D'Alembert strategy on Deriv: how the stake moves, step by step
7 min read
The D'Alembert strategy raises the stake by a fixed amount after a loss and lowers it by the same amount after a win. The stake climbs in a straight line rather than doubling, so a losing run costs less than it would on a martingale, and a win takes back one step instead of clearing the whole run. Deriv's Bot Builder offers it as a Quick Strategy, and BinaryTick runs it as a free bot you can read and tune before you press Run bot.
How the D'Alembert strategy moves the stake
On BinaryTick the strategy runs in D'Alembert Ratchet, one of the free Even / Odd bots. Its stake follows three rules, each shown on its card.
- After a loss, the bot adds one step.
- After a win, it removes one step. The step count never drops below zero, so the stake never falls below the Initial stake.
- The stake is the Initial stake plus the steps times the Step size. At the defaults that is 1 USD plus 1 USD per step.
The step count is capped by Max steps, 6 by default, so the stake tops out at 7 USD. The card's Max stake, 20 USD by default, is a second ceiling: if the stake would pass it, the bot drops back to the Initial stake. With the default Step size and Max steps the stake never gets near 20 USD, so the max stake only comes into play if you raise those two.
A run, worked through with real payouts
D'Alembert Ratchet trades one tick Even / Odd on Volatility 25. When we ran it, Deriv paid 1.95 USD on a 1 USD stake, 3.91 USD on 2 USD and 5.86 USD on 3 USD. Here is an example of six trades at the default settings, worked through with those payouts: two losses, a win, a loss and two wins.
| Trade | Stake | Result | Profit on the trade | Session profit | Next stake |
|---|---|---|---|---|---|
| 1 | 1 USD | Loss | -1.00 | -1.00 | 2 USD |
| 2 | 2 USD | Loss | -2.00 | -3.00 | 3 USD |
| 3 | 3 USD | Win | +2.86 | -0.14 | 2 USD |
| 4 | 2 USD | Loss | -2.00 | -2.14 | 3 USD |
| 5 | 3 USD | Win | +2.86 | +0.72 | 2 USD |
| 6 | 2 USD | Win | +1.91 | +2.63 | 1 USD |
Three losses and three wins, and the session ends 2.63 USD up. That is the idea behind D'Alembert: losses add to the stake, so the wins that follow them land at a higher stake than the losses that started the run. Trade 3 shows the other side. One win at 3 USD takes the stake down to 2 USD, not back to 1, so the bot keeps working its way down one step at a time.
What a losing run looks like
Each loss in a row costs 1 USD more than the one before it. At the defaults, four losses in a row cost 1, 2, 3 and 4 USD, 10 USD in all.
The card's limits meet at that point. Max losses in a row is 4 by default, and Stop loss is 10 USD, so a run of four losses from the starting stake ends the session on both counts at once. Both are checked after each trade. If a run starts after a few losses have already pushed the stake up, the stop loss usually arrives first.
Take profit stops the bot once the session is 10 USD up. Every run starts these totals from zero.
The profit lock
D'Alembert Ratchet adds one thing the plain strategy does not have: a trailing lock on profit. Once session profit reaches the Profit lock trigger, 5 USD by default, the lock arms. From then on the bot stops if profit falls the Profit lock distance, 2 USD by default, below its highest point.
Say the session climbs to 6.50 USD. The bot keeps trading, and stops if profit drops to 4.50 USD or lower. If it climbs on to 8 USD, the stopping point moves up to 6 USD. Either the lock or Take profit ends a good session, whichever comes first.
D'Alembert or martingale?
Both raise the stake after a loss. They differ in how fast, and in what one win does.
- Martingale multiplies the stake after each loss, 2.1x on most BinaryTick bots, so one win recovers the run, as long as it comes before the stake reaches the max stake and the ladder resets. The martingale guide works through it.
- D'Alembert adds one fixed step, so a losing run grows the stake slowly, and it takes about as many wins as losses to step back down.
If you want each loss to cost only a little more than the last, D'Alembert fits. If you want a short losing run cleared by one win, use a martingale bot with a max stake you are comfortable with. For a system that holds the stake through losses and raises it after wins, see Oscar's Grind. For a ladder where each stake is the sum of the two before it, see the Fibonacci strategy.
When D'Alembert Ratchet trades
The staking decides how much the bot stakes. The entry rule decides when. D'Alembert Ratchet looks at the last digits of the most recent ticks and works out what share of them were even.
- If evens make up 58% or more of the last 50 digits, it buys Odd.
- If they make up 42% or less, it buys Even.
- In between, it waits.
Each tick is a fresh draw, so the scan leaves the odds of the next digit where they were. It changes how often the bot is in the market: it waits for a lopsided window, then trades against the bigger side. The last digit analysis guide explains what digit statistics can and cannot tell you.
The settings that matter
Step size. How much one step adds to the stake. A smaller step makes every loss in a run cheaper, and every win at a raised stake smaller.
Max steps. How many steps the stake can climb. At the defaults, 6 steps cap it at 7 USD.
Initial stake. The stake with no steps added. The minimum on the card is 0.35 USD.
Max stake. The ceiling that sends the stake back to the start if the steps ever push it past.
Sample size and Bias edge (%). How many recent digits the scan reads, 50 by default, and how far from an even split the share must sit before the bot trades, 8 by default. A larger edge means fewer, more selective trades.
Cooldown after a loss (ticks). The bot sits out 2 ticks after a losing trade before it looks for a new entry.
Profit lock trigger and Profit lock distance. When the lock arms, and how much profit it gives back before it stops the bot.
Using D'Alembert from Deriv's Bot Builder
Deriv's Quick Strategy version asks for an initial stake, a unit, a profit threshold and a loss threshold. Its own guide describes the same rule: add a unit after a loss, remove one after a win, never trade below the initial stake. It has no cap on how many units it adds, so the loss threshold is what ends a long run.
If you built one there, you can run it on BinaryTick: open Bot Builder, press Load and pick the XML file. The XML upload guide walks through it. To see the rules this post describes, open D'Alembert Ratchet with Edit in Bot Builder.
Run it now
Open D'Alembert Ratchet and run it on your demo balance first. Watch the stake climb a step after each loss and come down a step after each win. Then try a smaller Step size or a lower Max steps and see how the run changes. The Even / Odd bots guide covers the rest of that family, and the stop loss and take profit guide explains every limit on the card.
Ready to try it yourself?
Every bot on BinaryTick is free, and you start on a demo balance, so you can explore at your own pace.
Run a free botQuestions about this guide? Email us at support@binarytick.com.