Ready to go further
Next: see how survivors skew an average
The survivorship bias guide shows how a list of what remains can make the typical result look better than it was.
What is included
- What it is, in plain words
- The formula or mechanism, with exact arithmetic
- How it hits grid, martingale and DCA bots
- A check you can run before you switch a bot on
Who it is for
- People who run or plan to run a trading bot
- Anyone reading a backtest or a bot sales page
- Readers who want the numbers, not the pitch
- Not advice, and not a signal service
Key facts
- Collective2 is described here only as a type of product: a strategy marketplace or signal listing, where someone publishes a strategy and a track record and others can follow its trades.
- A listing shows a record, not the reason for it. A record can be short, self-reported, from a different account size, or kept only for the strategies that survived.
- Survivorship is the biggest trap. In a made-up group of 200 strategies, the 80 still listed averaged +14% and the whole group averaged -7.6% (calculated by us on 31 Oct 2026, exact). See survivorship bias.
- Fees come off the result every month whether the strategy wins or not. A made-up $50 a month on a $5,000 account is 12% a year, which would take a 14% year down to 2%.
- Following a listing means each trade is copied after the strategy's own, with its own price. Delay and slippage cost the follower extra. See copy trading. We do not offer or sell signals.
What a strategy marketplace is
A marketplace of this type lets a trader publish a strategy, show its history and charge others to follow it. The follower's account then places the same trades, either automatically through a broker connection or by hand from alerts. The platform sits in the middle. It lists the strategies, displays their statistics and usually handles the fee. How each product does this, what it checks and what it charges are facts that change, so read the vendor's current pages and note the date.
The key point is that you are choosing from a list. Someone built the list, and the order depends on a statistic such as return or rank. A list sorted by past return rewards luck as much as skill. Before you read any listing, find out what the page will and will not tell you about how the record was made.
The general risks of a listing
Short records
A year of results is one market mood. A strategy that has not met a sharp reversal or a long sideways period has not been tested on them. A short record also has a wide margin of error: a few good weeks can drive the whole figure.
Self-reported or selected records
Some records come from a trading account linked to the platform and some come from the person listing it. Ask which kind you are looking at and whether the platform can show trades that were closed, deleted or moved. A record built only from the strategies that kept going is not the record of the whole market place. See survivorship bias.
Size and risk hidden behind percent
A percent return says nothing about how much was risked to get it. Read the largest drawdown, and the lowest point of open losses, not only the closed trades. See maximum drawdown and equity curve.
Worked example: survivors against the whole group
Everything here is made up. Suppose 200 strategies were listed at the start of a year. By the end, 120 had stopped or been removed after losing money, and 80 were still listed. The 80 survivors averaged +14 percent. The 120 that stopped averaged -22 percent. A reader who looks at today's list sees only the first group.
| Group | Strategies | Average year (percent) | Weighted contribution |
|---|---|---|---|
| Still listed | 80 | +14 | +1120 |
| Stopped | 120 | -22 | -2640 |
| Everything that started | 200 | -7.6 | -1520 |
Calculated by us on 31 Oct 2026, exact. The list a visitor sees averages +14 percent. The group the visitor would have picked from a year earlier averaged -7.6 percent. The weighted contribution column is each group's number of strategies times its average, and the last row divides the total by 200 to get the average. Real proportions are unknown. The point is the direction: leaving out the failures raises the average.
Then add the fee
A fee reduces what a follower keeps even in a good year. Suppose a made-up fee of $50 a month on a $5,000 account: $600 a year, or 12 percent of the account. A strategy that made 14 percent before the fee leaves 2 percent. That is before the follower's own slippage and before tax.
How it hits bots
- Automated strategies are often listed with a smooth line. A grid or martingale style can win small amounts for months and then lose a large amount at once. See martingale.
- Changes in version: a record may be the result of several versions of a bot joined together. Ask whether the rules changed during the record.
- Delay: a follower's order is a second order. The faster the strategy, the more of its edge disappears. See slippage.
What to verify on the vendor's current pages
- How a record is built: linked account, manually entered, or simulated. Simulated results should be labelled as such.
- Whether stopped or deleted strategies still show in the statistics or the search.
- What fees apply to followers and to the person listing, and when they are charged. Note the date you read them.
- How trades reach your account, what happens when a trade is missed, and what size rules apply.
- What the terms say about losses, refunds and the platform's responsibility.
- How long the record is, and the largest drawdown, not only the return.
What this page does not tell you
It does not say whether any listing or marketplace is good or poor. We have not run the product, and it states nothing about its fees, strategies or users. It says how a record can mislead. The numbers are made up and nothing here is a recommendation.
Keep Going
Related Tools and Guides
Keep exploring
FAQ
Frequently Asked Questions
What kind of product is Collective2?
In general terms it belongs to the class of strategy marketplaces or signal listings, where strategies are published with a track record. We have not run it, so check its current pages for how it works.
Why can a list of strategies be misleading?
The list often shows only strategies that survived. In our made-up example the survivors averaged +14 percent while the whole group averaged -7.6 percent.
How do fees change a result?
A fixed monthly fee is charged win or lose. In our made-up example a $50 monthly fee on a $5,000 account removed 12 percent a year.
What should I check on the vendor pages?
How records are built, whether stopped strategies are counted, the current fees, how trades reach your account and what the terms say. Note the date you read them.
Can following a strategy lose money?
Yes. Past or simulated results are not a guide to future results. You can lose all the money you put in.
Are the numbers on this page real?
Example numbers are made up to show how the maths works, and the page says so. A figure only counts as exact when the page says it was calculated on a stated date. Nothing you enter in a calculator leaves your browser.
Can I use this to decide whether to run a bot?
It can show whether a bot's numbers hold together and how much an account can lose. It cannot say what the market will do, and it is not advice. Use it to ask better questions, then read the terms of your own broker or exchange.
Is this financial advice?
No. It is general education. All example numbers are made up and nothing here is a recommendation to follow any strategy.
Know the Risk Before You Run It
Free calculators and guides, added every week. Leave your details and we will email you when new ones go up.
Trading carries a risk of loss