Short answer: no, not in the way “pay-to-win” usually gets used. Genshin Impact has no ranked PvP ladder and no leaderboard where a bigger wallet buys a higher rank. But spending real money does buy something concrete: faster, more reliable access to five-star characters that clear the game’s hardest PvE content with less grinding and fewer wasted resources. Whether that counts as “winning” depends on what you’re trying to win. This piece breaks down the actual math, using HoYoverse’s own published drop rates and the game’s current 2026 shop pricing, so you can decide for yourself whether a given pull is worth the money.
How Genshin Impact’s gacha system actually works
Genshin Impact runs on a wish system. Every pull on a limited-time Character Event Wish banner costs 160 Primogems, the game’s free-to-earn currency, or an equal amount of Genesis Crystals, the currency you buy with real money. HoYoverse discloses the odds for every wish type directly in the client, and those numbers are reproduced consistently across the Genshin Impact Fandom wiki’s Character Event Wish page and third-party trackers.
The base rate for pulling a five-star character on a Character Event Wish banner is 0.6% per pull. That rate holds flat from your first pull through your 73rd pull on a given pity counter. Miss for that long and the system starts compensating you, a mechanic known as soft pity.
Soft pity vs hard pity: the curve behind the guarantee
HoYoverse doesn’t publish the exact soft-pity percentages in a single official table, but the shape of the curve is well documented by outlets like Eurogamer and Game8, and confirmed repeatedly through years of aggregated pull data from the community. Odds stay at 0.6% through pull 73, then climb by roughly six percentage points per pull starting around pull 74, until the system guarantees a five-star outright.
That guarantee, hard pity, lands at your 90th wish. If you’ve pulled 89 times on a banner without a five-star, your very next pull is locked at 100%. This is the one number HoYoverse confirms directly in the client and it hasn’t moved since the game’s 2020 launch.
Put the curve into a cumulative view and the picture gets clearer. Most players who go the distance don’t actually see their five-star at pull 90. They see it in the soft-pity window, somewhere between pull 75 and pull 85, because that’s where the per-pull odds spike hardest.
The 50/50 rule and what happens when you lose it
Pulling a five-star doesn’t automatically mean pulling the character on the banner art. Every Character Event Wish banner runs what the community calls the 50/50: when you land a five-star, you have a 50% chance it’s the featured limited character and a 50% chance it’s one of the fixed standard-pool five-stars (Diluc, Jean, Mona, Qiqi, Keqing, Tighnari, or Dehya, depending on version).
Lose that coin flip and the game doesn’t just shrug. Your next five-star on any Character Event Wish banner is guaranteed to be the featured character, no matter which banner it drops on. That guarantee carries over between banner rotations until you cash it in, then resets. It’s a real mechanic that HoYoverse’s in-game wish description spells out, and it’s the single biggest reason average pull counts to a specific character run higher than the 90-pull hard pity number most players quote from memory.
The worked pity-cost table
Here’s what the math looks like when you convert pull count into cumulative odds, Primogems spent, and real dollars. USD figures below use the best per-dollar Genesis Crystal rate available in the shop, the $99.99 tier, which nets 8,080 crystals after its recurring purchase bonus.
| Pulls made | Cumulative odds of any 5-star | Primogems spent | USD cost (best-value pack) |
|---|---|---|---|
| 10 | 5.8% | 1,600 | $19.80 |
| 30 | 16.5% | 4,800 | $59.40 |
| 50 | 26.0% | 8,000 | $99.00 |
| 73 (soft pity starts) | 35.6% | 11,680 | $144.54 |
| 80 | 91.8% | 12,800 | $158.40 |
| 85 | 99.9% | 13,600 | $168.30 |
| 90 (hard pity) | 100.0% | 14,400 | $178.20 |
| 93 (expected pulls to featured, see below) | n/a | 14,951 | $185.02 |
| 180 (worst case, double hard pity) | 100.0% | 28,800 | $356.40 |
The jump between pull 73 and pull 80 is the whole story of soft pity in one row: cumulative odds go from 35.6% to 91.8% in seven pulls, because each individual pull’s rate is climbing roughly six percentage points at a time. That’s the mechanic doing its job, capping how unlucky you can realistically get.
Step-by-step: calculating expected value to the featured character
Expected value, or EV, answers a different question than “what’s the max I’ll ever need.” It answers “what should I budget for, on average, across many attempts.” Here’s the calculation, broken into steps using only HoYoverse’s published rates.
- Step 1: model the per-pull hazard rate. Odds are 0.6% flat for pulls 1 through 73, then rise roughly 6 percentage points per pull from pull 74 through pull 89, hitting 100% at pull 90.
- Step 2: sum the survival probabilities. Expected pulls to any five-star equals the sum, from pull 0 to pull 89, of the probability you still haven’t pulled one by that point. Running that sum against the curve above lands at approximately 62 pulls to any five-star, which matches independently published community modeling of the same curve.
- Step 3: layer in the 50/50. Half the time you win the coin flip and get the featured character in that same ~62-pull run. Half the time you lose it, spend roughly 62 pulls landing a standard five-star, then spend another guaranteed run (averaging roughly 62 more pulls, capped by hard pity) to land the featured character on the next attempt.
