Climbing
Anchors: the principles, not the acronym
Every climber learns a mnemonic for anchor building and most can recite it. Far fewer can explain why each letter is in there, which is the part that matters when the rock does not cooperate.
There is a mnemonic. There are several, depending on who taught you, and they are broadly fine as a memory aid. The trouble is that people memorise the letters and never internalise the reasoning, so when the stance is awkward and nothing quite fits the textbook picture, they have no way to reason from first principles.
So here are the principles, and then why each one exists.
Strong placements first
Nothing downstream fixes bad placements. Three poor pieces do not add up to one good one, because they will fail sequentially rather than together, each failure shock-loading the next.
Judge each piece as if it were the only one. Is the rock solid — does it ring hollow, is it detached, is there a crack behind it that runs the wrong way? Is the placement oriented to take load in the direction load will actually come? A perfect nut set for a downward pull is worth very little if the anchor will be pulled sideways or up.
Redundancy at every link
Look at the whole system and ask: if any single component vanished right now, would anything catastrophic happen? Any component. The sling, one carabiner, one piece, the knot, the loop of cord.
This is where people find the flaw in an anchor they thought was bomber. Two excellent pieces joined into a single carabiner is not redundant at that carabiner. A cordelette with an abrasion running over an edge is not redundant at the edge. The chain is only as good as its narrowest single point.
The usual exception, applied thoughtfully, is a single large modern bolt or a substantial tree — components with strength far beyond anything the system will see. Even then, most people build off two.
Load distribution
You want the pieces sharing the load rather than one taking it all until it fails. Perfectly equal sharing is a myth — rope stretch, sling angle and slight differences in placement mean real anchors are never truly equalised — but roughly even is achievable and worth having.
The part people get wrong is angle. As the angle between two legs of an anchor widens, the force on each piece increases, and beyond roughly ninety degrees it climbs sharply. A wide V between two pieces can put more force on each piece than the total load applied. Keep the angle narrow. If the pieces are far apart, extend the anchor downward to bring the angle in, rather than accepting a wide spread because it looks tidy.
No extension
If one piece fails, the remaining pieces should not receive a shock load from the system suddenly lengthening. A sliding-X without limiter knots does exactly that: lose a piece and the whole thing drops several centimetres onto whatever is left, at speed.
Limiter knots solve it. A fixed-point anchor — legs tied off with a knot rather than left to slide — solves it. The trade is that a fixed anchor equalises only for the direction you tied it for, which brings us to the last principle.
Direction of pull
An anchor is built for a direction. Belaying a second coming up produces downward load. Belaying a leader going up produces upward load, potentially violently so if they fall before the first piece. Those are different anchors, or one anchor with pieces placed to handle both.
This is the most commonly neglected principle and the most likely to actually fail in practice. A beautifully constructed downward-facing anchor can be lifted straight out by an upward pull. If you are about to belay a leader, put a piece in that resists upward force, and put it in before they leave.
The trade-offs nobody mentions
You cannot maximise all of these at once. Perfect redistribution conflicts with no-extension. Total redundancy conflicts with speed, and speed is itself a safety factor on a long route where weather is the real threat.
Experienced climbers make these trades consciously. On a bolted belay with two good bolts, two carabiners and the rope, thirty seconds, done. On a marginal gear stance in a chimney with weather coming, more pieces and more thought. The anchor is not a ritual to be reproduced identically; it is a response to a specific situation.
Test what you build
Before you commit, weight it. Sit back on the anchor, hard, in the direction you expect load. Watch every piece. Things that are going to shift usually shift under body weight, and finding that out while standing on a ledge is an entirely different experience from finding it out during a fall.
