Quote:
Originally Posted by Mihkel05
Some of that is just outright false.
Okay, could you point out what is false and why? Or if you think there's too much, then just a few things?
I'm not claiming to be a better bettor or have more sports betting knowledge, I believe that you do. I'm asking these things honestly because my understanding is different, but I don't mind being pointed out why my understanding is wrong if so.
Quote:
Originally Posted by bgordon
Not gonna read the paper since I don't do this professionally anymore
The second article doesn't really make an argument, it's just assertion with a link to the first paper
The third article just makes up a metric and doesn't attempt the task of explaining why it's more useful than expected growth. Uncertainty about the true odds is a real reason to bet less than Kelly, but it's something you can adjust for.
The fourth article argues that hedge fund managers have goals other than making the most money for their clients. Whatever
The article on pinnacle is just some guy saying he doesn't like math
The reason I posted those things is not to claim that each one is 100% correct and thus it is wrong to bet Kelly, or anything else that you might infer, it was simply a response to the quotes I included at the start of that post. The point was that doctor was correct and lvr was wrong, there are indeed rational reasons not to go full Kelly criterion on a single bet, that are not emotional ones. And to show that in a helpful way, instead of just replying "lol no u" to lvr, I included some samples of possible rational reasons. Some may be incorrect, but it seems that there are rational reasons in there.
I did not even need to do that, since technically doctor is correct and lvr is wrong, just by virtue of people's acceptance of half Kelly or any fraction of Kelly. If lvr was right that there are no good reasons to not go full Kelly, then it would be wrong to ever go half Kelly.
Besides that though, some of the links do seem to provide useful commentary. There has been positive commentary on the first paper from others. The summary said:
Thus, bet sizes that maximize this measure or its marginal increase are also reasonable choices. Both theoretical analysis and computer simulation show that these alternative choices of bet sizes are much more conservative compared to what the Kelly-Thorp theory suggests and makes sense in practice.
You can choose not to read the paper, but this is not an argument against the above. And for our purposes, you don't have to agree that their methodology is better for bet-sizing than Kelly, but simply that there is rational basis for believing there are non-emotional reasons not to bet full Kelly.
Yes, the second article referred to this same paper. I included it because it is quicker to read than the entire paper and summarizes some relevant points, namely that "risk aversion and a finite investment horizon" are "practical considerations" not emotional ones, that affect proper bet-sizing.
The third article just shows why half Kelly or other Kelly fractions can be preferable due to cutting your volatility by a greater amount than your expected return is cut. Some of the reasons to prefer doing so could be risk management or as you said, uncertainty about true odds. In either case you refer to a "real reason to bet less than Kelly" which is what was being addressed initially.
The fourth article is applicable to sports bettors as well. The point was:
"Notice this goal is not the same as many hedge managers’ or their investors’ goal. They often want to maximize their Sharpe ratio, not growth rate, for the reason that their investors want to be able to redeem their shares at any time and be reasonably sure that they will redeem at a profit. Kelly criterion is not for such investors. If you adopt the Kelly criterion, there may be long periods of drawdown, highly volatile returns, low Sharpe ratio, and so forth."
This is applicable to any sports bettor that draws down from his bankroll. Perhaps someone regularly draws down a significant enough portion of his bankroll to live on and needs to account for that, or perhaps someone will be buying a house in the near future, or whatever. Kelly optimizes long-term growth-maximization, but may not be a superior criterion when accounting for volatility and near-term drawdowns. None of these are emotional reasons.
The fifth article on Pinnacle is not necessary but simply makes some of the above points about variance and acceptability of loss in cruder fashion.
Again, you do not have to buy every one of the above arguments, but it does seem that there are rational reasons not to bet full Kelly, which was the original point.
Quote:
Originally Posted by fannypack
so dont bet full kelly because it creates uncomfortable situations?
As shown above, there are more reasons than simply "uncomfortable situations."