Some consonant pairs are doors. Others are walls. They look identical until you try them.
A consonant cluster is two consonants next to each other in a four-letter word, and in a ladder it is either a door or a wall. Doors are clusters that appear in many words with many neighbours; walls are clusters that appear in a handful of words with nowhere to go. The two look identical when you scan a board, which is why clusters deserve a separate lesson.
Take a common cluster like the one at the end of a word ending in a doubled consonant. Words with it are legal, familiar, and structurally nearly isolated: change one of the doubled letters and the word stops existing. A route through that cluster usually costs a rung and returns nothing. Meanwhile a cluster in the middle of a word that appears in several unrelated words is a junction, and walking through it opens options.
The way to learn the difference is not a table. It is a handful of deliberate failures. Play a few ladders where you intentionally take cluster words and watch what happens. The failure rate is informative, and the memory of a specific trap is stronger than any rule.
After a week of noticing, the scan starts to skip certain shapes automatically. That is the goal: not a dictionary of clusters, but an eye that recognises walls.
For a low-stakes experiment, visit the site and take the first cluster word you see on purpose. Then practice whenever you like until the difference between the door and the wall is a reflex rather than a rule.
The cluster lesson also transfers to the endgame. Final hops often run through a cluster, and knowing which clusters are doors tells you whether the last two rungs are realistic. A route that leaves a wall as its final step is not a route; it is a hope with a budget.