Stake strategiesUpdated 2 min readBy the FXNOD team
Does digit analysis work on Deriv synthetic indices?
Digit analysis counts how often each last digit has appeared over recent ticks. It describes the past accurately. On an index generated by a random number generator, as Deriv describes its synthetic indices, each digit is equally likely on every tick regardless of history, so the counts do not predict the next digit. You can test this yourself.
What the tools show
- The percentage of recent ticks that ended in each digit.
- The most and least frequent digit in the window.
- The share of even against odd, or of digits over and under a level.
- Current streaks.
Why the counts are never level
Over 100 ticks, a perfectly random source gives each digit 10 appearances on average, with a normal spread of about 3 either way. Seeing one digit at 15% and another at 6% is what randomness looks like at that sample size. Over 1,000 ticks the percentages crowd closer to 10%, and over 30 ticks they are all over the place. An uneven table is not evidence of a pattern.
| Window | Expected count per digit | Typical range |
|---|---|---|
| 30 ticks | 3 | 0 to 6 |
| 100 ticks | 10 | 4 to 16 |
| 1,000 ticks | 100 | 81 to 119 |
The two opposite theories
Some traders bet the hot digit will keep appearing. Others bet the cold digit is due. Both read the same table and reach opposite conclusions, which is a sign that the table does not contain the answer. On independent ticks, both win at the rate chance gives.
Test it yourself
Write the rule exactly
For example: when a digit is above 14% over the last 100 ticks, buy Differs on it.
Run it on demo with a flat stake
At least 500 trades.
Compare with chance
Differs wins about 90% by chance, Matches about 10%, Even or Odd 50%.
Allow for luck
Over 500 Differs trades, anything from about 87% to 93% is within normal variation.
What digit statistics are good for
- Giving a bot a rule for when to enter, so it trades less often.
- Choosing among contracts with known odds.
- Seeing for yourself how lumpy random data is.
Trading less often is a real benefit when each trade costs a little on average. That is a benefit of pacing, not of prediction.
In FXNOD
Auto Hub offers several digit-contract bots with fixed entry rules. A recent pattern does not predict the next tick. FXNOD cannot verify Deriv's random generator independently.
Questions people ask
- Can you predict the last digit on Deriv?
- Not if the index is random as Deriv describes. Each digit has the same chance on every tick.
- Why does one digit appear more than the others?
- Because small samples of random data are uneven. Over more ticks the percentages move towards 10% each.
- Are digit analysis tools a scam?
- The statistics are real. Claims that they predict the next digit, or guarantee a profit, are not supported by how the index is generated.
See the stakes before you start
Bot Forge shows the stakes of a losing streak before a bot runs, will not start without a stop loss, and caps the largest stake. Try any stake setting on your Deriv demo account.
Get started