On this page
The bias trap
Your opinion is costing you money. How often have you seen bad news over the weekend end ES futures open down big on the Globex open? By the time the US cash open comes around, the market is deeper in the red. At this point everyone is bearish and everyone thinks we're going lower, even if there's a little retracement first. The thinking goes that the overnight low isn't going to provide much support in the face of continued selling. Then reality bites. A face-ripping rally right off the open. A momentary new low, and then nonstop buying as every local tries to sell every level on the way back up, thinking at least one of them will work.
This is the most common bias trap, and I can think of at least two historic days that followed the same script: Trump's first election win in November 2016 and the hot US CPI print in October 2022. These are extremes, but they happen on a smaller scale all the time. Plenty of other bias traps exist, such as "the gap has to fill" or the classic "we've gone too far now". The quicker you can eliminate any thought of "I think the market will do X", especially when it comes with "because of Y and Z", the better. The market doesn't care. Trade the market you see, not the market you want.
From "what do I think?" to "what usually happens?"
Opinions are free, but they can cost you a fortune. The fix is to swap one question for another. Stop asking "what do I think the market will do?" and start asking "what does the market usually do from here?". The market dynamic I describe above leads to one of my favourite trade setups: the "caught in the hole" trade. A look below and failure of the overnight low, when everything is pointing bearish and we're already down big, is a great low-risk, high-reward trade, when the context is right.
Markets have habits. Sometimes it's a setup. Sometimes it's odds that show up time and time again over hundreds of sessions. That's all an edge really is. You don't need to know what happens next. You just need to know which way the odds lean, or where the majority is caught offside.
The stats I lean on
If you're not an intraday trader, stick with me. This is my example, and the next section is for you.
You all know the stats I lean on personally from an intraday perspective. I talk about them enough on the webinars and in the chat (I generally use them for ES, NQ and GC, but they apply across a lot of the US markets). For clarity, and for anyone reading this in the future, I like to wait for the Initial Balance (IB) to form: the first 60 minutes of the RTH session. You can find the full list of stats in the Market Stats space, but essentially I'm looking for two things to happen:
- The IB low is made before the IB high.
- The IB closes above its midpoint.
In this scenario the odds of touching the IB high before the IB low run at around 75–80% in the majority of the major markets I track. The opposite is true for the IB high forming first and a close below the midpoint. Those are fantastic probabilities. They don't guarantee a winning trade, of course. In fact, they don't even guarantee a trade at all. But they do give me a bias to work with until the IB is taken out in either direction. Once one side of the IB is taken out, there's a 70–80% chance we don't touch the other side in the current session. That gives me another bias to work with later.
Similar stats exist for taking out only one side of the overnight high and low. So the IB high and low and the overnight high and low give me good targets for trades if they're untested, and a bias to work with once they are. It's important to know your market, though. Not everything trades the same. For FDAX and FESX, which open at 07:00 UK, the one-side-only stat flips completely: in two out of three sessions they take out both sides of the IB. Same idea, different session, completely different behaviour. A stat that works on one market or one open can be useless, or even backwards, on another. Know the numbers for the session you actually trade.
The beauty of these stats is what they allow you to do as a trader. You no longer need to guess where the market might be going, or which way you should be positioned. They immediately give you one side of the market to focus on.
At least have a filter
Not everyone is an intraday trader. Not everyone looks at the same markets. So these stats are going to be meaningless for lots of traders. That doesn’t mean you can’t find some of your own.
Finding your own stats is simpler than it sounds. Log every trade of your setup along with its context: the trend, the time of day, where it sat against key levels. After 30 to 50 trades, look back and see where it works best and where it doesn’t. Maybe your breakouts work far better in the morning, or your pullback longs only really work when the market is above its 50-day. It’s not quite as robust as the ones I talk about, but it’s decent enough. This is where AI can really help these days: it can do all of that for you if you just plug in all your trades.
