You set the hook in five meters of sand, backed down at 1500 rpm, felt the boat pull up short, and killed the engine. Two hours later the wind has clocked forty degrees, the swell is coming in on the beam, and something feels off. Are you still holding, or has the boat been walking sideways since you opened the second bottle? Knowing the difference, in the dark, with tired eyes, is one of the harder pieces of small-boat seamanship. Here is how to actually tell.

What dragging actually looks like

Dragging is rarely dramatic. A well-set anchor that fails usually does so slowly : the flukes lose grip on a patch of weed or shell, the boat slides a few meters, resets, slides again. You can lose fifty meters in an hour and never feel a jolt. That is why relying on your gut ("we feel like we're in the same spot") is a bad idea. Reference points at night change with the tide, other boats are also swinging, and the shore lights all look the same after midnight.

Real dragging shows up in three ways :

  • Position drift in a direction inconsistent with your swing circle.
  • Change in the load on the rode, sometimes a slackening as the anchor lifts, sometimes a shuddering as it skips.
  • Change in the boat's heading behaviour, especially yawing that suddenly becomes wider or asymmetric.

Any one of these, on its own, is not proof. Two together, and you should be on deck with a torch.

The old-school checks that still work

Before any electronics, learn to use your eyes. Pick three transits when you set the hook : two objects on shore that line up (a light behind a rock, a chimney over a headland), then a second pair on a different bearing. As long as those transits hold, you are not dragging in that direction. If one opens up, you are moving.

The hand on the chain is the second classic. Rest your palm on the rode just forward of the bow roller. A well-set anchor in a steady breeze gives you a smooth, heavy tension with the occasional pulse as the boat yaws. Dragging feels different : a rhythmic vibration or a series of small jerks as the anchor skips over the bottom. Once you have felt it, you never forget it.

The third check is bearing. A hand-bearing compass on a fixed shore object, taken every fifteen or twenty minutes, will tell you within a couple of degrees whether your position has shifted beyond the normal swing. If you have not done it before, the technique is covered in more depth in our guide to how to anchor your boat safely, alongside scope, snubber and holding-ground basics.

The chartplotter anchor alarm, and why it lies

Every modern MFD has an anchor alarm. Most of them are poorly implemented. The typical version drops a waypoint at the helm position (not the anchor), lets you set a radius, and beeps if the GPS antenna crosses that radius. Problems, in order :

  1. The waypoint is at the helm, not where the anchor is sitting. On a 12-meter boat with 40 meters of chain out, the anchor is up to 45 meters ahead of the sensor. Your true swing circle is not centred on the boat.
  2. GPS jitter can trigger false alarms every few minutes, so owners set the radius huge, at which point real drag goes undetected until you are on the rocks.
  3. The alarm only sounds at the chartplotter. If you sleep in the forepeak with the door closed, you will not hear it.

The workaround, if you only have an MFD, is to walk to the bow when you drop, mark the waypoint the moment the anchor hits the bottom (not when you finish paying out), and set the radius to your maximum scope plus boat length plus a small buffer for GPS drift. Ten meters is usually enough on a decent receiver. Then leave the plotter on all night, which is a load on the house bank you need to have planned for.

What good monitoring actually looks like

The right way to monitor an anchor watch is to log position continuously, filter out GPS noise, and push an alert to the phone in your pocket, not to a screen at the helm. That means a device on the NMEA network that reads GPS, heading and depth, timestamps them, and has its own cell connection so the alert reaches you even when you have walked to the beach for lunch.

This is one of the jobs the Oria Box does out of the box. Once it is clipped onto the NMEA 2000 backbone, you draw the anchor position on the Oria app, set a radius, and get a push notification the moment the boat leaves the circle. Because it uses the boat's own 4G (SIM included), it does not care whether the chartplotter is on, whether the phone has boat wifi, or whether you are aboard. The same logic covers the boat when you are ashore and someone else's chain wraps yours, which is a different failure mode but ends the same way.

For the wider question of onboard connectivity that makes this kind of monitoring reliable, we looked at the trade-offs in Starlink or 4G for internet on your boat.

Signals from the boat you should not ignore

Beyond position, the boat itself tells you things if you know how to listen :

  • Wind shift plus heading lag. A boat at anchor should point roughly into the apparent wind, with some lag depending on windage and current. If the wind veers 30 degrees and the boat takes ten minutes to follow, you may be trawling the anchor rather than pivoting on it.
  • Depth change beyond tide. If your depth reading is drifting deeper (or shallower) faster than the tide table predicts, you are moving over the bottom.
  • New neighbours. If the boat that was two lengths off your port quarter is now on the bow, either they dragged, you dragged, or the wind shifted and everyone re-oriented. Work out which within about ninety seconds.
  • Chain angle. A rode that suddenly goes from a shallow catenary to nearly straight, with no increase in wind, means either the tide is running harder or the anchor is losing ground.

All of these are more useful when you have a baseline. A boat that logs its own behaviour over dozens of nights at anchor gives you a normal pattern to compare against. That, again, is where continuous data logging changes the game, versus the "eyeball it and hope" approach.

What to do when you confirm you are dragging

Do not immediately haul the anchor. In moderate conditions, paying out more chain often lets the anchor re-set, especially if you were on marginal scope to begin with. Try that first. If you are already on 7:1 or more and still moving, start the engine, take up slack under power to keep the rode from wrapping the keel or the prop, and recover the anchor cleanly. Reset in a better spot, or leave for a different anchorage.

The awkward case is dragging in a crowded anchorage at night with sleeping crew. Wake at least one person before you touch the windlass. Someone at the helm, someone on the foredeck, one radio between them. Losing the boat and losing crew overboard from a mid-night scramble are not comparable problems.

If you anchor regularly in areas where dragging is common, it is also worth reviewing the anchoring rules in France, because Posidonia zones, restricted areas and mandatory eco-mooring buoys change what "a good spot" even means in the Mediterranean now. And if you find yourself hunting for a berth instead of anchoring because the conditions have turned, our notes on finding a berth quickly and alternatives to ports are worth having bookmarked.

The honest answer

You know the boat is dragging when the data disagrees with your assumptions. Transits, chain feel, bearings and depth are all forms of data, and each of them requires you to be awake, on deck, and paying attention. The reason continuous monitoring matters is not that it replaces seamanship, it is that it lets you sleep, or eat ashore, without giving up the anchor watch. What would change about your night at anchor if you actually trusted the alarm to wake you before the shoreline did?