CVCX Question
- Thread starter Kasi
- Start date
Actually, I'm more hung-up on "physical round" stuff and could care less about hourly stuff lol.QFIT said:
While back-counting
I've always assumed "SCORE" in its purest sense meant "per 100 rounds seen".
In Don's Tables N0 always also always assumed "100 rounds seen" as near as I could tell.
For the sim in question, when you list "15930" rounds as N0, I assume, in that sim, it represents 15930/80=199 hours of play. If it represented "100 rounds seen" as SCORE does, it would represent 159.3 hours of play.
It appears to me the SCORE in that sim of 62.77 probably actually does assume "per 100 rounds seen".
N0 assumes number of rounds seen per hour based on how many rounds seen per hour.
So, my question still is how many hours of play does the "15930" represent?
Does it represent a number of rounds with a different assumption than SCORE does in this case in effect, so I can understand pics of your sims better, understanding the number of hands listed for N0 represents the number of rounds given playing x rounds per hour?
In other words, SCORE and N0 are not always based on the same assumption.
In other words, I've concluded the number under N0 represents number of rounds seen, given the number of rounds seen per hour.
To summarize my understanding at this point
the $29.27 average bet is avg initial bet per physical round actually played?
the $50.68 SD per hand is per physical round actually played?
the $0.29 under "Win/hand" means an average amount of won per round while observing 80 rounds per hour but the physical win rate is $0.55/rd.
the "%W/L" is an average % based on rounds seen rather than physical rounds played. But, in this case that would be a W/L% of $0.55/$29.27=1.884% rather than the 1.374% in the sim.
SCORE always assumes "100 rounds seen".
N0 assumes hands seen given x hands played for hour as an asumption in the sim.
Does all that make sense to you?
Thank you for hanging in there with me as I think I may have learned something, believe it or not.
With all that how long it takes to switch tables or not etc all I'm gonna know at the end of the day is my estimate of how many dealer upcards I played against and work from there lol.
So, personally, I'd prefer that "SD/hand" and "Win/hand", even N0, all operate on the same assumption. I'd like it to mean physical hand lol. Or, maybe at least change the titles to "avg win/hand" or "Hands seen or played" for N0 etc.
Quite possibly your documentation makes all this crystal clear to users and it is only me suffering because I don't have it lol.
QFIT
Well-Known Member
I'm sorry, but your questions are very long and I get lost trying to figure out what you are looking for. N0 is hands. It has the exact same meaning as the mening in Don's book since the tables in Don's book were generated with CVCX. So if N0 is 15390, then it represents 15930 hands. No matter what you set hands/hour to, N0 won't change. However many hands you play in an hour, divide that into 15930 to get the number of hours.
That's OK QFIT. Thanks for trying. I forgive you in advance from responding any further to my silly questions.QFIT said:
I agree N0 won't change, like I think you say, in terms of physical rounds played, but I'd have to think, if someone ran the exact same sim at 100/hds/hr, hands under N0 would change from the 15930 and the 52.6% would also change for starters
No?Like I get somewhere around 4620-4640 physical rounds needing to be played for this sim to achieve N0 so it seems the 15930 number would have to change if 100/hds-seen/hr is assumed.
So, maybe you probably meant, divide the number of hands under N0 by assumed hands/hr and not that the actual number of "15930" would never change?
Edited to add, thanks to BJ Avenger for making me realize, when I said "4620-4640 as # of physical rounds", it actually was $ of EV and SD. The N0 of physical rounds, my latest best guess lol, would be 8379 physical rounds played.
Like 15930 hands / 80 hds.hr=199 hrs. 199 hrs * .526*80hds/hr=8379 physical rounds.
199 hrs * $23.20/hr =$4620 of $EV.
Or, 8379 physical rounds * $0.55/physical hand = $4620 of $EV.
And, $50.68 SD/pysical round * sqrt of 8379 physiacl rounds = 91.5*$50.68=$4640.
So, $EV and $SD are the same after 8379 physical rounds. Or close enough.
That's how I interpret the sim.
sagefr0g
Well-Known Member
i'm not sure if this will clarify.Kasi said:...
I agree N0 won't change, like I think you say, in terms of physical rounds played, but I'd have to think, if someone ran the exact same sim at 100/hds/hr, hands under N0 would change from the 15930 and the 52.6% would also change for starters
...
here is the help on N0:
"N0 - N0 is another strategy scoring technique developed by Brett Harris. It is defined as "the number of rounds that must be played, with a fixed betting spread, such that the accumulated expectation equals the accumulated standard deviation As such, it is a measure of how many rounds must be played to overcome a negative fluctuation of one standard deviation with such a fixed spread." See the article The Theory of Optimal Betting Spreads on http://www.bjmath.com (Archive copy) for more information. Values are provided for both the Custom Bets and the calculated Optimal Bets."
http://www.qfit.com/downloads.htm
ok, with CVCX when you run a sim, you don't tell it how many hands per hour are played until after the sim has been ran.
once the sim has been ran and completed you can then adjust how many hands per hour are played.
Thank you Wise One.sagefr0g said:
I think, maybe, I am beginning to understand that cvcx is a "post-sim" calculator anf does not actually run any sims by itself. That, maybe, that is what cvdata does.
