yep, what you refer to there is quite the 'law of averages" but something more like a reference to the law of large numbers.
anyways, i hevent been playing fantasyball for two long but ive found out that trying to predict relievers wins, especially over a relatively short period of time is not that resourceful. even harder than predicting sterters wins which is a somewhat smaller crapshoot. notice its called the law of LARGE numbers, and obviously the sample sizes with relievers arent even remotely close to being "large". with relievers theres even more random crap going on than SP.
much easier* to try to predict their actual performance.
now that Gordon's in the mix, id say its a throwup between him and Mota. Gordon's been unhittable and Mota hasnt shown control comparable to last year which slightly worries me but he should be fine. hes thrown a ton of innings in the past 1.5 years. id give Gordon a slight edge for this year from now on.
*"easier" here is very relative.