Granblue Fantasy does not use the rising “soft pity” curve that drives most modern gacha games, and it has no 50/50 coin-flip either. The SSR rate stays flat at 3% on a standard banner and 6% during a Gala or Flash Gala rate-up, pull after pull, with no escalating odds as a player keeps missing. The only mechanic that guarantees anything is the Spark: every single draw banks one Cerulean Spark, and at exactly 300 Sparks a player can exchange them for any one item, including the newest limited SSR, from that banner’s list. That means the fully guaranteed path to a specific featured character costs 300 draws, which is 90,000 Crystals, which gacha-mechanics trackers put at roughly 800 dollars in real cash. Everything short of that is a flat probability roll where the odds never move.
Does Granblue Fantasy Have Pity? The Short Answer
If you searched for granblue fantasy pity odds hoping for a simple percentage, here it is: a normal premium draw lands an SSR 3% of the time, and that number doubles to 6% during a Gala or Flash Gala banner, according to the gacha-mechanics breakdown published by Game Design Lab’s case study on top gacha games. There is no soft pity ramp sitting underneath that number. The rate on draw 1 is the same as the rate on draw 299.
The one true guarantee is the Spark system. Cygames calls the currency a Cerulean Spark, and every draw deposits exactly one of them regardless of whether you pull a Rare, Super Rare, or Super Special Rare. Reach 300 Sparks during an active banner and you get to pick any item from that banner’s exchange list for free, no roll required. For a player chasing a specific limited character, that 300-draw threshold is the only point where luck stops mattering entirely.
That single design choice separates Granblue Fantasy from the escalating-pity model used in many newer gacha titles. Instead of nudging the odds upward as a player gets unlucky, Cygames keeps the odds frozen and instead hands out a slow, parallel currency that eventually buys certainty outright. Both approaches guarantee a result eventually. They just get there in very different ways, and that difference changes how a player should budget their spending.
How the Spark System Actually Works
A single draw in Granblue Fantasy costs 300 Crystals, the game’s premium currency. A 10-part draw costs 3,000 Crystals, the same rate per pull with no bulk discount, per the official draw mechanics listed on the Granblue Fantasy Wiki’s Draw page. Every one of those draws, win or lose, adds one Cerulean Spark to a running counter tied to that specific promotional period.
Hit 300 Sparks and the exchange list unlocks. You choose, you don’t roll. The catch that trips up new players is timing: the Sparks are tied to the active draw period, and they expire once that promotion ends. Carrying a partial Spark count from a banner that closed last month into a brand-new one is not how the system works. Each banner effectively runs its own 300-draw countdown, so partial progress on an old, expired Gala doesn’t carry forward.
There is a secondary, smaller guarantee that often gets confused with pity. On a standard 10-part draw, the game floors the tenth slot at Super Rare or better if none of the first nine pulls produced an SSR. That protects you from an all-Rare 10-pull, but it does not guarantee an SSR, let alone the specific SSR you want. It is a rarity floor, not a drop-rate pity, and conflating the two is one of the more common misunderstandings about this game’s economy.
The Real SSR Base Rate: 3% Normal, 6% During Gala
The 3% baseline applies on any ordinary premium draw, any day the game isn’t running a special promotion. During a Legend Festival, commonly shortened to Legfest, or during a Flash Gala, that rate doubles to 6%. These Gala periods run roughly every couple of weeks rather than being a rare, once-a-year event, which is precisely why experienced players save their Crystals and wait rather than pulling on a normal banner.
What the 6% figure does not tell you is which specific unit you’ll get if you land an SSR. During a Gala, the featured new characters sit inside the same SSR pool as the game’s existing SSR catalog, and Cygames has not published a separate, isolated sub-rate for “this exact new unit versus every other SSR in the pool.” That is a meaningful difference from games that publish a clean 50/50 split between a featured unit and the standard pool. In Granblue Fantasy, hitting the 6% doesn’t tell you who you got until the result actually resolves.
Because of that gap, the Spark exchange is the only place in the entire system where a player can point at a specific character and say “I will own this” with full certainty before spending a single Crystal. Everything before that threshold is a bet on an unknown slice of a known-size pool.
