I wouldn't be surprised to hear there's a 10x difference in, say, error rates between the best and worst doctors. In fact, I'm almost definitely sure that's a lower bound, given my experiences with doctors in third world countries. Barbaric, some of them still are. Depending how bad the worse doctors are, 10x from "average" might be reasonable.
In software development, I think that 10x number is also a lower bound if anything. Bad developers are everywhere, churning out volumes of terrible, pointless, code. In a code review a while back for a client, I made an offhanded comment about a minor thing like "make sure this list is sorted and deduped". The code owner then started "discussing" that and estimated it'd take about 2 hours. I was stunned and added the code ".Distinct().OrderBy(...)" right there in the review. I'm at a loss at how to explain that kind of difference, and that's on a micro-scale.
At the other end, I've seen projects/companies go down colossally wrong paths because they simply couldn't understand basic algorithmic complexity. So, say, I dunno, a week or two to do it right, versus a year spent chasing down fake paths? That doesn't even measure the business impact of floundering around for a year while impacting customers.
Focusing on "coding 10x faster" is simply misleading. Like measuring an author based on how fast they can type.
Doctors are also somewhat "presorted" by ability. Certain specialities just aren't attainable by any but the most "gifted", whatever this means in that context, for example tenacity, ability to remember vast amounts of facts and, of course, dexterity.
Objective Performance in doctors is harder to judge than in developers. Some specialities lose a high percentage of patients no matter how good the doctors are. Sometimes even a difference of "10x" in error rates may require more cases to judge than a particular doctor actually has in a year.
10x difference between error rates? No, this is not the case. Well, it might be in some special situations, like a scammer and a specialist, but you'll wont find that in the case of GPs, for example. Obviously you should only compare doctors on the same field (N0tch would be quite unproductive if he suddenly had to work with COBOL, for example), and between those the difference is measured in percentages, not 1000 percentages. (I can dig up the data if you want but I definitely read this when I recently had to look into diagnostic errors.)
I would be interested in the source and how they defined the error rate. (Although I'm not sure error rates are the best way to define a doctors "productivity")
Are you taking into account doctors in shitty countries? I've personally seen people seriously harmed and killed due to what, in the US, would be viewed as criminal intent. Yet inside such a country, the cases were viewed as nothing notable.
And are you really saying that inside the US, there are no practitioners with an error of 1/1000 where another one has 10/1000? That seems awfully tight.
In software development, I think that 10x number is also a lower bound if anything. Bad developers are everywhere, churning out volumes of terrible, pointless, code. In a code review a while back for a client, I made an offhanded comment about a minor thing like "make sure this list is sorted and deduped". The code owner then started "discussing" that and estimated it'd take about 2 hours. I was stunned and added the code ".Distinct().OrderBy(...)" right there in the review. I'm at a loss at how to explain that kind of difference, and that's on a micro-scale.
At the other end, I've seen projects/companies go down colossally wrong paths because they simply couldn't understand basic algorithmic complexity. So, say, I dunno, a week or two to do it right, versus a year spent chasing down fake paths? That doesn't even measure the business impact of floundering around for a year while impacting customers.
Focusing on "coding 10x faster" is simply misleading. Like measuring an author based on how fast they can type.