Push your luck, but not too far: what thousands of simulated games say about banking.
Every turn of Higher or Lower comes down to one nagging question. You’ve got a streak of three going, the next card looks a bit risky, and the Bank button is right there. Do you take the points, or chance one more guess? A good higher or lower strategy is mostly about answering that question well, so we stopped guessing and ran the numbers.
We played tens of thousands of games with the real Randomfella rules, trying different banking habits against the game’s own Normal computer. Some of the results were obvious. One of them was a genuinely useful rule of thumb.
The push-your-luck problem
In this version of higher or lower, every right guess adds 1 to your streak, and banking adds the streak to your score. A wrong guess wipes the streak out. The first to bank 10 points wins (11 for whoever went first).
That creates a classic tension. Bank too early and you crawl towards the target one point at a time. Bank too late and you keep throwing good turns away. Somewhere in between is a sweet spot, and it isn’t where most people’s instincts put it.
How quickly streaks end
Let’s start with the raw survival rate. Even calling the better side every single time, here’s how often a turn reaches each streak:
| Streak reached | How often |
|---|---|
| 1 right guess | 77% |
| 2 in a row | 57% |
| 3 in a row | 42% |
| 4 in a row | 31% |
| 5 in a row | 23% |
| 6 in a row | 17% |
The drop is steep. Fewer than half of all turns make it to three, and fewer than one in four make it to five. Long streaks happen, and they feel brilliant when they do, but they’re not something to plan around.
Points per turn, by banking habit
Now multiply each streak by the chance of reaching it, and you get the average points a habit earns per turn:
| Always bank at | Average points per turn |
|---|---|
| 1 | 0.77 |
| 2 | 1.14 |
| 3 | 1.26 |
| 4 | 1.25 |
| 5 | 1.16 |
| 6 | 1.03 |
So if you’re going to pick one number and stick to it, 3 or 4 is the sweet spot. Banking after every single right guess is the worst habit of all: it feels safe, but it leaves far too many points on the table.
A better higher or lower strategy: watch the odds
A fixed number ignores the most useful information you have, which is the card in front of you. A streak of 3 on a 2 is very different from a streak of 3 on an 8. On a 2, the next call is right about 94% of the time. On an 8, it’s worse than a coin flip.
So we tried a smarter rule: bank whenever your best call drops below about 60%, and otherwise keep going. Here’s how each approach did over 20,000 games against the Normal computer:
| Your habit | Games you won |
|---|---|
| Bank after every right guess | 13% |
| Always bank at 2 | 42% |
| Always bank at 3 | 50% |
| Always bank at 4 | 54% |
| Bank when your best call is under 60% | 58% |
That’s a clear winner. Watching the odds beat every fixed habit we tried, including the cautious-feeling ones. We also tested thresholds of 55%, 65% and 70%, and they all did slightly worse than 60%.
What 60% looks like in real cards
With a full deck, your best call is under 60% when a 7, 8 or 9 is showing. Everything from a 6 down and a 10 up is 63% or better. That gives you a version of the rule you can play without any maths at all: bank on the middle cards, push on the edges.
Of course, the deck doesn’t stay full. As cards are played, the odds shift, sometimes a lot. A 10 can become a risky card once most of the low cards have gone. That’s why the game’s odds bar is so handy. It counts the cards still in the deck for you, so the 60% rule works at any point in the game.
When the score should change your mind
A rule of thumb is a starting point, not a commandment. The score matters too.
- When you’re well behind, small banks won’t catch up fast enough. It can be worth pushing one card further than usual, especially on an easy card.
- When you’re well ahead, the opposite applies. Bank a little earlier and make your opponent take the risks.
- When one bank would win the game, take it. There’s no prize for winning by more.
The game’s Hard computer plays a simple version of this, banking sooner when it leads and later when it trails. In our tests that score-aware habit beat the Normal computer about 53% of the time, a real edge but smaller than the one you get from simply watching the cards.
A quick word on luck
Even the best strategy here wins a little under six games in ten against a sensible opponent. That’s the nature of a game built on a shuffled deck: good decisions tilt the odds, they don’t remove them. Some games you’ll do everything right and still watch an 8 turn into a bust.
That’s also why Randomfella keeps it purely for fun. There’s no money and nothing to buy, just points that matter for exactly as long as the game lasts.
Put the higher or lower strategy to the test
So here’s the whole higher or lower strategy in one line: bank when your best call drops below 60%, push on easy cards, and let the score nudge you either way. It’s simple, it’s easy to remember, and in our simulations it won more often than anything else we tried.
Try it for yourself in a game of Higher or Lower against the computer, or teach it to a friend and see who sticks to it longer. If you’re new to the game, start with our guide on how to play higher or lower, which covers the rules and the odds for every card.