The talk’s core advice about planning ahead for stress is well supported, but several scientific and medical examples are simplified or context-dependent.
Claim-by-claim breakdown
Accurate (97% confidence): The pre-mortem technique was developed by Gary Klein and involves imagining that a plan has failed in advance to identify what could go wrong. Gary Klein’s premortem method was published in Harvard Business Review in 2007, and AHRQ describes it as a way to imagine a project has failed and then identify likely causes before they happen.
Accurate (92% confidence): Stress triggers cortisol release, and sustained stress can have harmful effects on cognition and health. NIH and review sources describe cortisol as part of the normal stress response, while noting that chronic exposure or sustained elevation can become maladaptive and contribute to health problems.
Accurate (94% confidence): London taxi drivers have changes in hippocampal structure associated with learning city navigation. UCL and the original MRI literature report hippocampal gray matter differences in London taxi drivers linked to navigational experience, including increases in posterior hippocampal measures.
Accurate (79% confidence): Squirrels use spatial memory, not just smell, to recover buried nuts. Behavioral studies of squirrels show spatial-memory-based retrieval of cached food, while other sources note that smell can still contribute, especially for locating other squirrels’ caches or under certain conditions.
Accurate (72% confidence): The number needed to treat for statins is around 300 in some populations. A USPSTF evidence review found NNTs for statins that varied by outcome, including an NNT of 286 for all-cause mortality over 1–6 years; other studies report different values depending on risk and follow-up.
Misleading (84% confidence): The talk’s statin side-effect framing makes the drug seem more harmful than helpful by combining different statistics too directly. Modern evidence summaries show statin benefits and adverse effects are outcome-specific and vary by baseline risk; the talk’s '15 times more likely harmed than helped' framing is not a generally valid comparison.
Accurate (67% confidence): Radical prostatectomy for prostate cancer has a number needed to treat much higher than 1, meaning many surgeries are done for each patient who benefits. The ProtecT trial summary reports much larger NNTs for preventing prostate cancer death over the study’s follow-up, while the talk’s quoted number appears to refer to a different older estimate or endpoint.
Accurate (93% confidence): Making important decisions while stressed can impair judgment and increase the chance of mistakes. Review literature and stress-response sources describe stress-related cognitive impairment and maladaptive decision-making under sustained or acute stress.