The Granblue Fantasy Pity-Cost Table
Below is the full cost ladder for a Gala banner, from a single 10-pull up to the guaranteed Spark exchange. The cash figures are scaled directly from the roughly 800 dollar cost that gacha-economics trackers at GachaWiki’s pity-systems comparison and Game Design Lab both attach to a full 90,000-Crystal Spark, since Cygames does not publish a single official USD price list for Crystals on the standalone browser and mobile version of the game. Treat the dollar column as an informed estimate, not a storefront quote.
| Draws (Gala, 6% SSR) | Crystals spent | Estimated USD | Chance of at least one SSR | Guaranteed unit? |
|---|---|---|---|---|
| 10 (one 10-pull) | 3,000 | ~$27 | 46.1% | No |
| 30 | 9,000 | ~$80 | 84.4% | No |
| 50 | 15,000 | ~$133 | 95.5% | No |
| 100 | 30,000 | ~$267 | 99.8% | No |
| 300 (full Spark) | 90,000 | ~$800 | 100% | Yes, by exchange |
The jump between 100 draws and 300 draws is the part most players misjudge. At 100 draws you already have a 99.8% chance of having pulled at least one SSR, yet you still have zero guarantee it’s the one you wanted. The last 200 draws of a full Spark run exist purely to convert “probably got something good” into “definitely own exactly what I wanted,” and that certainty is what the extra 533 dollars actually buys.
Step-by-Step: How to Calculate the Real Odds Yourself
You don’t need a simulator to check these numbers. Granblue Fantasy’s flat-rate design makes the math simpler than almost any other major gacha game, because the probability never changes between draw 1 and draw 299.
The formula for the chance of landing at least one SSR within n draws is 1 minus (1 minus p) raised to the power of n, where p is 0.03 on a normal banner or 0.06 on a Gala banner. Plug in n equals 10 at the Gala rate: 1 minus 0.94 to the 10th power equals 1 minus 0.5386, which comes out to 46.1%. That single formula generates every percentage in the table above.
For the expected number of draws before your first SSR, the math is even simpler: it’s just 1 divided by p. At the Gala rate of 0.06, that’s 1 divided by 0.06, or about 16.7 draws on average. At the normal 0.03 rate, it’s 1 divided by 0.03, or about 33.3 draws. Multiply either number by 300 Crystals per draw and you have the expected Crystal cost of landing any SSR, though not necessarily the one you were chasing.
Finally, the expected number of SSRs across a full 300-draw run is n times p. At the Gala rate that’s 300 times 0.06, or 18 expected SSRs over the course of earning a full Spark. That number matters because it reframes the 800 dollar Spark cost: you’re not just buying one guaranteed unit, you’re statistically buying roughly 18 SSR pulls along the way, with the guaranteed exchange as a backstop on top.
Worked Example 1: One 10-Pull During a Gala
A player spends 3,000 Crystals, roughly 27 dollars by the scaled estimate, on a single 10-part draw during a Gala. Using the formula above, their odds of walking away with at least one SSR sit at 46.1%. That’s close to a coin flip, but remember it says nothing about which SSR they got. If the pool that Gala contains, say, four or five featured characters plus the entire back catalog of older SSRs, a single lucky pull is far more likely to land an older unit than the new, specific one the player wanted.
Worked Example 2: Spending Roughly $267 for 100 Draws
Scale that same player’s budget up to 100 draws, around 30,000 Crystals or an estimated 267 dollars. Their chance of at least one SSR climbs to 99.8%, which feels like a near-certainty. But a near-certainty of pulling some SSR is not the same as a certainty of pulling the right one, and this is exactly where impulse spending in Granblue Fantasy tends to go wrong. A player can spend 267 dollars, pull three or four SSRs, and still not own the one character they opened the game to get.
Worked Example 3: Budgeting the Full $800 Spark
Now take the same player and have them commit to the full 300-draw Spark from the start: 90,000 Crystals, an estimated 800 dollars. This is the only path in the entire system with a 100% success rate on the specific unit, because it bypasses RNG and becomes a straight exchange. Along the way, at an expected 18 SSR pulls across those 300 draws, the player is also statistically walking away with a stack of extra SSR characters or summons they didn’t have to pay anything extra for. Framed that way, the Spark isn’t just a guarantee, it’s a guarantee plus a free bonus haul.
