Does this phrase even make sense? To say that

[a particular approach] "achieved, on average, mixed results at best".

I'm concerned about the combination of 'on average', 'mixed results', and 'at best': does it make sense to use two (never mind three) of these together like this?

(The phrase comes from a DevOps article by Damon Edwards.)

  • Yes, it means the best that can be said of [the approach] is that the typical outcome was mixed (i.e. sorta good, sorta bad). There is nothing better than can be said about [the approach]. That's the best we can say. In re: "on average", that means that there may have been some outcomes which were stellar, and some which were abysmal, and some which were non-descript, etc, but variation is normal, and those specific outcomes aren't really characteristic of the process when viewed holistically. So yes: the approach achieved, on average, mixed results at best.
    – Dan Bron
    Mar 19, 2016 at 12:48
  • @DanBron - that's why I said an almost tautology. It depends a bit on context. How many results are we talking about? However, I think your comment is more in line with how I first read the sentence. Mar 19, 2016 at 12:52
  • 3
    I'm afraid I can't agree with the comments. It achieved mixed results at best, meaning that the best thing that can be said was that it achieved mixed results. Agreed. It achieved, on average, mixed results. Meaning that the average results were mixed. This is somewhat problematic already, since the average is taken from all results. If the results were mixed, then an average would fall somewhere in the middle of the worst and best results. ...on average...at best. Seems to be a contradiction in terms, are we talking about its best results, or the average of its results?
    – Karl
    Mar 19, 2016 at 13:26
  • @Karl Agreed. The writer wants his readers to feel that a certain style of industrial approach has been definitively discredited, largely because his article is essentially an advert for a company that purports to do things differently. To promote this favourable view, he deploys (among other things) this phrase that actually says nothing usefully intelligible: mixed results are no surprise at all; on average implies a scale of measurement that he does not reveal; and at best implies generally unsuccessful (not mixed!) results in some unexplained sense. It's meaningless handwaving. Mar 19, 2016 at 13:38
  • @Karl Agree completely. Thanks for not rewarding a question that needs to be explained much more clearly with an answer (though this would be a great one). Mar 19, 2016 at 14:45

1 Answer 1


It is simply an example of poor writing, including a typo (‘on averaged’) that you understandably do not reproduce.

The writer’s intent is obvious enough, of course: individual organisations have different requirements, so (he claims) the attempt to refine a uniform approach across contexts would be misguided. He strongly implies that such an attempt has been proven to fail, although he does so by using the overwrought phrase that you mention rather than giving actual evidence. He strives to convey the impression of an informed historical overview and of inconclusive-or-destructive results, but does not actually provide either.

Cursory inspection of this article reveals many other infelicities created by the author and/or his editor. If you plan to cite the article in support of some kind of case, I would suggest carefully paraphrasing the writer’s general view rather than direct quotation. On the other hand, direct quotation would be a good idea if you plan to give an example of an opinionated argument that advertises its own lack of evidence by trying to disguise it.


Obfuscatory—that’s the word I was groping for. The writer hopes that you will leave with a sense of of his preferred case having been made, behind the smokescreen of the kind of construction you have quoted, when he has actually established nothing at all.

  • 1
    I can't wait to find out why someone reckons that spending the time consulting the actual source in order to answer the actual question is in some sense not useful!! Some people really are funny. Anyway, when downvoting, it is always useful and polite to spend a moment helping the offender to understand the offence, and this question is specifically about interpretation. So come on: let's play nice. Mar 19, 2016 at 13:54
  • 1
    I'm glad to see that you're rapidly becoming more understanding. // The downvote was probably for not addressing the exact question/s on the actual phraseology OP should have asked to fit in with ELU requirements; Karl does this in his comment above. Mar 19, 2016 at 14:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.