A conviction based solely on identification evidence of poor quality cannot stand in law. Where identification rests on a brief observation (15 seconds) of a stranger in difficult conditions (poor lighting, nighttime, stress, obstruction), made by a single witness without corroboration, and where no formal identification procedures were followed, the risk of mistaken identification is too high to justify a conviction. The quality of identification evidence must be rigorously assessed against the factors outlined in S v Mthetwa and R v Turnbull, including duration of observation, lighting, distance, prior knowledge of suspect, stress factors, and corroboration. When such evidence is weak, other supporting evidence is essential; absent such support, the accused must receive the benefit of doubt. Furthermore, an accused bears no onus to prove an alibi; the burden remains on the State to disprove it beyond reasonable doubt. A trial court errs in law by dismissing an alibi for lack of corroboration by the defence or requiring the accused to produce proof of the alibi. Where identification evidence is suspect and an alibi raises a reasonable possibility that the accused was elsewhere, the conviction cannot be sustained.