When should you cut your bet size after a series of losses?
I assume the optimal way to do it is to keep the ROR at a constant rate, like 1%.
Realistically, how often should you adjust your predetermined bet size? After every win or loss?
If I max bet $200 per hand and lost $2000 during a shoe, which decreased my bankroll by 10%. For the next shoe, should I decrease my max bet then?
Optimally, you adjust after each bet but that is not practical.
I like setting an RoR (usually .5) that I am comfortable with and the reevaluate after a days play that experiences a big move either way (usually 15-20% of bank).
Luck is nothing more than probability taken personally!
I used to not adjust until I didn't have enough money to double down my max bet 16 times, especially when I was losing. At a time, I was down to $4500 in total BR, I still bet $200 max bet cause I thought if I have money to double down or splitting, I would have the edge and everything's gonna be fine. I couldn't reduce my bet so profit from card counting is still worth the time, and if I lost more, I could earn money from jobs to increase my bankroll again. I believe this mindset is only correct when my bankroll is too small to worry about lose it all.Originally Posted by Stealth
Adjust daily sounds like a good idea, more practical than adjust per shoe. When you adjust, you simply keep you RoR at 0.5% again regardless of winning or losing right?
15%-20% bank swing with 0.5% RoR sounds huge, how often does this happen?
Sent from my iPhone using Tapatalk
Optimal is to use a 13.5% roR set before each bet. That is adjusting your bets after each win or loss to maintain a constant 13.5% RoR. That is not practical. Most will try to keep their RoR in a range that is comfortable for them. Often that is a fraction of the latter kelly style betting.
This depends on your RoR. The higher it is the more you need to adjust to maintain that RoR.
Your BR is not huge. You need to tend to it more than many. You never say what your RoR was and what it will be if you don't resize. That is important. It sounds significant enough to consider resizing. What you want is a series of BR amounts and the to resize up or down to maintain a comfortable RoR. Once your BR is large enough these concerns will not be a factor as often. Losses will not have much effect on RoR and increasing your bets may not be an option due to casino tolerances or table limits. But, the example you give shows a sufficient BR that needs more maintenance when it comes to resizing.
It isn't. Three didn't express it properly. 13.5% ROR is for an original Kelly bettor who NEVER again adjusts his bet size. He either wins all the money in the world or goes broke trying. By definition, if you keep resizing your bet after every move of your bankroll, which is the very definition of pure Kelly betting, you will NEVER go broke. But, for practical purposes, no one plays that way.
There isn't, however, as Three incorrectly explained, any such notion of having a series of 13.5% RORs.
Don
Don
Well, I will defer to Don on dotting the "i"s and crossing the "t"s when it comes to terminology. But it is not the way the term is used in conjunction with the software in BJ. I think a good definition would be able to be used to fill the RoR field when generating a bet ramp. If you say RoR assumes you never resize and then redefine RoR for players that intend on resizing then which RoR should a player use as an input in ramp generation. Right or wrong I refuse to have such inconsistencies in the way I use a term. To me RoR always assumes you never resize and I always know that I will resize when necessary so RoR as it was originally defined and as a consistent definition gives no direct information on my odds of busting out, but is a very useful metric. If Don says that is wrong I am sure he is right. That doesn't mean I am going to make my posts an order of magnitude longer to keep true to a shifting multiple definitions or use explain something for each of the many different definitions of RoR while making a post. I just use RoR as a stat that makes an assessment of the chances that you will bust out for your BR and approach to the game based on the errant assumption that you will never resize your bets. I believe that is probably how most do it but I could be very wrong about that. If I am bear in mind how I use the term when I make comments unless you want posts explaining how all the ways defining RoR could affect your attack to the game every time I use the term. I am not shy about writing long posts so if that is what Don wants I can oblige him.
Wouldn't it be better to establish the definition of Risk of Ruin as the Risk Function itself? Rather than Don and you post endlessly about the semantic/wordy definition, use the universally accepted definition (found in BJA) based on the mathematical construct of risk with respect to expectation and volatility.
That's what I I think I do. I use RoR as a stat that, based on an assumption you will never resize, gives the stat for the odds you will bust out. Since you are most likely going to resize it is not actually the odds you will bust out because the assumption the figure is based on is false. The problem comes when you want the stat to be your likelihood of busting out. Then the inputs into the sims aren't your chances of busting out. And you have to invent alternate definitions of RoR, like Don did in his book, that depend on your resizing intentions that you may or may not actually follow in the future. It would be less confusing if the odds of busting out if you plan on resizing were defined as a separate stat in Don's book rather than giving RoR multiple meanings.
Again it is all semantics. Don decided on multiple meanings for RoR so it is hard to communicate ideas while keeping to the multiple definitions of RoR. I just stick to the definition where it (errantly) assumes you will never resize and doesn't actually indicate the chances you will bust out except for the rare cases that a player would bust out without resizing. Certainly if your BR is small enough you can't resize after a BR loss. If you have a large enough BR you will most likely never resize but then your RoR is essentially 0.
If the BR is too small, let's say $5000, it would be a waste of time to resize a smaller bet ramp, I'd rather make money faster by washing dishes than resize a small bet ramp.
If the BR is too big, casino tolerance doesn't allow us to bet more to generate higher EV and increase the RoR from roughly 0% to 13.5%.
I think resizing only make sense when the BR is between $10,000 to $100,000 for solo play. For a large sum of BR in use, RoR is only "adjustable" for BP team play.
I have another RoR question.
How does RoR calculation consider the last few actions before ruin?
Unlike tossing a coin that has two simple results, Blackjack has double down, split & resplit options.
With a set RoR, If I never resize my bet ramp, keep losing until I have $20 left, which means I have a much worse EV if I keep playing with no money to double or split, does the ROR calculation assume I keep betting regardless of having enough money to double or split?
In reality, I would stop betting if I don't have money to double or split, so I would never ruin my entire BR, at worst with a tiny amount left.
It's more or less academic at that point, no? If you start with, say, $50,000, and you're down to your last $20, you're effectively ruined whether you can double or not. You're not going to build your $20 into another bankroll, are you? So, you can bet $10, so that you can double, or you can "splurge" and bet the $20; it makes no difference. You're done.
Don
Posting Permissions
Reply With Quote
Bookmarks