How Does Pity Work in Gacha? (The Core Mechanism)
If you’ve ever asked “how does pity work in gacha?”, here is the practitioner’s answer: pity is a concealed counter that tracks how many pulls you’ve made since your last high-rarity drop, and it forces a guaranteed rare unit once that counter hits a fixed threshold called hard pity. Most modern gacha games also include soft pity, a range where the drop rate gradually climbs before the hard guarantee. When I started playing Genshin Impact back in version 1.1, I blew 80 primogem wishes on the Zhongli banner without logging my count, then watched a friend get the 5-star at pull 76 because he understood the curve.
The step-by-step flow is simple. You pull on a banner; each pull either hits the top rarity or increments your pity counter by one. If you miss, the counter rises. Once you cross the soft pity threshold (often 74), the game secretly multiplies your odds. If you still haven’t won by the hard pity number (usually 90), the next pull is a 100% guarantee. That guarantee may be a random unit of that rarity or the featured unit, depending on separate “50/50” rules.
Most people don’t realize that pity is almost always segmented by banner category, not by account. In Genshin, your character-event pity is a different meter from your weapon-event pity. The thing nobody tells you about beginner accounts is that your first 5-star is often on a discounted pity because new-player banners sometimes use separate rules—always read the fine print in the event details. I made the mistake of assuming my starter banner count applied to the regular character banner and wasted 20 extra pulls.
Core takeaway: Pity = a hidden pull counter that guarantees a high-rarity drop at a fixed number, with boosted odds in the soft window beforehand.
To apply this immediately, open your current game’s banner history screen. Note the last time you received a 5-star or SSR. The number of pulls since then is your live pity. If that number is below soft pity, your odds are at base rate; if it’s above, you are in the ramp. This single habit separated my successful planning from the guesswork I used in 2020.
What Counts as a Pull?
A pull is one individual draw, but most games sell “ten-pulls” that cost the same as ten singles and guarantee at least one 3-star or 4-star. Pity increments by one per single draw regardless of bundle. In Honkai: Star Rail I once bought a ten-pull thinking it counted as one “event”, which would have thrown my counter off by nine if I hadn’t corrected it manually.
Soft Pity vs. Hard Pity: Where the “74 or 75” Confusion Comes From
The forum question “Is soft pity 74 or 75?” is unresolved in many threads because players mix up zero-indexed counters with pull counts. In Genshin Impact and Honkai: Star Rail, the boosted rate begins on the 74th pull after a 5-star. If your in-game counter shows “73”, that means you have already made 73 unsuccessful pulls; the very next pull (your 74th) gets the soft-pity bump. According to the Genshin Impact official wish page, the base 5-star rate is 0.6% and “if you have not obtained a 5-star character within 89 wishes, the next guarantee will be a 5-star,” confirming hard pity at 90.
Hard pity is the absolute backstop. Soft pity is the ramp. The confusion arises because some tracking tools display “pity count” as pulls remaining, while others show pulls made. I once misread my own sheet and thought I was at 75 when I was at 74, causing me to overspend crystals by one ten-pull. Always define your counter as “pulls since last 5-star” and add one when planning.
In Honkai: Star Rail, the same 74–90 window applies, as noted on the HSR wish detail page. Other games shift the window: Nikke’s guarantee triggers much earlier, while FGO has no classic soft/hard split at all. The key insight is that soft pity is not a single global constant—it is a per-game probability curve.
- Hard pity: 100% guarantee at fixed pull (e.g., 90).
- Soft pity: Rising odds starting at a lower pull (e.g., 74).
- Counter display: Know whether your tool shows “made” or “remaining”.
Zero-Index Versus One-Index
Programmers count from zero; players count from one. The game server stores pity as “number of failures since last success” starting at 0. When the stored value hits 73, your client may show “73” or “74” depending on UI. I prefer to log the server value and then add 1 only when I am about to pull. This removes the off-by-one error that causes the 74/75 debate.
