The steering gear has a little play anyway by virtue of its operation. There is stub shaft that has the output spline that connects to your steering column. On the stub shaft is a cylindrical valve spool inside of a cylindrical valve body. The stub shaft/valve spool and the valve body are connected to the worm ball drive. If the worm drive were held fixed the valve body won't move but the stub shaft and valve spool can torsion twist a little. So the valve spool could move relative to the valve body.
The valve spool and valve body form the hydraulic passages to each end of the piston. When there is no resistance on the pitman shaft, like sitting on a work bench, the stub shaft, the valve spool, the valve body, and the worm all move together to drive the piston and hence the pitman shaft.
When there is resistance, like the tires on the ground, the stub shaft acts like a torsion bar first twisting against the resistance of the worm to move. The valve spool moves relative to the valve body. This positions the valve spool to block some fluid passages from return and send hydraulic pressure to one end of the piston offering steering assist to move the piston and turn the pitman. The worm drive turns too and so does the stub shaft (your turning it) and the valve body. When you let off the wheel coming out of the turn, the torsion bar effect releases and allows the valve spool and valve body flow passages to realign to neutral and provide the same return pressure to both sides of the piston. The road resistance allows the wheels to straighten out and spin everything back to straight (or you could power assist it by turning the wheel).
The amount of torsion twist is small but even at 2 degrees on a 15.5" steering wheel is about 3/10 of an inch along the circumference of the steering wheel. So I would imagine there is always some play. Sorry for the dissertation but the steering gear is sort of a really cool device.