But you don’t even need statistics. What you do need, at the very least, is some sort of filter. Again, most traders struggle because they don’t know which way they should be looking. They’re long one minute and short the next.
I remember trader-turned-podcaster Anthony Crudele saying that he always had one side of the market he would trade each day. He had his rules, and if they said longs only, it was longs only. For some it might be a simple line in the sand: bullish above, bearish below. For others it might be moving averages stacked in the same direction, or an anchored VWAP (buyers in control above it, and so on).
There are many ways to create a bias, but you need one, and you need to be consistent in how you find it. Decision-making gets a lot easier when you take one side of the coin off the table. I prefer one that isn't subjective, like stacked MAs: the 10 above the 20, the 20 above the 50, for example. In that environment, imagine two periods of consolidation. In one, the market tries to break out; in the other, it tries to break down. Which break do you think is more likely to succeed?
In bullish environments, false breakdowns can be some of the most powerful setups going. The opposite is just as true in bearish regimes. Context is always king.
Trade with the wind in your sails
One filter is good. Several pointing the same way is better. Think of each one as wind in your sails: the more of it behind a trade, the less work the trade has to do.
A long on a day when the IB low was set first and the IB closed above the mid is one thing. Add an untested overnight high sitting above as a target, MAs stacked bullishly, and an LVN on the volume profile, and you've got a trade with everything behind it.
Flip it around and the same long is sailing into a headwind: IB high set first, an untested overnight low below, MAs pointing down. You might still make money, but you're fighting the odds to do it.
That's really what the alignment line on the daily card in the Market Stats space is doing: counting how much wind is behind each side. When everything lines up, be ready to strike. When it's mixed, the best trade is often no trade.
You can apply the same logic to any asset class, any timeframe. A swing long where the weekly trend is up, price pulls back to a rising 50 MA on the daily, and the sector is leading the market has the wind behind it. The same pullback in a downtrending stock, in a weak sector, is just a falling knife with a nice-looking entry.
Where stats (and traders) go wrong
Even an 80% statistic is still wrong one in five. And that's just the statistic itself. What about the trade you lost on that still made it to the target eventually? A stat is a filter, not a trigger. It tells you which side of the market to focus on, and for newbies that's half the battle.
Here's the thing. The stat only says price will touch a level. It says nothing about where you got in, how much heat you took on the way, or whether your stop got hit first. The stat can be right and your trade can still lose.
And it's not just published stats. The same goes for your own setup. If it wins 55% of the time, you'll still get runs of four or five losers in a row. That's not the setup breaking, that's just what 55% looks like.
The good news: you’d be surprised how often at least three days a week play out as expected. We track it every week on the scorecard, so you can see how often expectations meet reality, and how consistent the stats are over the past one and three months. They rarely deviate much once you go out more than a couple of weeks.
Think about each trade as one data point. You're trying to prove your edge with many data points, not with the trade itself. You are going to be wrong sometimes even when you do everything right. But you do the right things enough times and your edge will win out eventually.
Matt 'Pax' Kenah put it perfectly when he said "losses are business expenses". Reframe losses not as personal failures, but as standard overhead costs.
Mathematically, the outcome of one trade with adequate risk allocation is irrelevant. What matters is this: would I take this trade a thousand times without knowing the outcome? If not, why not? I need to know the answer: the odds, the stats.
The takeaway
The aim is to build a probabilistic mindset. Stats are just that: probabilities, not certainties. You want to think about the likely scenario, and then see if the market gives you a way in. If it doesn't, you sit on your hands. If it does, take it.
Mark Douglas made the same point in Trading in the Zone: the job isn’t to win each trade, it’s to execute well over a series of them, letting your edge play out over time.
You don't need to be right about the day. You need to know what usually happens, pick your side, and stop arguing with the market. And once the odds are on your side and you've made some money, the next job is making sure you don't hand it all back. That's the next post.