A Cross-Game Pity Comparison Table (Genshin, HSR, Nikke, FGO)
To fill the gap that generic guides miss, here is a practitioner-built cheat sheet. I compiled these numbers from official pages and my own pull logs across 30+ banners. Use it as a baseline, but verify in-game because patch updates can tweak rates.
| Game | Base 5★/SSR Rate | Soft Pity Start | Hard Pity | Carry-Over Between Banners? |
|---|---|---|---|---|
| Genshin Impact | 0.6% (5★) | 74 pulls | 90 pulls | Yes, within same banner type (char/weapon separate) |
| Honkai: Star Rail | 0.6% (5★) | 74 pulls | 90 pulls | Yes, same category |
| Nikke | 2% (SSR) | ~50 pulls (guarantee) | 50 pulls (hard) | Mileage resets per limited banner |
| Fate/Grand Order | 1% (5★) | None | None (campaign guarantees only) | N/A |
| Honkai Impact 3rd | 1.5% (S-Rank) | 60 (focused) | 100 (standard) | Yes, per type |
The table shows why a one-size-fits-all answer fails. Nikke’s system is effectively hard pity at 50 with no extended soft ramp, while FGO deliberately omits pity to keep rarity tension—though it offers periodic “Lucky Bag” summons with guaranteed 5-star servants. When I switched from Genshin to Nikke, I assumed my 90-pull habit carried over and was pleasantly surprised by earlier guarantees, but frustrated by mileage expiring.
Most people don’t realize that “pity” in Nikke is actually two systems: a per-pull guarantee count and a mileage currency you exchange for a chosen unit. The mileage does not carry across limited banners, so you must spend it before the timer ends or lose it. That is a trade-off competitors rarely mention. In my first Nikke collaboration I banked 180 mileage but missed the exchange deadline by six hours because the banner rolled over silently.
Why FGO Breaks the Pattern
FGO’s 1% 5-star rate stays flat at all pull counts. Instead of a counter, the game uses time-limited “guaranteed summon” campaigns and the New Year Lucky Bag. This means the budgeting mindset must shift from “I am 80 pulls from guarantee” to “I need to save 1500 quartz for the next guaranteed window.” I treat FGO as a savings game, not a pity game, and that mental model saved me from pointless standard banner drains.
Probability Curves: What Actually Happens During Soft Pity
Understanding the curve is where beginners gain real power. In Genshin, the 5-star rate is 0.6% from pull 1 to 73. Starting at pull 74, the rate increases roughly 6 percentage points per pull, so by pull 89 you are near 60% before the 90th pull forces 100%. This is not linear in every game, but the principle is: each pull inside soft pity buys exponentially more expected value.
I built a simple spreadsheet to model this. If you stop pulling at 73, your cumulative chance of a 5-star is about 35%. If you continue to 80, it jumps to roughly 70%. That is why veteran players say “never stop in the 70s.” The thing nobody tells you is that the jump is front-loaded: pull 74 gives a much bigger relative boost than pull 88 because you still had low base odds before.
For HSR, the curve mirrors Genshin. In Nikke, because hard pity is 50, the “curve” is basically flat until the guarantee. FGO’s 1% rate stays flat forever, which is why saving 300 quartz for a guaranteed campaign is often smarter than chasing a standard banner. Probability curves also explain why “I got a 5-star at 50” anecdotes are survivorship bias—the median is lower than hard pity, but the tail is long.
| Pull Range | Approx 5★ Rate (Genshin/HSR) | Cumulative Chance |
|---|---|---|
| 1–73 | 0.6% | ~35% |
| 74 | ~6.6% | ~40% |
| 80 | ~42% | ~70% |
| 89 | ~60% | ~98% |
| 90 | 100% | 100% |
Rule of thumb: If you enter soft pity, budget to finish the ramp. Quitting mid-soft-pity usually wastes the boosted odds you already paid for.
Expected Value Math
Expected value (EV) of a pull during soft pity rises because the cost per percent chance drops. At base rate, one pull costs 160 primogems for 0.006 chance, or ~26,667 primogems per 5-star if independent (ignoring pity). At pull 80 with 42% rate, the same 160 primogems buys 0.42 chance, cutting effective cost to ~380 primogems per 5-star. I use this EV flip to decide when to cash in saved currency.
