I think that you can generalize the lesson even more, into "variability is built into the robot game, and the key to success is building robust solutions that take that into account".
Consider: from game to game (on the same table), battery power level might change. Models might be put in slightly different positions, and the robot might be in a slightly different starting position.
Between tables, additional factors come into play: the illumination is different, the mats aren't the same distance from the walls, and so on.
Then there's the variability built into the equipment: traction will vary from tire to tire, there's slack in the motor gearing, and so on.
(Calibrated) sensors will help you build robust solutions, as will simple things like starting jigs and bumpers so that the robot can drive into a wall to square up itself.