You can’t get escape the concept of utility. You’re implying that you assign higher utility to returns that have one shape vs. another.
Every strategy has a probability distribution over outcomes and people differ on which probability distributions they like.
Conversation
You can say “as long as you like more money vs. less then it implies you should prefer Kelly” (for *almost* all percentiles), but this goes back to the insane infinite time assumption.
In real life the 10th percentile outcomes of Kelly tend to be pretty bad.
1
They look a lot better than the 99th percentile outcomes in almost-all-in (the strategy Sam prefers)!
And law of averages kicks in over time; it doesn’t for almost-all-in.
1
1) I think that and I were talking about what _we_ think, and we do believe in utility, so I don't think it's appropriate for you to respond the way you did.
2) can you please address twitter.com/SBF_Alameda/st? Percentile outcomes as a metric fails.
1
I wasn’t saying you were being inconsistent with yourself
On percentiles: yes, directly optimizing median or any given percentile would probably not be coherent. But Kelly doesn’t try to do that. It just ends up doing that for all percentiles (other than 100%) eventually
1
Dan: "yes, directly optimizing median or any given percentile would probably not be coherent"
Also Dan: twitter.com/danrobinson/st
Could you please, in a single tweet thread here, define what exactly it is you _are_ aiming for?
Quote Tweet
Replying to @SBF_FTX @SBF_Alameda and @elliot_olds
No! I am not trying to maximize EV of anything!
I want to pick the strategy that beats yours 99.99% of the time. That’s my terminal goal
Kelly takes that input and spits out that I should maximize EV(log(wealth)), but that preference is the consequence, not the cause
1
1
Where’s the inconsistency? There’s nothing incoherent if the desideratum is “I prefer strategy A to B if A outperforms B at all percentiles other than 100%,” right? That avoids the independence of possible outcomes problem, right?
1
Probably should have been clearer—Kelly doesn’t directly optimize median outcome in a given gamble (that would be incoherent), which is what I meant by “doesn’t try.” But the fact that it does tend to do that over time is one of the properties that makes it appealing, for sure
1
wait but 99.99% is *not* the same as "all percentiles other than 100%"!
Enough mottes and baileys. What is the thing you actually believe? How will you decide how to act in non-infinite scenarios where nothing is 100%?
1
This Tweet was deleted by the Tweet author. Learn more
There are lot of things I can abide by.
But if there's one thing in the world you don't fuck around with, it's me + misusing the Kelly Criteria.
Everyone has to have their limits _somewhere_.
Have you ever seen the Kelly criterion being used correctly? The way it is stated all but guarantees misuse.
1
4
Heh basically never, I think it's almost always a mistake to use.
If you want to max log wealth, say that! If you say Kelly it's very likely because you're trying to slip past people the difference between "almost surely" and "surely"