Changing the surface changes which tissue takes the impact

Ground hardness and shoe traction decide how much force is returned to the body and how easily a planted foot can release.

A squad moves indoors for a fortnight of bad weather and the physio room fills up with calves and knees. Nothing about the training changed. The floor did, and the floor is part of the athlete's mechanics whether anyone accounts for it or not.

Two separate properties are involved and they get muddled. The first is stiffness in the vertical direction. A hard surface returns energy quickly and deforms very little, so the deceleration on landing happens over a shorter distance and the peak force rises. The body compensates by adjusting leg stiffness, flexing more at the knee and ankle to soften the arrival, and that compensation costs muscular work in the calf and quadriceps that the surface used to do. Hence the calf soreness after a fortnight on concrete.

The second property is horizontal, and it is the one that hurts knees. Traction determines how firmly a planted foot is fixed to the ground. When the shoe grips completely and the athlete's body continues rotating, the rotation has to be absorbed somewhere above the foot, and the first joint capable of absorbing it is the knee, in a direction it does not like. Artificial surfaces combined with long studs are the classic version of this, and it is why the shoe and the surface only make sense as a pair. A boot that behaves perfectly on soft ground becomes a fixing point on a dry pitch in August.

Which is my objection to the way boot choice is treated as personal preference. It is a mechanical setting, and it should change with conditions.

The other underrated variable is uniformity. A cambered road loads one leg differently from the other for the entire run, and a runner doing the same loop in the same direction for a year has been quietly building an asymmetry. Trail is uneven in a way that varies, which spreads the load across tissues rather than concentrating it.

The transition is the risky part rather than the surface itself, because a body that has spent months on grass has tuned its stiffness for grass.

Give any surface change the same respect you would give a change in volume, and treat the first session on it as the exposure it actually is.