The Indicators IronGate Uses
Indicators aren't magic — each one is just an honest, repeatable answer to one question about recent price. Here are the four your bot actually leans on, and the question each one answers.
6.1Moving averages — "where has price been lately?"
A smooths the last N closes into one line — the market's recent centre of gravity. Use two of different speeds and think of them as boats: the fast EMA is a speedboat, swerving with every wave; the slow EMA is an oil tanker, turning only for a real change of course. When the speedboat overtakes the tanker, something recent is genuinely stronger than the established course — and that overtake, the cross, is the MA Crossover strategy's entire signal.
⚙ EMA crossover explorer — drag the periods
Make the fast EMA very fast: more crosses, more false starts. Slow both down: fewer, later, but more reliable signals. This is the same noise-vs-lag trade-off as timeframes — it never goes away, you only choose your position on it. (Your card's "EMA 10/50" is a classic middle choice.)
6.2ATR — "how big are the waves right now?"
Picture the market as the sea, and every candle as one wave. Some days the water is glassy — waves of a few pips. Some days it's a storm — waves ten times taller. is simply the average height of the recent waves. It's direction-blind on purpose: a wave is a wave whether it lifts you or drops you.
Now the part that matters: your stop loss is a mooring line. Tie your boat too tight in heavy swell and an ordinary wave — not a storm, just the sea being the sea — rips the cleat out: that's being stopped out by normal noise. Your bot ties the line at 1.5 waves of slack, so only a genuine storm surge (a real move against you) can pull it.
And that's not a metaphor stretched over something complicated — it's literally the computation: ATR = the average height of the last 14 candles. Prove it to yourself:
⚙ ATR lab — raise and calm the sea
The amber ruler is the ATR — the average wave height. When you hear "ATR 17 pips," picture the sea state: typical waves of about 17 pips right now. The stop is just the mooring line, tied 1.5 waves out — and Lesson 1's formula shrinks the boat (the lot) when the sea gets big, so a snapped line always costs the same 1%.
6.3ADX — "is this a trend worth respecting?"
is harder to picture because it isn't a distance — it's a ratio. Here's the picture that makes it click. Two hikers each walk exactly 10 km today. Hiker A walks mostly north and ends 8 km from camp. Hiker B zigzags all day and ends 1 km from camp. Same effort — completely different progress.
ADX asks: how much of the market's walking was actual progress? Lots of movement that goes somewhere = high ADX. Lots of movement that cancels itself out = low ADX. That's why it says nothing about direction — a determined march south scores just as high as one north.
⚙ ADX lab — same walking, different progress
Every path here contains the same total movement — only the agreement between the steps changes. Now re-read Lesson 5's dial: "ADX 34" just means "recent bars mostly agreed on a direction." The bot uses it twice: scoring pairs (strong trend = opportunity) and classifying regime (which strategies may compete).
6.4RSI — "has this move stretched too far?"
measures how one-sided recent closes have been, on a 0–100 dial. Think of a rubber band: every close in the same direction stretches it further. Above 70 or below 30, the band is stretched hard — and in a ranging market, stretched bands snap back:
6.5The specialist instruments
The four above are the principal players — they drive the recommendation engine itself. But several strategies hire session musicians for their particular sound. You'll meet these inside Sentinel, BreakoutGuard, RevertX and the scalpers, so here's a proper introduction to each:
— the gap-meter between the boats. You know the speedboat and the tanker; MACD watches the gap between them. Gap widening = the move is accelerating; gap shrinking = it's tiring, even while still "winning." The histogram bars draw that gap directly:
— the room's statistical walls. An envelope two standard deviations around a 20-bar average: wide in rough seas, narrow in calm. The narrowing — the squeeze — is Lesson 7's crowd pressing at the door (BreakoutGuard), and the band edges are the walls RevertX fades from.
— the second rubber band. Like RSI but asks a slightly different question: where is the close sitting inside the recent high–low span? RevertX demands both bands agree before it trades — two witnesses, one story.
— noise-cancelling headphones for candles. Each candle is averaged with its neighbour, smoothing the flickers so trends read as long unbroken runs of one colour. Sentinel reads these instead of raw candles when judging trend health.
— the traffic light. A high/low moving-average channel that flips a simple state: price above = green (long side), below = red (short side). One of Sentinel's five lookouts — the one that just says which side of the road we're on.
— the scout. A momentum dot engineered to turn with minimal delay. Where heavy averages confirm late, the scout signals early turns — another Sentinel lookout.
— the seismograph. Combines the MACD gap with band width to measure how much directional energy a move carries, with a dead zone below which Sentinel simply won't trade: no energy, no trade — the seismograph must actually shake.
— the speedometer, raw. Price now versus price N bars ago. No smoothing, no cleverness — which is exactly why the Scalping strategy uses it: at seagull speed, you want the instrument with zero lag.
— the crowd-noise meter. How many price updates arrived per bar — a proxy for activity. BreakoutGuard requires the crowd to actually roar on the breakout candle: a door bursting open in silence is usually a false exit.
6.6The cheat sheet
Moving averages — the speedboat and the tanker → did something recent overtake the established course? (drives MA Crossover)
ATR — the wave height → how rough is the sea, and how much mooring slack does the stop need? (sizes every stop, and therefore every lot)
ADX — the hiker → how much of the walking was actual progress? (gates which strategies may trade)
RSI — the rubber band → how hard is it stretched, and is it free to snap back? (drives the mean-reverters — in ranges only)
And the specialists: MACD the gap-meter · Bollinger the walls · Stochastic the second band · Heiken Ashi the headphones · SSL the traffic light · NonLagDot the scout · Waddah the seismograph · Momentum the raw speedometer · Tick volume the crowd noise.
No indicator predicts. They all describe the recent past in a way that's precise enough to build rules on. The edge never comes from a magic indicator — it comes from matching the right rule to the right regime and sizing it so you survive being wrong. You now know every ingredient on your recommendation cards.