Here what I found elsewhere: "Betting half the Kelly amount, for example, reduces bankroll volatility by 50%, but growth by only 25%."
The big question is, is volatility the same as risk? If yes, then you are right, reducing max bet reduces risk more than the loss in growth. But if volatility does not equal risk, then you just lost growth in return for nothing.
I would say, volatility does not equal risk. If I have a statistical certain win, then volatility is not risk, since loss is always temporary with statistical certainty.
Wrong. By your logic, the risk is always zero, regardless of the max bet, therefore there is no reduction in risk when the max bet is reduced.
"Risk" and "volatility" usually refer to the standard deviation of the return, and I think in your quote "bankroll volatility" is the standard deviation of the bankroll, which would be the same. In this thread we also use the term "variance", which is technically the square of the standard deviation.
Not to get all pedantic on everyone, but we use "risk", "variance" and "volatility" imprecisely, to capture the idea that the investor's balance can be up one day and down the next. And that's OK. Now, for something like "reduces bankroll volatility by 50%", then we need to get specific and ask whether we're talking about a standard deviation or the square of a standard deviation, because that 50% number can't be correct for both of those at the same time.
Even if you persist in your colloquial use of "risk", we can talk about a volatility/growth ratio, and see that an investor might reasonably prefer a situation with half the volatility and three-fourths the growth.