once said a summer without spectators still had us rehearsing for an imagined audience. In the gacha model, the audience does not need to imagine anything, because they never left.

The counter-risk is clear. A system dependent on probability-disclosure and consumer-protection rules is directly exposed when those rules change. An esports ecosystem has more supports to diversify risk.

The contrarian angle: the problem is the source, not the schedule

This is where I argue against myself.

Across the information set I am analysing, there are 28 data points. Twenty carry no source. Only one is tied to an official source. Three are explicitly the author's opinion.

Many named entities cannot be cross-verified against known game state. That means a substantial share of the content may be rumour or generated without verification.

Where could I be wrong?

First, the character names and version numbers may be accurate and simply beyond my current knowledge. If so, my schedule analysis needs updating, while my revenue-architecture analysis stands.

Second, this may be a traffic-oriented media product in which factual accuracy is not the top priority. If so, its analytical value lies in showing how a community consumes information, not in the specific content.

Third, I may be applying an industrial framework to an entertainment product and thereby ignoring experiential factors players actually care about.

But on any branch, one conclusion holds: readers are making spending decisions on an information set that is largely unsourced, and the most authoritative source itself concedes the schedule is unconfirmed.

That is the real risk in this problem. It is not about which character is stronger.

The blind spot: a schedule is not a value

A common analytical error is equating a character's early appearance with a character's worth investing in.

In sport there is a version of this error: equating heavy transfer spending with future success. History shows that correlation is far weaker than the crowd feels.

In the gacha frame, two simultaneous debuts in 7.1 phase one says nothing about their strength. It only says currency-allocation pressure in that phase is higher.

Conversely, a character appearing in a rerun phase says nothing about their value either. It only says the publisher decided when to re-monetise.

That is why I always separate two questions. The first is "when". The second is "is it worth it". A schedule explainer answers the first and leaves the second blank. Readers tend to fill the second with feeling.

I was once a laughing stock for my pronunciation, and now I am the voice they choose every evening. The lesson from that period was not pronouncing names correctly. It was being forced to separate what I know from what I want to believe. In a market of unsourced information, that is the only skill that truly protects the reader.

What to watch

The official announcement of the 7.1 banners is the single most important signal. When the publisher publishes the phase-by-phase character list, every prediction above is confirmed or refuted.

Cross-checking the character names in the current information set against official archives will establish the reliability of the entire body of content.

Developments in probability-disclosure and consumer-protection rules will directly affect the revenue architecture described above.

And if the real goal is esports industry analysis, re-classifying this content into a games-and-revenue-models category will let the downstream analytical framework work correctly.

A thought to take away

I do not believe a banner schedule can be built on unconfirmed names. But I do believe the architecture behind that schedule — the guarantee threshold, the two-tier probability distribution, the shared counter, the unfixed rerun calendar and a secondary revenue lane for the back catalogue — is one of the most complete revenue designs interactive entertainment has produced this decade.

And like every good design, it does not need players to understand it in order to operate.

The only thing players control is the question they ask themselves before opening a banner. If the question is "when", they will always get an answer. If the question is "is it worth it", they must answer it themselves — and that is the moment real decision-making returns to their hands.

The widest stadium is not the crowded one, but the one where people agree to listen. In a system where the information publisher is also the one collecting the money, listening carefully becomes the cheapest and strongest skill anyone can train.