> they are very likely to grow more slowly than average (i.e., revert to the mean)
I mostly agree with you, but this is basically the gambler's fallacy.
If I'm flipping a coin every second for days, and I hit a run of 10 heads in a row, "reversion to the mean" just means that the next 10 flips are likely to be less extreme than the previous 10. It does not mean that I should expect "more tails than usual" for the next flips. We revert towards the mean, not past it.
I mostly agree with you, but this is basically the gambler's fallacy.
If I'm flipping a coin every second for days, and I hit a run of 10 heads in a row, "reversion to the mean" just means that the next 10 flips are likely to be less extreme than the previous 10. It does not mean that I should expect "more tails than usual" for the next flips. We revert towards the mean, not past it.