You've slightly conflated the assumption and the conclusion there. NASA *did* assume that they were going to lose people to Apollo, due to the levels of unknown they were facing and the sheer size of the challenge, and as a result leaned very heavily into risk management processes to minimise this outcome.
The Shuttle programme, as a result, inadvertently took a different approach and assumed that the Apollo success would carry through, and so minimise the risk management, leading to 2 failures (and dozens more close-calls).
https://ntrs.nasa.gov/api/citations/20190002249/downloads/20190002249.pdf
The "normalisation of deviance" is another interesting thing to look at, where NASA effectively learned to ignore issues outside of tolerance *because the mission didn't fail* and allowed them to accumulate. There were multiple o-ring issues and tile degradation events before the losses.