I wrote most of the order-flow detection in this platform, and my own footprint chart has about two thirds of it switched off.
That sounds like a strange thing to admit in public. It's not false modesty about the features, and it isn't that the detections don't work. It's that a chart showing you everything it knows is a chart you can't read, and I had to learn that the slow way, by building all of it and then spending months quietly turning things off again.
So this is the honest version. Which signals actually earn their place on my chart, which ones I treat as context rather than triggers, and which I don't display at all. You'll disagree with some of it, and that's fine, because a lot of this comes down to what you trade and how fast.
The trap of having built everything
When you've written the detection for twenty different order-flow events, the temptation is to run all twenty. I did that for a while. Every imbalance variant on, every delta marker on, volume bubbles on, three POCs per bar, all ten alert slots firing.
What happens is predictable and slightly humiliating. The chart becomes decoration. You stop reading any individual marker because there's always a marker somewhere, and within two weeks you've trained yourself to ignore the exact thing you built. The alert sounds become background noise you mute during the open, which is the only time they matter.
The useful frame I landed on eventually: a footprint doesn't give you signals. It gives you evidence about who was aggressive and who was passive at a specific price. Most of what I turned off wasn't wrong, it just wasn't evidence I was going to act on.
The ones I actually trust
Absorption, with one condition
This is the one I'd keep if I could only keep one. Price drives into a level, aggressive orders keep hitting it, and price doesn't move. Someone's sitting there filling all of it.
The condition is that it has to happen somewhere I already cared about. Absorption in the middle of a range is just two participants disagreeing at a random price, and it resolves either way. Absorption at the session high, or at yesterday's value area edge, or at a level I'd marked before the open, is a completely different piece of information. Same detection, and its usefulness depends entirely on where it prints.
That's why absorption is worth having drawn for you rather than spotted by eye, and why the outline extending until the level gets tested again matters more than it sounds. I want to see where the passive orders were sitting twenty minutes ago, not just get a marker on the bar that's forming.
Stopping volume
A move running into a wall of trade and just halting. I trust this one because it's hard to fake and because it tends to show up before anything's visible on the candle. It's the moment a big opposing order steps in, and on a 1-minute ES chart it's usually the first hint that a drive is over rather than pausing.
I don't trade it on its own. It makes me stop adding, and it makes me pay attention to the next two bars.
Unfinished business, but as a target
A high or low that printed with both sides still active didn't finish, and the market has a habit of coming back to complete it. I've found this genuinely reliable, with one important caveat: I use it for targets, almost never for entries.
The reason is timing. "Price will probably come back here" is a useful thing to know when you're deciding where to take profit. It's a terrible thing to build an entry on, because "probably" and "eventually" don't tell you anything about the next twenty minutes. Some of the worst trades I've taken were me deciding an unfinished high was going to get filled today.
Stacked imbalances, as zones rather than triggers
A run of imbalances in the same direction means aggressive traders paid up repeatedly through that range and cleared out what was resting there. I care about the zone it leaves behind, not the moment it prints.
Marking those zones and then watching how price behaves on the retest is, for me, the single most practical thing a footprint does. It converts "there was buying here" into a specific price band where I'll find out whether that buying is still around.
The ones I keep on, but don't believe on their own
Delta divergence
Price makes a higher high, delta doesn't. The buyers pushing the move are stepping back. It's a real thing and I do watch it.
It's also the signal I've seen cost people the most money, because in a genuine trend it fires over and over on the way up. If you treat every divergence as a reversal you'll spend an entire trend day getting run over, and you'll be right on the last one, which somehow feels like vindication.
I use it as a reason to tighten a stop or stop adding. Almost never as a reason to reverse. If I'm going to act against a trend on a divergence I want absorption or stopping volume in the same place, and I want it at a level that already mattered before the divergence showed up.
Exhaustion
Aggressive activity thinning out at an extreme. I like it conceptually and it's genuinely useful late in a move. My problem is that "the aggressive side stopped showing up" and "everyone went quiet for a minute" look identical on a slow tape. It's excellent at the open and around a level, and much less trustworthy at 11:40 when nothing's happening anywhere.
Big delta and market sweeps
Someone tearing through several levels in one motion, or an outsized imbalance at a single price. This is real intent and I want to know about it. What I've learned is that it tells me somebody's in a hurry, not which direction is right. Sweeps into a level are one thing. Sweeps out of a level are another. The detection can't know the difference, so I treat both as "pay attention" rather than as a direction.
What I don't display at all
Most of the delta bar markers. There's a family of them, rise, drop, flip, reversal, tail, trap, sweep, slingshot, and they're all detecting something real. I found that on a 1-minute chart they produced more marks than bars, and I was reading the marks instead of the tape. If you're on 5-minute or higher they make far more sense, and plenty of people use them happily. On my chart they're off.
POC 2 and POC 3. Up to three points of control per bar is genuinely useful for seeing where volume split inside a bar. I found I only ever acted on the main one. The others became visual noise I was scanning past.
Volume bubbles, except on one template. Bubbles sized by trade size are great when you're specifically hunting large orders, and the Big trades template exists for exactly that. As a permanent overlay on top of cells and imbalances, they were the thing that finally made me admit my chart had become unreadable.
Most of the alert slots. There are ten. I run three. Stacked imbalance at a level, absorption, stopping volume, and only during the sessions I actually trade. The moment I had six or seven going, I stopped hearing any of them.
Two templates, not fourteen
There are fourteen one-click templates and I use two.
Bid/Ask when I'm studying a level, because I want the raw numbers on both sides of the row and I'm willing to work for them. And Volume Cluster when a trade is running, because at that point I don't want to read anything, I want to see where size is and whether it's building against me. Solid blocks, no digits.
The reason there are fourteen isn't that anyone needs fourteen. It's that the two you want depend on what you trade and what your eyes do, and I'd rather ship the templates than have everyone hand-build a look from forty settings. Almost every trader I've watched work settles on two. Which two varies enormously, and I stopped trying to predict it.
The related setting nobody uses enough is tick aggregation. Grouping several ticks into one row turns an unreadable wall of cells into something with shape. On CL I run it aggregated basically always. On ES I don't. If you've looked at a footprint on a wide-tick instrument and concluded footprints are useless, that's probably the setting you were missing.
The thing that actually changed my results
None of the above.
What changed things was using the footprint to decide where I was wrong instead of where to get in. That sounds like a platitude, so here's the concrete version: I stopped putting a stop at a round number of ticks and started putting it beyond the price where the evidence would be invalidated. If I'm long because a level absorbed selling, the trade is wrong when price trades meaningfully through the absorption rows. Not eight ticks. There.
That one change did more for me than every detection module combined, and it's the part I'd want a new order-flow trader to take away. The signals help you find situations. Knowing exactly what would prove you wrong is what makes the situations tradeable.
The last thing I'd say is that I read the footprint less than I used to, because several of these reads now live inside strategies instead of my head. The plots are available as building blocks, so absorption at a level can be a condition in a rule rather than something I have to be present for. I wrote about how I use that side of things here. It hasn't made me a better tape reader. It's just meant the setups I trust don't require me to be awake.
If you want the full list of what's in there, including all the things I've admitted to turning off, it's on the footprint page.