Direct answer

To identify support and resistance honestly: mark candidate levels from prior reaction points before the next move, prefer zones over lines, check the order book for actual resting depth at the level, verify the candles involved had meaningful trade counts, and decide in advance what would invalidate the level. The order-book check is what separates a level with a mechanism from a line on a picture.

Step 1 — Mark candidates before, not after

Only levels marked before the next move are evidence of anything. A level drawn after the reaction is a description of the past wearing a prediction's clothing.

Write them down, with a timestamp, before the move you intend to judge them by. This one discipline eliminates most of the self-deception in the practice.

Step 2 — Use prior reaction points

Reasonable candidates:

Prior swing highs and lows where price visibly turned.
Areas of repeated reaction, not single touches.
Round numbers, which attract attention and therefore orders.
Prior consolidation ranges where a lot of trading occurred.

Two touches make a line; any two points do. Prefer three or more reactions, and be honest that you selected them from many candidates.

Step 3 — Draw zones, not lines

Price does not respect single ticks. Mark a band covering the reaction area.

A zone is also more honest: it admits the imprecision that a single line hides, and it makes "the level held" a harder claim to fudge after the fact.

Step 4 — Check the order book

The step that turns an observation into evidence, and the one almost every guide omits.

Ask what is actually at the level:

Is there resting depth there? Look at the book directly — see how to read market depth.
How much, relative to typical trade sizes?
Is it spread across participants, or one enormous order?

A level with substantial two-sided depth has a mechanism: orders will absorb size before price proceeds. A level with nothing resting at it is a line on a picture.

Important caveat: resting orders can be cancelled instantly, and a wall placed to create an impression is typically withdrawn exactly when it would matter. Recompute depth excluding the largest single order — if the level collapses, its "support" was one participant's claim.

Step 5 — Check the candles are real

If the reactions that formed your level came from candles built from a handful of trades, the level is a handful of trades.

Check trades_count on the candles at your level, via the candles API. On thin markets this frequently disqualifies the level entirely — see reading token price action.

Step 6 — Define invalidation in advance

Before using a level, write down what would prove it wrong: a close beyond the zone by some margin, a sustained period trading through it, or depth disappearing.

A level that explains every outcome explains nothing. If price falls through and you conclude "it became resistance", check whether you would have accepted any result as confirmation.

Step 7 — Size for being wrong

Whatever your level, position size determines what a mistake costs. That decision matters more than the level's quality, and it does not depend on the level being right.

The checklist

Level marked before the move I will judge it by
Based on multiple prior reactions, not one touch
Drawn as a zone
Order book checked for actual resting depth
Depth survives excluding the largest single order
Candles forming the level have meaningful trade counts
Invalidation condition written down in advance
Position sized for the level failing

Why this is deliberately strict

The methodological problem with level identification is researcher degrees of freedom: with enough candidate levels, timeframes, and tolerance for "close enough", something always appears to have worked.

The steps above remove degrees of freedom — commit before, define invalidation, require a mechanism. What remains is either genuine or at least honestly tested.

Common mistakes

Drawing levels after the reaction.
Skipping the order-book check, leaving a line with nothing behind it.
Trusting one large wall as support.
Applying levels to illiquid tokens.
Redefining the level after it fails.
Searching timeframes until a level fits.

Educational content. Not trading advice. CoinDock publishes no signals or forecasts.

How to Identify Support and Resistance Honestly