Trading simulator:
what it needs to be worth your time

Price has to move like the market

A random number generator produces a chart that looks like a chart and does not behave like one. Four things happen in real markets that a smooth random walk does not do, and they are exactly the things you need to learn to read.

It changes regime

There are stretches where price is going somewhere and stretches where it bounces inside a range. They alternate and they last as long as they last. Trading a range with trend technique is the most common way to lose slowly, and the only way to learn the difference is to see a lot of both.

Volatility comes in clusters

After a big candle come more big candles. After half an hour of flat it usually stays flat. Books call it volatility clustering, and it has a practical consequence: the same ten point stop is generous mid-session and laughable at the open.

The open is not like the rest

The first minutes of New York move several times more than late morning. A simulator with constant volatility trains you for a market that does not exist.

There are sweeps

Price goes looking for the stops sitting just under the previous low, collects them and turns around. If the simulator never does this, you get used to putting your stop where everyone else puts theirs, and live it will get taken. The candlestick guide shows how to spot one.

If the simulator does none of this, you are practising on noise: you can get very good at a game that does not resemble the market.

Trading has to cost money

This is the point most simulators fail and the most expensive one, because the error is invisible: your stats look good right up until the broker starts charging.

Every trade pays two things. The spread, the gap between what buyers bid and sellers ask: you buy at the top and sell at the bottom, so you are born losing. And the round turn commission the broker charges to open and close.

With the Micro Nasdaq, which is what most people practise with, the numbers are these:

You should not always get the price you ask for

When the market runs, price has moved between your click and your fill. That is slippage.

What matters is not that it exists, it is when it shows up: on the macro release, at the start of the impulse, when your stop gets hit. In other words, exactly on the trades that weigh most in your result. A simulator that always gives you the exact price is handing you the best entries of the day.

How to spot it

Put in a market order in the middle of a big candle and look at your fill. If it is exactly the last price you saw, execution is fake.

There has to be a consequence

This is what separates practising from killing time, and it is the fundamental difference between a demo and a tournament.

If you can blow the account and reload, the part of your brain that learns never finds out. Learning needs an error signal, and "nothing happened" is not a signal.

What counts as consequence

• A loss limit that throws you out of the session. Not a warning: out.
• A cut that leaves you out of the competition if you are not among those going through.
• A public scoreboard where your placing shows.
• A clock. Without one anybody trades well, because you can always wait for price to come back.

What does not count: losing play money with nothing attached. Nothing hurts there, and what does not hurt is not learned.

The same market for everyone

If everyone plays different data, your result means nothing. You can finish on top because you got an easy tape.

With the same tape for all, the comparison says something: you traded exactly what everyone else traded, and whoever finished ahead did better. That turns the scoreboard into information instead of decoration.

It is also the only way competing makes sense: in BullHero, everyone entering the same fifteen minute slot gets the same price tape.

Sessions have to be short

You learn by repetition with immediate feedback, not by accumulated hours. Three minutes, see the result, understand what happened, again.

A demo open all afternoon gives you three or four decisions and none of them matter. A three minute format with a cut gives you forty decisions in the same stretch and all of them count. The difference in learning speed is huge and has nothing to do with talent.

Here is the full calculation of how many reps you need and how long it takes with each format.

It has to be able to go badly

If most of your sessions end green, something is miscalibrated.

Live, a beginner loses most sessions. That is normal and it is information: it tells you that you do not have an edge yet. A simulator that agrees with you is setting you up for a punch, because you arrive live convinced you know, carrying statistics that were never yours.

The warning sign

If you have done thirty sessions and lost three, you are not good: the simulator is soft. Check the costs, check whether price sweeps, and check whether there is a loss limit.

The ten minute test

Open whatever simulator you are looking at and do this in order. Ten minutes and you know whether it deserves your time.

1. Enter and exit immediately, with price unmoved. If you lose nothing, there are no costs. Fail.
2. Watch thirty minutes of chart straight. There have to be quiet stretches and nervous ones. If everything moves the same, it is noise.
3. Find a clear previous low. Wait and see whether price goes for it and turns. If that never happens, there are no sweeps.
4. Lose on purpose until you blow through the limit. If nothing happens, there is no consequence.
5. Check whether you can rewind. In practice mode fine; in competition it should not let you.
6. Look at the default size. If it hands you a hundred thousand, set it to what you will actually have. That alone changes how you trade.

With those six points you can judge any of them, ours included. Better to find out a simulator is soft in ten minutes than in six months.

Frequently asked questions