Icicle Hitch

Most rope-to-object hitches assume a roughly uniform, reasonably grippy cylinder to wrap around - a fence post, a straight branch, a spar. Genuinely smooth or tapered objects break that assumption, and a standard rolling hitch or clove hitch tied around something narrowing toward one end tends to slide toward the narrow tip under load rather than holding firm. The icicle hitch, whose name nods to gripping something as literally smooth and tapered as an icicle, was developed specifically to solve that problem.

Its tying pattern lays four or five turns along the taper, then brings the working end back alongside the standing part leaving a long bight hanging behind the object - and that bight is passed over both ends and over the end of the object itself. That last move is the whole knot. It is what locks the turns, and it is why the hitch grips a taper where a rolling hitch simply migrates.

Riggers and arborists reach for it specifically when a load needs to be attached to something without a uniform cross-section - a tapered branch during a felling or rigging operation, a smooth pole, or any similarly awkward shape where a standard hitch's usual assumptions don't hold.

Because its advantage over simpler hitches only shows up on genuinely tapered or slick surfaces, it's largely unnecessary on ordinary parallel-sided rope or timber, where a rolling hitch does the job with far less tying complexity - the icicle hitch earns its extra steps specifically where those simpler hitches would slip.

The hitch has an unusually precise origin for a knot: it was devised by John Smith of the International Guild of Knot Tyers, and demonstrated at the Guild's eighth annual general meeting in 1990. That makes it a genuinely modern, deliberately designed knot rather than an evolved folk technique, which is part of why its tying sequence feels systematic compared with older hitches that accumulated their form over centuries of maritime use.

Getting the wrap direction correct relative to which end of the taper the load will pull toward is the single most important detail - tied backwards, the hitch can actually slide faster toward the narrow end than a simple rolling hitch would, the opposite of its intended behavior. Riggers who use it regularly always tie a test pull before committing real load, specifically checking that the hitch bites down rather than migrates.

Published sources give the turn count as a flat four or five, laid so each turn beds against the one before it. We have seen the claim that the count should scale with how sharply the object tapers, but we could not source it, so it is not repeated here. What sources do agree on is that the turns must be snugged down properly and the finishing bight drawn tight by pulling both ends perpendicular to the object - a loose set of turns is the failure mode, not an insufficient number of them.

Compared with a rolling hitch, the icicle hitch's reputation as a skill rather than a sequence comes from the finishing move rather than from any variability in turn count. Getting a long enough bight, and passing it cleanly over both ends and over the end of the object without disturbing the turns already laid, is the part that takes practice - and it is the part that, left out or fudged, leaves the hitch unfinished and free to slide.