regression to the mean
The statistical tendency for extreme initial measurements to move closer to the average upon repeated or further measurement.
As defined in William MacAskill on Effective Altruism and Doing Good Better
Other definitions
A fundamental principle that overvalued or undervalued conditions (particularly in financial markets) inevitably correct back to long-term historical averages over time
Markets Overvalued by 150% as Dishonest Metrics Hide the Coming 'Catastrophic' Collapse: Dave Collum
Where it comes up · 8
- A Historic Market Bubble2 claims
- Daniel Kahneman | The Knowledge Project #681 claim
- #295 I had dinner with Charlie Munger1 claim
- Philip Ball1 claim
- "China is digging out of a crisis. And America’s luck is wearing thin." — Ken Rogoff1 claim
- William MacAskill on Effective Altruism and Doing Good Better1 claim
- 150 - The Map Of Misunderstanding1 claim
- 367 | Jared Diamond on the Course of History and the Role of leaders1 claim
Most important claims · 9 of 9Search all claims for “regression to the mean” →
- Kahneman's favorite bias is non-regressive (extreme) prediction: drawing extreme conclusions from very weak evidence, as in job interviews where interviewers form confident predictions of performance that persist even after being told such predictions are worthless.150 - The Map Of Misunderstanding
- Coaches are often fired after a losing season, but much of that losing reflects random factors, so by regression to the mean the next season tends to improve regardless; the right test compares next-year performance with and without firing, which shows the coach's real effect is smaller than naive comparisons suggest.367 | Jared Diamond on the Course of History and the Role of leaders
- The best prediction is always less extreme than your intuitive impression because intuition does not account for regression to the mean—a statistical phenomenon where extreme cases tend to move back toward the average on re-measurement—and this is a subtle error that is very hard to correct even when recognized.Daniel Kahneman | The Knowledge Project #68
- The expansion and increasing institutionalization of science may make it a victim of its own success, giving rise to a more conservative, cautious approach—a regression to the mean toward mediocre and safe work—as high-risk frontier and blue-skies research becomes harder to fund and publish amid demand for short-term commercial return, which could partly explain a slowdown in transformational discoveries.Philip Ball
- When initial results suggest something is exceptionally good, you should statistically expect the effect to regress toward the mean as further evidence accumulates; the downward revision of deworming's impact in reanalysis is exactly what one should expect and does not undermine the case.William MacAskill on Effective Altruism and Doing Good Better
- Regression to a long-term normal is what rules everything in markets, so overpriced trends, including 75-year trends in the S&P, eventually revert, although after the 1929 and 1965 bubbles, the Fed's support in 2000 let the S&P bounce back above trend before breaking it years later.A Historic Market Bubble
- Rogoff expects a rebalancing in which foreign equities outperform dollar equities going forward—not because US equity growth slows but because Europe in particular has substantial catch-up potential and the dollar, currently very strong, should mean-revert lower (the euro doing well); European remilitarization would further boost the euro, European tech, and Europe's geopolitical power."China is digging out of a crisis. And America’s luck is wearing thin." — Ken Rogoff
- The NASDAQ fell 82% and the S&P 500 fell 50% after 2000, and Grantham's firm had forecast about minus 75 percent for NASDAQ and minus 50 percent for S&P, with the S&P not overshooting the trend as it had historically.A Historic Market Bubble
- Munger's view is that mimicking the herd invites regression to the mean, so investors who trust their own judgment and ignore others' opinions have an advantage.#295 I had dinner with Charlie Munger