Worked Example 4: Gala Timing Roughly Halves Your Cost-to-SSR
Compare the expected Crystal cost of landing any SSR on a normal banner versus a Gala banner. On a normal banner, the expected cost is 33.3 draws times 300 Crystals, or 10,000 Crystals, around 89 dollars. On a Gala banner, it’s 16.7 draws times 300 Crystals, or roughly 5,000 Crystals, around 44 dollars. Waiting for a Gala before spending real money doesn’t change the 800 dollar cost of a guaranteed Spark, since Sparks accumulate one per draw no matter the rate, but it roughly doubles your odds of incidental SSR luck along the way for the same money.
Worked Example 5: What a “Dry” Run Actually Looks Like
Take the unluckiest realistic case: a player on a Gala banner who draws 100 times, 30,000 Crystals, an estimated 267 dollars, and pulls zero SSRs. The formula says that outcome has only a 0.2% chance of happening, since 99.8% of players would have hit at least one SSR by that point. It’s rare, but it is not impossible, and it’s exactly why Granblue Fantasy’s Spark exists at all. Without it, that unlucky 0.2% would have no backstop beyond spending their way to 300 draws anyway. With it, that same player is already a third of the way to their guaranteed exchange, since every one of those 100 draws banked a Spark regardless of the outcome.
Normal Banner vs Gala Banner: Where the Real Savings Are
The table below lines up the two base rates side by side at the same draw counts, so the gap is easy to see at a glance.
| Draws | Normal banner (3% SSR) odds | Gala banner (6% SSR) odds |
|---|---|---|
| 10 | 26.2% | 46.1% |
| 30 | 59.8% | 84.4% |
| 50 | 78.1% | 95.5% |
| 100 | 95.2% | 99.8% |
| 300 (Spark) | 100% | 100% |
Notice the Spark row is identical regardless of rate. The deterministic exchange doesn’t care how lucky or unlucky the odds were along the way, which is exactly the point of including it in the system. It’s the floor under the floor, a guarantee that holds even in a world where the SSR rate never moved.
Why There’s No 50/50 or Escalating Soft Pity in Granblue Fantasy
Plenty of gacha games layer a rising soft-pity curve into the final stretch before a hard cap, pushing the odds of a rare drop upward the longer a player goes without one, then resolving a coin-flip style guarantee between a featured unit and the standard pool. Granblue Fantasy does neither. The rate stays flat from the first draw to the 299th, and there is no coin flip deciding whether a landed SSR is the new featured character or an older one from the catalog.
Instead, the entire certainty mechanism lives outside the roll itself, in the Spark counter. It’s a structurally different answer to the same question every gacha game has to solve: how do you promise a player they will eventually get what they’re spending for, without it happening on literally every single roll? Granblue Fantasy’s answer is to decouple the guarantee from the probability entirely. The odds never change, but a separate, parallel currency eventually buys the outcome outright.
That tradeoff cuts both ways. A player who gets unlucky early gets no comfort from rising odds the way they might in a game with soft pity. But a player tracking their Spark count always knows exactly how many draws stand between them and a guaranteed result, with zero ambiguity about when that guarantee kicks in.
Myths vs Facts About Granblue Fantasy’s Gacha
| Myth | Fact |
|---|---|
| Landing an SSR during a Gala guarantees the new featured character | The 6% rate covers the entire SSR pool, old and new characters alike, with no published isolated rate for the specific featured unit |
| The game has a 50/50 system like other gacha titles | Granblue Fantasy has no 50/50 mechanic at all, featured and standard SSRs share one flat pool |
| Odds get better the longer you go without an SSR | There is no soft pity curve, the base rate stays at 3% or 6% on every single draw regardless of your luck streak |
| The 10th slot in a 10-pull guarantees an SSR | It only guarantees SR or better, not specifically an SSR |
| Cerulean Sparks carry over to the next banner if unused | Sparks are tied to the active promotional period and expire when it ends |
| You need to spend real money to reach a Spark | Free Crystals and event draw tickets count toward the same 300-draw Spark total |
Is Granblue Fantasy’s System Actually Fair?
