Swings/Bankroll

[Oneoffthecount]

My first post, describing me. Have played a couple of hundred thousand hands on CVBJ and recently ordered CVCX. Both are great. I'm using Hi-Lo, and play eight deck in Atlantic City. I've lost a few thousand and wanted to step back to re-evaluate and learn before going back. I can tell you that I think CVBJ and CVCX are absolutely fantastic as well as support offered by QFIT. If you can afford to sit down at a table and loose a grand, you can't not be able to afford these products.

My basic question is on swings. I just played a session of 831 hands and went down 255 (3825 on a $15 table) units a loss rate of -12.69%. My cumulative error cost for the session was about $36. According to CVCX my chance of this result is less than 1%. Howver, I see swings at 1000 hands betting this ramp on a miniscule amount of data (46K hands) betting this ramp of 6 to 7 K. I dumped my CVBJ data into a spread sheet and measured change in bank roll by subtracting each bankroll balance from the bankroll balance 1000 hands before it. So, I throw it out there. How many units down would you consider an extremely unusual result, e.g. less than 2% probability in units. I'm playing about a 13 unit spread.

I am very slightly ahead overall over the last 46k hands. It seems to me that the loss itself isn't important but being able to plan for sings and absort them is, as well as, being able to absorb them psychologically. A $4k loss or a $6K loss woudn't bother me if I planned for it and expected it eventually. Looks like a $10K bankroll on a $15K table at a minimum. Even playing the ramp I'm on, CVCX thinks there is a 9.2% chance of wiping out my bankroll. Once again, if you have a thought on min bankroll relative to max bet, spread etc. Let me know.

[KJ]

First of all, congrats on your approach to the game. Having any kind of success is more about preparation than what actually happens at the table, and that preparation includes being prepared financially (having necessary BR) and being prepared mentally for the negative swings that will occur. 9.2% RoR is high. Way to high for my liking, but I am now ultra conservative in that aspect as my BR is non-replenishible. I understand that folks that are able to generate a new BR should they go bust can play to a higher RoR than I will accept, but it is important to really understand this concept. Are you really able to lose your entire bankroll and how quickly can you replenish it? 9.2% means there is a real possibility this could happen. And already playing the lower limit tables, you are limited in your ability to cut back your unit and spread should you lose a significant portion. Players, playing underfunded often cheat by playing to a higher RoR. God knows, I did early on and got away with it. You do so because playing to a more reasonable RoR would mean smaller profits and slower growth. Sometimes I think players don't really weigh that against what would happen if they go bust. Do they have to sit on the sidelines for a year while building a new BR? These comments aren't directed at you specifically Oneoffthecount, as I don't know your specific situation. They are just my general thoughts on playing to a higher RoR. So if your going to start out playing to a 9.2 RoR, your top priority has to be to get that lowered as quickly and safely (if that's even possible) as you can.

Tthree is right. Finding and playing the best games is one of the biggest favors you can do your RoR and win rate. Unfortunately, you mentioned that you are playing AC, so your options are limited in that regard, especially at red chip level. But do the best you can, seek out the best games and especially the best penetration you can find. Penetration was the one big variable when I played there the first 5 years of my career. You could find different penetration if you looked and occasionally find something decent. (decent being subjective) Playing AC is tough. No doubt about it. But, if you can find the ability to win there, you can win almost anywhere.

One thing you didn't mention is wonging? When playing 8 deck games, even with a decent spread, playing all will kill ya. You don't have to wong into games as that used to be difficult in AC when I played due to crowded conditions, but you need to at least get out of some of those negative counts.

[zerg]

How many units down would you consider an extremely unusual result, e.g. less than 2% probability in units.

You can use your software to determine the exact number that is 2% probability. If you are looking for real world results, I'll share an answer this in max bets rather than units. I have played for about 15 months, probably a little over 1000 hours. I have had two bad losing streaks. About 60 & 80 max bets. I play a little less than .5Kelly on average and think those bad runs were in or near the 2% range.

if you have a thought on min bankroll relative to max bet, spread etc. Let me know.