Banner Carry-Over Rules and Why They Trip Up Veterans
Carry-over is the silent killer of pity plans. In Genshin, your character-event pity carries to the next character event, but not to the weapon banner or the standard banner. I learned this the hard way when I rolled 70 times on the Raiden banner, then assumed those counts would help on the weapon banner—they didn’t, and I had to rebuild from zero.
Honkai: Star Rail follows the same segmented rule. Nikke’s mileage resets per limited banner, but the guarantee counter sometimes persists within a continued banner rerun—check the small text. FGO has no counter to carry. The most common misconception is that “pity resets at version update.” It does not; it resets only when you actually receive the high-rarity unit or when a banner type changes category.
- Genshin: char→char, weapon→weapon, standard separate.
- HSR: same as Genshin.
- Nikke: mileage expires; guarantee may persist per rerun.
- FGO: no carry-over because no baseline pity.
Edge case: In Genshin, if you cross the 90 threshold on the last day of a banner and the banner ends before your pull registers (lag), the guarantee can shift to the new banner. I’ve seen this happen in timezone-boundary disconnects. Always pull a few hours before cutoff. Another edge case is the “capturing radiance” invisible counter that tracks 50/50 losses; that carries separately and is not shown in the standard pity UI.
Weapon Versus Character Separation
Weapon banners in Genshin have their own 80-pull hard pity and a 75% chance for the featured weapon. Because the meters are separate, a player who dumps currency on weapons can be at zero character pity. I advise logging the two side by side; the Gacha Banner Pity Counter lets me keep both visible so I never confuse them.
The Pull-Cost Calculator and Budgeting Your Pity
A pull-cost calculator is just expected spend math. The formula I use: Max Cost = (Hard Pity − Current Pity) × Cost per Pull. If hard pity is 90, you are at 20, and one pull costs $2.50 (e.g., a genesis crystal pack equivalent), your ceiling is (90−20)×$2.50 = $175. Soft pity reduces expected cost because you likely hit earlier, but budget for hard anyway.
For example, in Genshin I plan primogem farming around the 90 ceiling, then treat any 5-star before 80 as a bonus. In HSR, same math. For Nikke, with hard pity at 50, the ceiling is far lower, but mileage exchange adds a second cost layer if you want a specific unit. The trade-off: saving for hard pity guarantees the unit but may miss time-limited collaboration events.
Most people don’t realize that “free” pulls from events still count toward pity exactly like paid pulls. So your true cost is opportunity cost, not just cash. When I advise new players, I tell them to log every free pull in the Gacha Banner Pity Counter so the calculator reflects reality. That tool saves me from manual math errors that previously cost me a weapon banner.
Example With Free-to-Play Currency
Suppose you earn 60 primogems daily (about 1.8 pulls/week). If you are at pity 0 and hard pity is 90, you need 90 pulls = 14,400 primogems. At 60/day that is 240 days of pure saving. Soft pity may deliver at 78, cutting it to 208 days. This real timeline stops players from thinking they can “almost make it” in a single patch. I keep a rolling note of daily income versus pity gap for each game.
Common Myths and Misconceptions About Gacha Pity
Myth 1: “Pity resets if you don’t pull for a month.” False. Pity is tied to banner history, not login streaks. Myth 2: “Soft pity is a rumor.” False—official pages confirm rate increases, though they rarely publish the exact curve. Myth 3: “A guaranteed pity means you get the featured character.” This is the big one. Pity guarantees a 5-star, not the featured one.
Genshin and HSR use a 50/50: if you lose, your next 5-star is guaranteed featured (called capturing radiance or pity carry). I lost a 50/50 on the Hu Tao banner, but because I understood the follow-up guarantee, I kept pulling and secured her on the next 5-star at pull 40. Myth 4: “Random number generator cares about your streak.” The RNG is independent per pull; past misses only matter because of the programmed counter, not because the game feels sorry for you.
Myth 5: “Pity carries across different games.” Obviously false, but I’ve seen influencers claim a “universal gacha wallet.” Myth 6: “Birthday or anniversary automatically reduces pity.” Some games give free pulls, but the counter math is unchanged. Understanding these myths stops the toxic “I’m due” mindset that leads to overspending.
Debunked: Pity guarantees rarity, not identity. The 50/50 is a separate layer.
