It's hard to say w/o knowing the details of the experiment and result. For example, if everyone got an error that was 50% above the correct value, that would point to systemic error. If everyone got errors that were equally distributed from -50% to +50%, I would say that the errors are more likely random.