Worse(*) than that: even with balanced dice, how you throw them could have some significant influence on the results. Cubic dice tend to roll forward and not turn sideways, so if you do this carefully, you might make an opposite pair of numbers (1 and 6, or 2 and 5, or 3 and 4) less likely than the others.
This could be used to avoid extreme results (make 1,6 less likely), or to increase the chances of extreme results (make 3-4 less likely), or even to decrease the chances of cowering (make different pairs less likely with the two dice). If done expertly (say, so as to make the "less likely" pair never appear), this would have a much more significant influence on a game outcome than poorly balanced dice (here I am thinking of poorly balanced, not outright weighted dice - something like what would be possible with non-precision dice such as those found in ASL boxes).
I suspect it would be more difficult with, say, a dice tower than just rolling dice into a flat surface, and I admit I am no expert myself.
There is a pretty well-known mathematician (probabilist and statistician) named Persi Diaconis who, in an earlier career, used to be a professional magician; I know he has done some work on randomness, esp. coin flipping, but I have not managed to find anything of his on fixing dice rolls. He has a few videos on the Numberphile channel, though.
(*) I realize I said "worse", but really this is something that is unlikely to happen without the player deliberately attempting to cheat; and it seems to me that somone trying to cheat at ASL probably could do it more easily without becoming an expert at fixing his dice rolling.