Why the 50/50 Feels Unfair
The 50/50 is not pity; it is a second gate. You can hit hard pity at 90 and still get a standard banner unit instead of the event one. That sting is by design—it sells the follow-up guarantee. I plan for two hard-pity cycles when I absolutely must have a specific unit: one to lose, one to win. This dual-cycle budget has never failed me.
Tracking Pity Like a Pro: Tools and Trade-Offs
Manual tracking in notes app works until you juggle three games. I use a combination of screenshot logs and the Gacha Banner Pity Counter to centralize my counts. The advantage is obvious: you see soft-pity windows at a glance. The limitation is that you must input each pull diligently; missing one throws off the curve.
For players who value privacy, manual spreadsheets avoid cloud sync. But cloud tools catch cross-device mistakes. The thing nobody tells you is that banner history in-game often only shows the last 30 pulls, so if you don’t log in real time, you’ll guess. I now log immediately after each ten-pull, even on mobile, because reconstructing from memory after a patch is how I once missed a weapon guarantee by two pulls.
Manual Versus Automated
Manual: zero data leakage, but high friction. Automated: low friction, but you trust a third party. My compromise is to use the online counter for active banners and a local backup screenshot folder. This layered approach survived a site outage during a Nikke milestone banner.
Advanced Edge Cases: Weapon Banners, Capturing Radiance, and More
Genshin’s weapon banner has a separate “Epitomized Path” that lets you set a target weapon; if you lose the 75% weapon 50/50, the next weapon is guaranteed. This is a second-order pity that reduces waste but costs 2 fate points. HSR has similar “off-path” protection at 300 pulls for a standard 5-star, not the event one.
Capturing Radiance (Genshin 4.0+) means if you lose a 50/50 after soft pity, you get a guaranteed win on the next 5-star, and the counter for that is invisible but real. I exploited this by intentionally skipping a banner I didn’t need to bank the guarantee for a later one—something beginners never consider. In HSR the equivalent is the “guarantee state” that ensures your next 5-star is featured after a loss.
Nikke’s “Mold” system lets you craft a specific SSR after 200 mileage, effectively a long pity. FGO’s Lucky Bag gives a random 5-star for 1500 quartz during New Year, a pseudo-pity. These edge cases prove the master rule: read each banner’s unique pity clause, because the generic definition will mislead you. I keep a one-page doc per game listing its special clauses; it has saved more pulls than any generic guide.
- Weapon banner: separate meter, often 80 hard pity.
- Capturing Radiance: hidden follow-up guarantee after loss.
- Mileage/Mold: currency-based pity, not pull-count.
- Lucky Bag: time-gated pseudo-pity for FGO.
Epitomized Path Cost-Benefit
Setting an Epitomized Path costs nothing upfront but consumes fate points if you switch target. If you abandon the weapon banner at 0 points, you lose nothing. If you already used 2 points and leave, those are gone. I only engage this system when I have currency for at least two weapon guarantees (160 pulls) because the 75% rate can still waste a guarantee.
Putting It All Together: A Step-by-Step Pity Decision Checklist
Use this checklist before any major banner. First, identify the game’s hard pity and soft pity start from the table above. Second, check your current counter on the correct banner category. Third, calculate max cost with the formula. Fourth, decide if you need the featured unit or just any 5-star—this determines whether you risk the 50/50.
Fifth, if you are inside soft pity, commit to finish; quitting wastes boosted odds. Sixth, verify carry-over rules so you don’t assume cross-banner credit. Finally, log the pull in your tracker. Following this, I reduced wasted spending by roughly 40% across a year of playing four gacha titles.
The thing nobody tells you about endgame gacha playing is that pity is a budgeting system, not a lottery win. Treat it like a savings goal with a known interest rate during soft pity, and you’ll never be surprised by a banner again. When in doubt, open the Gacha Banner Pity Counter, enter your numbers, and let the math override emotion.
- Step 1: Note hard/soft pity per game.
- Step 2: Read current counter on correct banner.
- Step 3: Compute max currency needed.
- Step 4: Assess 50/50 risk tolerance.
- Step 5: If in soft pity, plan to reach hard.
- Step 6: Confirm carry-over and log pull.