- Step 4: average the two branches. EV = 0.5 × (62 pulls) + 0.5 × (2 × 62 pulls) = 1.5 × 62 ≈ 93 pulls to guarantee the specific character you actually want.
- Step 5: convert to currency. 93 pulls × 160 Primogems = 14,880 Primogems (14,951 in the precise, non-rounded calculation). At the best cash rate in the shop, that’s roughly $185. At the worst single-pack rate ($0.99 tier), it’s closer to $247.
EV(pulls to featured 5-star) = 0.5 × E[any 5★] + 0.5 × 2 × E[any 5★]
= 1.5 × E[any 5★]
E[any 5★] ≈ 62 pulls (derived from HoYoverse's published 0.6% base rate + soft/hard pity curve)
EV ≈ 93 pulls × 160 Primogems/pull = ~14,950 Primogems ≈ $185 (best-value pack)That $185 figure is an average, not a promise. Roughly half of all attempts will land the featured character for less. The other half will run past it, and a real minority of players will hit the true worst case: losing the 50/50 at pull 90, then needing all 90 pulls again on the next attempt.
What one wish actually costs across every pack tier
Genesis Crystals are sold in six fixed tiers, and the per-crystal rate gets better as the tier size grows, a standard mobile-monetization pattern. Here’s the full shop breakdown as currently listed, including the recurring bonus crystals each pack adds beyond its base amount.
| Pack price | Base crystals | Bonus crystals | Total crystals | Cost per single wish |
|---|---|---|---|---|
| $0.99 | 60 | 0 | 60 | $2.64 |
| $4.99 | 300 | 30 | 330 | $2.42 |
| $14.99 | 980 | 110 | 1,090 | $2.20 |
| $29.99 | 1,980 | 260 | 2,240 | $2.14 |
| $49.99 | 3,280 | 600 | 3,880 | $2.06 |
| $99.99 | 6,480 | 1,600 | 8,080 | $1.98 |
There’s also the Blessing of the Welkin Moon, a $4.99 monthly pass that pays out 300 Genesis Crystals up front plus 90 Primogems a day for 30 days, roughly 3,000 Primogems total for the month. That works out to under 28 cents per wish, by far the cheapest way to convert cash into pulls in the entire game. The catch is time: Welkin only pays out over a full month of daily logins, so it doesn’t help if you need pulls before a banner rotates out.
Worked example 1: the lucky day-one pull
A player with zero saved Primogems buys a single $0.99 pack for 60 Genesis Crystals, converts it, and needs 160 total to pull once. Add a bit of free income and they pull on day one. At 0.6% base odds, that’s a 1-in-166 shot at a five-star, and only a 1-in-333 shot it’s the featured character specifically. It happens to some players every version. It’s not the number to plan around.
Worked example 2: hitting soft pity on a $150 budget
A player buys two $49.99 packs and one $49.99 top-up ($150 total, ignoring tax) for roughly 11,640 crystals, enough for 72 wishes, just short of the soft-pity threshold at pull 73. Cumulative odds of any five-star at that point sit at 35.6%. That means nearly two out of three players spending exactly this amount walk away with no five-star at all and have to decide whether to keep going into the soft-pity zone or stop.
Worked example 3: the average-case spender chasing the EV
Following the step-by-step calculation above, a player targeting the statistical average spends roughly 93 pulls, or about $185 at the best per-dollar pack rate, to land the specific featured character. That’s the number that should anchor any “is this banner worth it” decision, because it already accounts for the 50/50 coin flip, not just the raw pity cap.
Worked example 4: the worst-case double-pity run
A player loses the 50/50 at pull 90 (drawing a standard five-star instead of the featured one), then needs the full pity cycle again to lock in the guaranteed featured pull on their next five-star. That’s 180 total pulls, 28,800 Primogems, and $356.40 at the best available shop rate, or $475.20 buying only the smallest $0.99 packs. This is the true statistical ceiling. HoYoverse’s own guarantee mechanic means no player can be forced past it for a single featured character, no matter how unlucky the RNG runs.
Worked example 5: comparing the EV cost to real whale spending
Industry data on gacha-style mobile RPGs, compiled by VoxBooster from Adjust and Sensor Tower figures, puts average monthly spend among paying gacha players at roughly $85, with the top 2% of spenders generating about 72% of total gacha revenue across the genre. That figure isn’t Genshin-specific, but it’s a useful yardstick: a single month of average whale-tier spending in this genre covers the $185 average-case EV to a featured character more than four times over, or the full $356 worst-case double-pity scenario more than twice over. For a genre where the median player spends nothing, per Unity’s 2025 monetization data cited by industry trackers, the gap between what a whale spends and what a single character actually costs on average is enormous.
Where the money actually matters: PvE, not PvP
Here’s where the “pay-to-win” label falls apart under scrutiny. Genshin Impact has no ranked player-vs-player mode built around its main combat system. Spiral Abyss and Imaginarium Theater, the game’s two rotating endgame challenges, are both PvE content with personal star-based rewards and no cross-player leaderboard built into the client. The only actual PvP in the game lives in the Genius Invocation TCG, a separate turn-based card mode that runs on its own card-collection system, not the character gacha.
