Apparently not in this case, though.
If a prosecutor bets the whole ball of wax on a first degree murder conviction, when the evidence may only support (or more likely support a conviction on a lesser charge) he risks confusing the jury and getting a loss. That is what I think happened here.