Building a rotation plan that actually holds.
Rotation plans fail in year two, not year one. The difference between a plan that holds and one that quietly reverts is who built it.
The year two problem
Year one of a new rotation runs on the enthusiasm of having decided something. Year two runs on habit, and habit reverts to whatever the operation did before unless the plan is genuinely owned by the people executing it.
This is why plans authored externally and handed over have such a poor survival rate.
Build it with the operators
The people who will run the rotation should build the model, with us alongside rather than in front. It takes longer and produces a plan that survives contact with a difficult season.
They will also catch constraints no external analysis would surface — which gate is impassable after rain, which block cannot be worked in sequence.
Leave room to be wrong
A rotation with no slack is a rotation that breaks the first time the weather does not cooperate. Building in explicit flexibility on a subset of blocks costs a little theoretical yield and buys a plan that bends instead of snapping.
The best plans we have seen are slightly less optimal on paper and considerably more durable in practice.
Recording the season honestly
The value of a rotation model is entirely in the quality of the actuals fed back into it. Seasons recorded optimistically produce a model that flatters the same mistakes each year.
When to revise versus hold
One bad season is weather. Two in the same direction is the plan. Resisting the urge to redesign after a single disappointing year is most of the discipline.
Handing it over
The plan is finished when the person who will run it next season can explain why each block is where it is in the sequence. If they cannot, it is still our plan rather than theirs.
Written by
