Is Kelly always best?

For bankroll management decisions, does it make a difference if your goal is to win a fixed amount of money and then stop playing (except maybe leisurely on occasion)? For simplicity I'll make up numbers...lets say you start with a $25k bankroll and you want to turn it into $50k and then stop playing seriously. We'll also say you have a 1-2% advantage. Would it be worth it to bet above Kelly and/or try some type of anti-martingale or other betting system along with your counting advantage (like keep pressing up to the max in TC >2)? I know the general answer is "it depends on what ROR you want," but if you're only trying to make a certain amount I would think it makes a difference since there won't be infinite ups and downs (once you're up a certain amount there is no more down...or up). Most of the card counting info I see is based on if you play indefinitely. I don't see a lot on if you want to win X amount and stop. Is Kelly still the best?

Setting specific goals sometimes works sometimes doesn't. What works more consistently for me is to just see where the session heads and get up and leave if whatever I am ahead starts going too far in the other direction. However, I do set minimum goals, which I stick to unless I am behind for that session, then I tend to get up and leave once I am back to even or a little ahead if I have come up from behind.

This money management strategy has worked for me for countless sessions.

Quote: camz1969

For bankroll management decisions, does it make a difference if your goal is to win a fixed amount of money and then stop playing (except maybe leisurely on occasion)? For simplicity I'll make up numbers...lets say you start with a $25k bankroll and you want to turn it into $50k and then stop playing seriously. We'll also say you have a 1-2% advantage. Would it be worth it to bet above Kelly and/or try some type of anti-martingale or other betting system along with your counting advantage (like keep pressing up to the max in TC >2)? I know the general answer is "it depends on what ROR you want," but if you're only trying to make a certain amount I would think it makes a difference since there won't be infinite ups and downs (once you're up a certain amount there is no more down...or up). Most of the card counting info I see is based on if you play indefinitely. I don't see a lot on if you want to win X amount and stop. Is Kelly still the best?

Like you said, it depends on what ROR you’re comfortable with. Just because you have a certain long term goal in mind it doesn’t change how you should bet. Eventually you’re going to play your last hand — why would you play any differently if it’s your first hand, 10,000’th hand, or your last hand? Hint: You wouldn’t.

Quote: camz1969

. Is Kelly still the best?

Firstly, you have to make your goal more defined: « Win a fixed amount IN WHAT TIMEFRAME? », « Does the time change the value I put on my goal? », « What do I consider as failure to reach the goal? », « Is the residual value zero if the goal is not achieved? »
Etc.

Secondly, you should define what you call ROR. I understand it means risk of ruin, so the question is twofold: what do you define as ruin for yourself, and what do you use as definition of risk? (What time frame, calculation a priori or running, etc.)

Supposing you consider the Kelly bet, it entails (a) that you bet according to a budget / bankroll / capital that you want to see grow ; (b) that you bet a portion of that capital. Consequently, if you define Ruin as the loss of your whole capital, it never happens with portions betting. So I guess your ROR means something else. What?

Thirdly, Kelly criterion seeks the best ‘rate of growth’ of the capital. This optimal value is a function of the bet conditions: amounts, probabilities. So, if you are limiting to one specific game (there are not a bunch of positive EV ones), it has a built-in rate of growth. As a result, reaching x2 your bankroll is essentially a question of how many hands are needed on average. In that case, choose Kelly if your goal is to reach it fastest.

But if you want to trade some of this average speed for some thrill and a possibility of winning quicker (or if the built-in duration is too long for your purpose), you might try some positive progressive martingale, limiting the total bets of a ‘martingale round’ to some Kelly-like portion of your

Anyway : Kelly betting IS a progressive system when winning, and degressive when losing !

You're right...I did not clarify some of those points. Basically I was asking is it worth the increase in risk to try to win X amount in a shorter time period using another betting progression if you have a stopping point versus playing indefinitely. I would think the math is different with a stopping point and the risk doesn't increase as much. Trying to get a 100% return is different than playing forever in which you are more likely to encounter really long down swings. My definition of ruin would simply be losing the whole initial bankroll. Even a gambler with a negative edge has a better chance if they have a stopping point, but of course I'm talking about counting with a 1%+ edge. There is a cover aspect to this as well since you would look more like a regular gambler if you use something like anti-martingale.