in the original experiment (sackman, erickson, and grant) they were able to successfully solve programming-contest-style problems in a tenth of the time that other people required. so maybe ten hours instead of 100 hours. this should not surprise anyone who's worked on a programming team or who has looked at code they wrote ten or twenty years previously. the actual ratios were from 5:1 to 28:1, but that was an understatement, because some of the programmers couldn't solve the problem at all in the allotted time, if I understand correctly in some cases because they decided to solve it in assembly language. sackman, erickson, and grant suggested firing the 1× performers:
> Validated techniques to detect and weed out these poor performers could result in vast savings in time, effort, and cost.
complaints and issue lists were not evaluated, but rewriting twice a month would have resulted in not completing the tasks they were being evaluated on
> Validated techniques to detect and weed out these poor performers could result in vast savings in time, effort, and cost.
https://www.construx.com/blog/the-origins-of-10x-how-valid-i...
https://dl.acm.org/doi/10.1145/362851.362858
complaints and issue lists were not evaluated, but rewriting twice a month would have resulted in not completing the tasks they were being evaluated on