So suppose you are trying to compare 2 people's consistency in Bowling where the max score is 300. Standard deviation seems like it would not be reliable to measure consistency in performance because large variations are seen without context.
If player A gets 104, 115, and 180 while player B gets 120, 123, and 127, player B is seen as the more consistently better one if you plainly use standard deviation. If you use the mean of both players' data, player A's average will be affected by the outlier. So I'm wondering which formula can be reliably used to determine who is more consistent as well as better performing overall.
Consistency and level of performance are quite different attributes and you are unlikely to find a single measure that statisfactorily summarizes both in just the way you have in mind.
You need to sharpen your criterion if there is to be only one. If two criteria then one measure of centrality and one measure of dispersion might be what you want. 'Robustness' is useless without clear purpose.
Are you looking for most valuable member of a bowling team or best individual player? Are there circumstances in which the one worst score is disregarded? Is (180, 181, 150) "better" than (170, 172, 169)?
For measures of centrality, consider mean, median, and trimmed mean. For measures of variability, consider standard deviation, inter-quartile range, mean absolute deviation, and perhaps coefficient of variation. (You can google the ones not in a readily available statistics text.)