Learn
Why Your Last 10 Bad Pulls Don't Mean the Next One Is Better
The gambler's fallacy, sunk cost bias, and gacha psychology explained with actual math instead of banner superstition.
The gambler's fallacy in one sentence
The gambler's fallacy is the belief that a random process should self-correct in the short run. Flip tails five times and people start feeling heads is due. Miss ten summons in a row and players start talking like the banner owes them a hit. That belief feels intuitive because humans expect balance to show up quickly, but most random systems do not work that way.
If each pull is independent, the eleventh pull has the same base probability as the first. A bad streak can make the next result emotionally important, but it does not make the next result statistically favored. The only time your odds change is when a published rule changes them, such as pity, rate-up thresholds, or drawing without replacement from a shrinking pool.
How gacha systems exploit that instinct
Gacha design leans hard on the phrase people say to themselves after a bad run: I am probably close. Animations, fake-outs, and near-miss language all reinforce the feeling that progress is accumulating even when the system is still rolling a base-rate event. Players end up interpreting time spent as probability earned.
This is powerful because the pain of stopping after a streak feels worse than the pain of starting fresh. The next pull feels like it contains the value of all previous pulls. Unless pity is actually active, that is fiction. The banner is not tracking your frustration as a hidden resource. It is tracking only the rules the designer wrote into it.
Sunk cost makes the mistake expensive
Sunk cost fallacy shows up when players spend more to justify what they have already spent. Someone who would never begin a banner at current odds suddenly decides to keep rolling because they are already eighty pulls deep. The past spend becomes a reason for future spend, even though the money is already gone and cannot improve the quality of the next decision.
Trading card openings produce the same trap. Once several weak boxes are open, another box starts to feel like the one that will finally make the math work. In reality, expected value should be judged from the present moment forward. The only rational question is whether the next purchase is worth its current odds and cost.
Pity is real, but it is not kindness
Pity systems are often presented like player protection. They do protect against the ugliest tails of variance, but they also create a predictable spend ladder. Once you are near pity, quitting becomes psychologically difficult because the remaining gap feels small relative to what you already paid. That is not accidental. It is a retention and monetization tool as much as it is a fairness tool.
This does not make pity evil. It just means you should read it as a pricing mechanic. If hard pity is 90 pulls and you are at 82, your next eight pulls may have better decision logic than your first eight. But the logic comes from explicit design, not from random chance finally deciding to be generous.
A better way to think about banners
Estimate the full expected cost of reaching your target. If a featured unit is guaranteed at a certain threshold, price the banner based on that threshold, not on your best-case fantasy. If there is no guarantee, decide the maximum you are willing to lose before you start. That turns the pull session into a bounded decision instead of a streak-driven spiral.
Once you frame pulls as a probability problem instead of a narrative, the compulsion drops. Bad runs stay annoying, but they stop feeling meaningful.