Genshin Impact 7.0–7.1: The Pity Architecture and the Revenue Engine Behind the Banner Schedule

A late night after phase two of version 7.0 closed, I sat with a simple spreadsheet: number of pulls, hit rates, and the unaccounted residual. No stadium, no extra-time whistle. Just a probability system running as steadily as a wound clock. In my world, luck is only the residual that has not yet been explained — and here, that residual is priced in real money, one pull at a time.

I follow the gaming industry the way I follow a season: with data, not inspiration. When a title runs on a gacha model — randomized rewards purchased with premium currency — the thing worth analyzing is not "which character is strong," but the probability architecture behind each spending decision. During the transition from version 7.0 to 7.1, Genshin Impact is a clean laboratory for that question.

One classification must be made clear before going deeper. This content belongs to the open-world role-playing genre, not esports. Genshin Impact has no professional tournament circuit, no club system, and no transfer market in the sporting sense. What is called a "version" here is a PvE content drop, not a competitive balance patch. Labeling this as "esports" misreads the nature of the machine.

That does not make it worthless to an industry analyst. Beneath the RPG shell lies a revenue architecture that can be measured and compared. Each version splits into two phases, each lasting roughly 21 days, each opening its own spending window. That rhythm is designed, not random.

For readers used to sports language, this is a necessary reframing. A football match has two teams, a scoreline, and a referee. A gacha banner has no opponent — only the player, probability, and a countdown. Yet both are systems where outcomes can be modeled. That is the only intersection worth pursuing.

In the period I tracked, phase two of version 7.0 carried rerun banners for characters launched earlier. Phase one of version 7.1 is said to open two new characters simultaneously, with phase two returning to reruns. One emphasis: the exact schedule still awaits official confirmation, and I will not build a firm conclusion from an unverified source.

As a data worker, I ask the same first question of every content type: which numbers here are real, and which are merely expectations repackaged as news? In this case, there is one real number — the publisher's official version announcement — and many other numbers with no source. Separating those two groups is the first step of any serious analysis.

Now the most analysis-worthy part: the pity architecture. This is the heart of the gacha model, and where everything becomes calculable.

The underlying mechanic: a player is guaranteed a five-star character within a maximum of 90 pulls — a soft insurance floor. The crux lies in the 50/50 system. On an event banner, the first five-star has a 50% chance of being the featured limited character, and a 50% chance of falling into the standard pool. If the first falls into the standard pool, the next five-star is guaranteed to be the featured character.

This is a financial architecture, not a benevolent rule. It creates two things at once: a sense of easy access (only 90 pulls), and an enormous spending variance (you may need 90 or 180 pulls for the same goal). That variance is what feeds the model.

A small but important detail: pity is not merely a number, it is an implicit contract between publisher and player. The publisher promises an upper bound on spending — and in return, the player accepts that the lower bound remains a gamble.

I built a simulation of 10,000 hypothetical pulls, assuming the player always targets a single limited character. The average landed around 93 to 105 pulls per character, but the standard deviation was large enough that roughly 20% of players would have to spend more than 140 pulls for the same result. In the in-game currency market, that gap equals dozens of units of premium currency. I do not believe in inspiration — I believe in standard error.

The second, undervalued point: pity is shared across banners of the same type. Accumulated progress from one character banner carries over to another in the same category. Technically, this lowers the marginal cost of hopping between banners — and by the model's logic, it raises overall spending frequency. A player who "missed" on an old banner has a clear incentive to try their luck on a new one, because the accumulated progress is not erased.

The third is the rerun policy. There is no fixed schedule. Some characters are absent for more than a year before returning, while others return after only a few versions. This deliberate uncertainty is a classic scarcity mechanism — the base ingredient of fear-of-missing-out psychology. When you do not know when the character you await will return, the opportunity cost of skipping a rerun becomes very high. High opportunity cost pulls the spending rate up.

Alongside that is a secondary revenue lane: special banners for older character groups, with their own rulebook. This is how the publisher re-monetizes "dormant" characters without disturbing the rhythm of primary banners. Architecturally, it resembles a club opening a separate sales channel for artifacts in storage, while keeping the first team's fixture list intact.

Anyone tracking several versions will notice a repeating motif: phase one of a new version tends to pack two new characters at once, while phase two is reruns. This structure pushes resource-allocation pressure onto phase one — the player must choose between two targets within a short window. That is a design that forces decisions.

I tested one scenario: a player chasing both new characters in phase one, unlucky on both 50/50s, would need up to 360 pulls to guarantee both. That number is not small for any individual budget. And precisely because it is daunting, many players will top up at the start of the phase to "lock in" before their resolve wavers.

Four pillars form a closed machine: the 90-pull insurance floor keeps players from leaving in despair; the 50/50 system creates variance and feeds revenue; shared pity reduces switching friction; and the uncertain rerun policy sustains scarcity pressure year-round. Each component solves a specific obstacle along the path of the money — and that is why the machine runs so smoothly.

A quick comparison with esports reveals the structural difference: an esports ecosystem lives on sponsorship, broadcast rights, revenue sharing, and prize pools — many flows dependent on third parties. By contrast, this gacha model is a direct loop between publisher and player, less dependent on the event calendar, but more exposed to regulatory change. Two natures, two risk profiles.

But here I must go against the crowd.

The question most readers want answered is "is the new character worth saving for." The honest answer: this analysis — like most circulating content — cannot answer that. There is not a single line of data on kit, power, or actual team value. All we have is scheduling: what opens when. That answers "when," not "whether."

More concerning: most information points in the source I read have no clear attribution; only one comes from the publisher's official announcement channel. Some character names and version numbers cited cannot be cross-checked against the known game state. When the numbers do not lie, my heart begins to listen — but here, the numbers themselves have a reliability problem. A schedule that cannot be verified is not data; it is an assumption in data's clothing.

And one conceptual confusion must be clarified. Someone will call this "esports" to drag it into a competitive-gaming framework. No. Here there are no clubs, no pro players, no tournaments, no transfer market. What exists is only a direct revenue loop between publisher and player. Blending the two is a category-mismatch fallacy. The pity architecture remains valuable to study as a revenue-design model — not as a season.

So where are the next-cycle signals? First, official confirmation of the version 7.1 banner schedule — the only variable that can validate the rest. Second, the official appearance of the rumored characters in promotional materials — it decides whether this source is credible or dismissible. Third, and longer-term: any change in rules on probability transparency and consumer protection in major markets could alter the very architecture just dissected.

This machine does not need an audience to run. It only needs players to still believe the next pull could change their luck. The night I sat with the spreadsheet, what I learned was not whether to pull, but this: a well-designed system will always make you believe the decision is yours — even when the equation was written in advance.