Fairness in a gacha context usually comes down to two separate questions: are the odds what the publisher says they are, and does the system eventually guarantee a result for a bounded amount of money. On the first question, Granblue Fantasy’s 3% and 6% SSR rates have held steady for years and are consistently reported the same way across independent trackers, which is itself a form of reliability that newer, frequently rebalanced games don’t always offer.
On the second question, the answer is unambiguous. The Spark caps total spend on any single featured unit at 90,000 Crystals, an estimated 800 dollars, with zero variance. Compare that to a pure-RNG system with no pity at all, where a sufficiently unlucky player could in theory spend far more than that chasing the same character with no backstop whatsoever. Granblue Fantasy’s flat rate plus hard Spark cap gives up the feel-good momentum of rising odds in exchange for a number you can write down in advance and never exceed.
That tradeoff has become more visible since the game’s global Steam debut on March 10, 2026, per the official Granblue Fantasy Wiki’s Steam launch guide, which opened the decade-old browser and mobile title to a wave of new Western players who are comparing its economy against newer competitors for the first time. Many of those newer games advertise escalating pity as a player-friendly feature. Granblue Fantasy’s pitch is the opposite: predictability over momentum, a flat number over a climbing one.
Whether that’s fair depends on what a player values. A flat rate with a hard, calculable cap is easier to budget against than a system with hidden, shifting curves, even if it feels less exciting in the moment a player is actually rolling. Tools like the community-run GBF Spark Calculator exist specifically so players can model their own odds against their own budget before they spend a single Crystal, which is itself a useful gut check against the sunk-cost pull to just keep rolling past a budget that was already lost to bad luck.
Should You Pull Early or Save for the Spark?
Given everything above, the practical answer splits by goal. A player who just wants more SSR characters in general, without caring which ones, gets the best value pulling opportunistically during Gala periods, where the expected cost of landing any SSR drops to roughly 44 dollars instead of 89. A player who wants one specific, named character and refuses to accept a substitute has exactly one rational plan: budget for the full 800 dollar Spark from the start, treat every pull along the way as a bonus, and ignore the temptation to stop early just because the odds say a pull is likely to hit by that point. Likely is not guaranteed, and in Granblue Fantasy, guaranteed has one price and one price only.
Frequently Asked Questions
Does Granblue Fantasy have a pity system?
Not in the escalating-odds sense most gacha games use. The SSR rate stays flat at 3% normal or 6% during a Gala. The closest thing to pity is the Spark system, which guarantees a free choice of any item on a banner’s exchange list after 300 draws.
How many pulls guarantee an SSR in Granblue Fantasy?
No specific pull count guarantees an SSR through the roll itself, since the rate never rises. The only guaranteed path to a specific SSR is reaching 300 draws and using the Spark exchange, which is a selection, not a random roll.
What is the Spark system in Granblue Fantasy?
Every draw banks one Cerulean Spark. At 300 Sparks during an active banner, a player can exchange them for any single item on that banner’s list, including the newest limited SSR character or summon, with no RNG involved.
How much does a full Spark cost in real money?
A full Spark requires 300 draws at 300 Crystals each, for 90,000 Crystals total. Gacha-mechanics trackers estimate that at roughly 800 dollars in real currency, though Cygames does not publish a single official USD price list for the standalone browser and mobile version of the game.
What’s the difference between Normal and Gala SSR rates?
A standard premium draw has a 3% chance of an SSR. During a Legend Festival or Flash Gala, that rate doubles to 6%, roughly every couple of weeks, making Gala periods the more efficient time to spend.
Does hitting an SSR during a Gala guarantee the new featured character?
No. The 6% rate covers the full SSR pool, which includes both the newly featured units and the existing SSR catalog. There is no published isolated rate confirming a landed SSR is specifically the new character.
Do Cerulean Sparks carry over between banners?
No. Sparks are tied to the active promotional period and expire once that banner ends. Partial progress toward 300 on one banner does not transfer to the next.
Is Granblue Fantasy’s gacha system fair compared to other games?
Its base rates have stayed consistent for years and the Spark caps the maximum possible spend on any one unit at a fixed, calculable number. It trades the feel-good momentum of rising odds for a flat rate plus a hard, predictable ceiling on worst-case cost.