Everyone has a different answer for this. I recommend for full Kelly betting for those with small replenishable bankrolls, and starting to back off when you can afford to play 25 tables. It is a personal preference. You need to look at your situation through your eyes to come up with the correct answer for you. It sounds like the ramp you are considering may be right for you.

Id like to mention a reason to consider betting less that I have been thinking about recently. It seems to me that one could probably play longer hours when betting less for a few good reasons. During my big losing streaks I lost some playing time just because I had lost so much money. I was betting at a level that those losses took a mental toll. It made it hard to play as much. Also, smaller bets in general get less attention which means you can probably play at any given place a little longer.

[Oneoffthecount]

Wow, thanks all of your advice makes perfect sense. I've been amazed by the willingness of experienced players to share their knowlege. Norm has answered many questions over time and email. There is a game in PA I like and may roost there. AC had offered me a free room....somehow it doesn't look as attractive now. I'm wasn't sure how long before I wore out my welcome in PA. They havn't recognized me as a counter, possibly because I'm doing it so poorly I have them confused!!!

[jaygruden]

Spread your play around the PA stores in the Eastern part of the state. If you're in NJ, you can get to all of them on a day trip.........Parx, Sugarhouse, Harrah's, Sands, Mohegan Sun, Mt. Airy & Hollywood. At least four of these have pretty good conditions. Add in Dover Downs in DE & sprinkle in an occasional trip to AC (2-3 stores have playable games if you scout) to further spread yourself and you'll be unlikely to burn any of them out.

[Gamblor]

Dare I say, BJ is a "streaky" game! Even at this point, almost every other day I sit there mystified by the wild swings and roller coaster nature of BJ. Its almost like wild swings are common.

Having also played poker a fair amount, I have NEVER had the same feeling at poker, its an outright grind compared to BJ. Still can't explain at this point, could be comparing apples and oranges.

[Lowrider]

I lost 85% of my bankroll when I first started which was about 2.5 standard deviations...the math gurus can chime in but I think the odds of hitting a serious bout of 3 standard deviations is about 1 in 100 before reaching N0. I redid the math every single night to make sure I wasn't crazy and almost quit....but then the slow, steady climb started and I caught up to even eventually and finally worked my way to my expected return and then past it to a nice profit level.

Unfortunately, life got in the way and my car died so I had to buy a late model used car....I like to "buy and hold" my cars looking to get at least ten years out of them....so my bankroll was devastated and I'm back down to below my original buyin, but I funded 80% of my car purchase with AP winnings and I don't have a car payment....very frustrating to be back to square one but every time I drive my new car I smile at the fact I basically funded it with my hard work as an AP.....

[muffdiver]

That's fiscal responsibility---------------------good job!

muffdiver

[Oneoffthecount]

Welp, I'm up to a 477 unit loss in 1531 hands. According to the software that is only 2 tenths of one percent probability. However, I have seen over the course of the last 56k hands losses in the 350 to 450 range before, so I'm confused about the probablity of a given loss of a course of a number of hands. If a loss has a .002 probablity, shouldn't that mean that it would only happen on average 2 times out of a thousand sessions played if the sessoins were 1531 hands each? This is probably my largest loss in a 2k hand period, however, I see others around the 400 range moving from top of bankroll to lowest point before going up again. I'm not trying to whine as much as I am trying to figure out what type of swing is normal and, as such, what size bankroll one would need to accomodate it. I spent about 52k, hands (although admittedly not playing perfectly) playing down to a 400 unit loss and then up and down, until I was finally about 666 units up, now I'm back down to about 166 units. My EV per hour was 23.80 and my standard deviation per hour (defined as 100 hands) was 33 units per hour. My actual result was about -32 units per hour, making my actual result -about 33.5 units away from the mean so basically one standard deviation negative. So if I'm one standard deviation or so away from the mean, then I have about a 15% chance of that result, which might be about right for what I'm seeing. So why does my software think there is only a 2 tenths of a percent chance of that result? Last, if two standard deviations happen about three percent of the time, then does that mean that in about 3 percent of 10 hour sessions I could see a swing of 800 units or more? I wish I had stayed awake in statistics now, I really really do...