Predicting save opportunities is really a crapshoot. Usually you want closers on good teams, since they win more games, but if they aren't winning close games, there won't be the save opps. In addition, some closers on terrible teams rake in huge save totals (ie Mike Williams and his 48 saves with the Pirates last year.). So, the best strategy in general is just to go with the best overall closers. Top notch closers always seem to get save totals exceeding 35 if they are healthy all year.