Every fastener I added was a decision I had failed to make

Using more bolts as a substitute for understanding the joint.

Adding extra fasteners to a joint that already had enough is usually a substitute for actually understanding how that joint carries its load, since a bolt count that has crept upward out of caution, without a specific reason attached to each additional one, is standing in for an analysis that was never properly done, and it rarely makes the joint any stronger.

A key ring nobody has sorted

Carrying a huge key ring loaded with keys for doors that may not even exist any more shows the same pattern in a different setting. Nobody carrying that ring can say for certain which keys are still needed and which are relics of a lock that was changed or a building left years ago. Working that out and discarding the rest takes effort, so in the moment it is simpler to add each new key to the same ring and let the old ones ride along indefinitely. The weight in the pocket makes nobody more secure or better prepared. It stands in for a decision about locks that was never made, much as an ever-growing fastener count stands in for a load analysis that was never done.

Where a joint's strength actually runs out

A well-understood joint has a specific number of fasteners because someone worked out, or measured, how much load the joint needs to carry and sized the fastening to meet it with an appropriate margin. Every bolt in such a joint is there for a reason that could be stated plainly if asked.

A joint whose bolt count was reached differently, by starting from a rough guess and adding more whenever the designer felt uneasy, ends up with a number that records the designer's uncertainty more than the joint's requirement. Past a certain point more fasteners stop adding useful strength anyway, because a joint's real capacity is often limited by something else: the strength of the surrounding material, the stiffness of the parts being clamped, or the share of the load path the fasteners are able to intercept. None of these improves because another hole was drilled and another bolt fitted into it, and each new hole removes a little of the material the load has to pass through.

Stiffness decides more than intuition expects. Where the parts are stiff, fasteners share the load roughly as the drawing assumes, and where they are flexible, the plate bends between fixings and the load gathers at whichever fastener the bending reaches first, so adding one more mainly changes where the trouble concentrates.

Asking what each bolt is for

Before a design is finalised, a useful review asks of every bolt on the drawing whether there is a specific, statable reason for it, tied to an identifiable share of the joint's load. A bolt that cannot be justified that way usually adds weight, cost and assembly time in exchange for a feeling of safety, and a proper analysis would either confirm it was unnecessary or, more usefully, replace the feeling with a figure worth trusting.

Extra fasteners are sometimes exactly right. A deliberate margin added to cover a specific, known uncertainty, an assumption about load that could not be measured directly, a material property with real variability, a manufacturing tolerance that might shift the load path, is a different thing from an unexamined habit of adding bolts whenever a joint feels uncertain. The test is whether the extra capacity was chosen on purpose, with a reason that could be written down and defended, or simply accumulated because nobody stopped to check. A margin with a stated reason behind it is engineering. A margin without one, however sensible it feels at the time, is a guess wearing engineering's clothes.

Turning caution into analysis

Telling the two apart is one of the clearer signs of a designer maturing past their earliest instincts. The temptation to over-fasten is strongest precisely when someone does not yet trust their own analysis, and the honest fix is to redirect that caution into doing the analysis, so that a real figure says whether the joint needs another bolt. The caution itself is healthy. Spent on an extra hole drilled because it felt safer than leaving well alone, it buys very little, and spent on working out where the load really goes, it can show that the problem lies in the parts between the fasteners.

My key error with this

The sensor mounts kept cracking, so I kept adding fasteners, on the reasoning that a joint which is coming apart needs more of whatever is holding it together. Each additional fastener produced a part that cracked somewhere slightly different a little later, which I read as progress rather than as the same failure moving around. The reason it kept moving is that the components involved were soft plastic and I had been treating the whole stack as though it were rigid, so my mental model of the load path, in which fasteners share a load between stiff bodies, did not describe what was actually happening, where a compliant plate deformed between the fixings and concentrated strain at whichever one had most recently been added. Nothing improved until I stopped adding fasteners and redesigned the plate itself. What replaced the belief is that every fastener I add without being able to say what load it carries is a decision I have declined to make, and that reaching for another one is usually evidence that the real problem is the geometry between them.