So spending money buys speed and comfort in solo content, not rank against other players. A free-to-play account with well-built four-star characters can clear Spiral Abyss’s hardest floors. It just takes longer to figure out and more careful team building. A whale account with every meta five-star can often brute-force the same clear with less optimization. Both outcomes exist in the same game, and neither one beats the other in any formal sense, because there’s nothing to beat.
Myths vs facts about Genshin’s pay-to-win reputation
| Myth | Fact |
|---|---|
| You need to spend money to beat the story or open-world content | All main story quests, exploration, and world bosses are completable with free-to-play characters and standard-banner pulls |
| Whales get a competitive rank advantage | There is no PvP ladder tied to the character gacha. Spiral Abyss and Imaginarium Theater are solo PvE with no cross-player leaderboard |
| The odds are secret or unverifiable | HoYoverse discloses the 0.6% base rate and 90-pull hard pity directly in the client’s wish details, reproduced by the Fandom wiki and outlets like Eurogamer |
| Losing the 50/50 means you’re stuck with the wrong character forever | Losing guarantees your very next five-star on any Character Event Wish banner is the featured one |
| Every player who reaches pull 90 spent $178+ | Most players who go the distance land their five-star in the soft-pity window (roughly pull 75-85), spending noticeably less than the full hard-pity amount |
What the revenue numbers say about how the game actually gets paid for
Genshin Impact’s lifetime player spending crossed $10 billion by the end of 2025, according to research compiled by ShaneTheGamer and industry analyst Niko Partners, making it one of the fastest games in history to reach that mark. Mobile-only annual revenue has declined significantly from its 2020-2021 launch peak, with AppMagic-sourced figures cited by PocketGamer.biz showing a roughly 42% year-over-year drop in one recent tracked period, alongside broader industry estimates placing 2025 mobile revenue somewhere in the $319 million to $505 million range depending on methodology and platform scope.
Sensor Tower’s own two-year mobile revenue analysis pegged the game at $3.7 billion in its first 24 months alone, an early indicator of just how concentrated the spending was in Genshin’s launch window. What that spending buys, per the mechanics broken down above, isn’t rank. It’s speed, and the option to skip the grind that free-to-play players absorb through time instead of money.
How this compares to other HoYoverse titles
The 0.6% base rate, soft pity ramp, and 90-pull hard pity aren’t unique to Genshin. HoYoverse runs the same core pity structure across Honkai: Star Rail and Zenless Zone Zero, its other live gacha titles, with only minor variations in currency names and banner cadence. That consistency matters for the “is it worth pulling” question: if you understand the math in one HoYoverse game, the same EV logic (base rate, soft-pity ramp, hard-pity cap, 50/50 layer) carries over almost exactly to the others.
Frequently asked questions
Is Genshin Impact pay-to-win?
Not in the competitive sense. There’s no ranked PvP mode tied to the character gacha, so spending doesn’t buy a rank over other players. It does buy faster, more reliable access to strong characters in PvE content, which some players consider a meaningful advantage even without a leaderboard attached to it.
How much does it cost on average to guarantee a featured 5-star character?
Working through HoYoverse’s published 0.6% base rate, the soft/hard pity curve, and the 50/50 mechanic, the expected value lands around 93 pulls, roughly 14,950 Primogems, or about $185 at the best available Genesis Crystal pack rate.
What’s the absolute worst-case cost for one featured character?
Losing the 50/50 at hard pity (pull 90) and needing the full 90-pull cycle again caps out at 180 pulls, 28,800 Primogems, or roughly $356 at the best shop rate. HoYoverse’s guarantee mechanic means no player pays more than this for one specific featured character.
Can free-to-play players clear the hardest content in Genshin Impact?
Yes. Spiral Abyss and Imaginarium Theater are both solo PvE challenges with no cross-player leaderboard, and both are widely documented as clearable with well-built four-star teams pulled from the free standard banner. It takes more optimization and game knowledge, not necessarily money.
What is the cheapest way to buy Primogems in Genshin Impact?
The Blessing of the Welkin Moon monthly pass, at $4.99 for roughly 3,000 Primogems spread across 30 days, works out to under 28 cents per wish, far cheaper per-Primogem than any direct Genesis Crystal purchase. The tradeoff is that the payout is spread over a full month rather than available instantly.
Does losing the 50/50 in Genshin Impact ever happen twice in a row?
It’s mathematically possible but rare, since the guarantee only applies to your very next five-star pull after a loss, not the one after that. Once you win a guaranteed pull, the 50/50 resets for your following five-star, so consecutive losses beyond the guaranteed pull are governed by the same 50% odds each time.
Are Genshin Impact’s drop rates independently verified or just self-reported?
The rates are disclosed directly by HoYoverse inside the client’s wish details screen, not independently audited by a third party in the way casino RNG is sometimes certified. Community data-mining and years of aggregated pull tracking from large player samples have consistently matched the disclosed base rate and pity thresholds, which is the main reason the numbers are treated as reliable across gaming outlets.
