That's exactly how it works....a 2.8 % prob of winning is equivalent to 1 positive outcome in approximately 35 tries. You are correct is saying that if we lose 10 (completely random) games in a row, that does not change the likelihood of winning/losing the next, but that is a separate argument.
Or, In layman's terms....In a game where mediocre teams beat very good teams on a fairly regular basis...it's just plain foolish to say that a mediocre team has virtually zero chance to beat a very good team. It may not happen to us but it has already happened multiple times in week one and will continue to do so as the season goes on.