Maybe you, of all people, can possibly run the Lonesome Gambler sim as it originally is, if you can figure out what it was assuming and, basically duplicate it's results first, as a base-line, and then run the exact same assumptions except at 100/hds seen/hr.
My half-assed predictions:
No way will N0 still stay at 15930 hands. $SD/hd will still be $50.68. SCORE will be the same. $SD/hr will change.
$win/hd will change and $win/hr will change, obviously. BC % of 52.6% will change. Maybe. lol.
Frequencies at each TC will change. Maybe. lol.
Avg bet of $29.27, whatever it was, will not change.
W/L% probably will probably change from 1.374% but not sure lol.
My best guesses lol.
Help put me out of my misery, I beg you. :whip:
sagefr0g
Well-Known Member
i'm not sure if cvcx actually runs a sim or if it uses data from sims that were already ran by maybe cvdata.Kasi said:
i'm only guessing maybe it actually uses some of cvdata's 'guts' to actually run a sim, but that's just a guess.
the reason i think this, is that you can evaluate canned sims and you can establish your own parameters that you dream up for a sim and then run a simulation, that takes a fair amount of time to produce results.
well, i can try, thing is i don't know how Lonesome Gambler resolved his true counts.
but the only thing i could do as far as the 100/hands seen/hr, is fool with the spinners after the simulation has been run.
SCORE i haven't really the foggiest about.
i don't know enough about the maths of N0 to be able to say if it is in any way time dependent or not. i kind of doubt it though, seems more like it would depend, playing strategy, betting, the nature of the game, but why would it care how long it takes to play a hand or a round?
i don't see why $win/hand, freq's at each TC or the W/L% would change as a result of playing more or less hands per hour. those things are what they are no matter how the clock ticks, lol, at least that's how it seems to me.
i can say for sure they don't change when you fiddle with the hands/hour spinner for a given cvcx simulation.
probably i'm missing your point, i'm definitely young (just turned 60, lmao) and easily confused.
:whip:Kasi,Kasi said:
Somethings wrong with the sim LG posted.
First of all the sim does not match his rules:
5D (3.5/5) S17 DAS RSA $5/250
BR: $5000
where the sim is s17 das, no rsa
the spread is 1-14 1x10- 2x70
using a departure point of 20 means he missing 20 hands every time he walks away from the table to find a new game which lowers the SCORE.
After doing a few different short sims which i stopped very short due to the results, the cvcx sim is very inaccurate vs the cvdata results.
The game simmed in cvdata using the written rules above, RPC with a larger optimum 1-24 spread 1x10, 2x120 is just barely break even coming in with SCORES between 1 and 5 and N0's in the 6 digits.
Just look at the original posted sim and it doesnt make sense.
s17 with only das, 70% pen 1-14 spread, WI at 0 and the SCORE's in the 60's.
I dont believe even if using Hi-Lo with the 1-14 spread that the results would come in any higher then the CVData sims I ran using RPC.
One last note, in cvcx it says that:
Departure Adjustment - This option only appears when Back-Counting is selected. CVCX normally assumes that you play all hands at or above the selected count. However, most people will stay at the table for awhile if the count dips below that Wong-In point. This feature is based on a methodology posted by DD’. When selected, an estimate is made of the SCORE and Win Rate as if you exited the table when the count has dropped to the first row of counts displayed (normally -1.) Note: This is only an estimate. CVData must be used to calculate the exact numbers. Also, you can enter a number. This is the number of hands that it would take you to find a new table. So, if you enter 10, you are estimating that it you will waste ten hands looking for a new table and the Win Rate will be decreased accordingly.
Now if my understanding of the above is correct, if you enter 20 in the departure adjustment, your leaving the table and missing a time frame = to the time to play 20 hands.
Going back to look at the OP's sim (http://www.blackjackinfo.com/bb/showthread.php?t=14612), if he's playing 52.6% of 80hands/hr back counting at TC0, thats 42 hands per hour and then if he's departing the table at the cost of an additional 20 hands per hr while he searches for a new table (42-20=22), he's playing only 22 hands per hr. To reach his "long run" or better known as N0 of 15,930 hands he will have to play for 724 hrs. Now what if he finds a table plays three hands and departs again? He may be playing only 17 hands an hr (just a guestimate) or even less. At this rate he'll play a lifetime to get to his N0 or the equivilent of 1 SD (per cvcx: N0: it is a measure of how many rounds must be played to overcome a negative fluctuation of one standard deviation with such a fixed spread." ).
Now based on the departure explanation CVCX only provide an estimate so we need CVData to give us the accurate info.
Wow
More facts to add to the confusion... BJC
QFIT
Well-Known Member
Changing hands/hour would have no effect on N0 or the % of hands played. You can change the number of hands after the sim and see the numbers that change instantly.Kasi said:That's OK QFIT. Thanks for trying. I forgive you in advance from responding any further to my silly questions.
I agree N0 won't change, like I think you say, in terms of physical rounds played, but I'd have to think, if someone ran the exact same sim at 100/hds/hr, hands under N0 would change from the 15930 and the 52.6% would also change for starters
Brett_Harris
Member
Interpretation of NO...
Guys,
I have to say, it has been 13 years since the idea of minimizing N0 came to me, and it took perhaps 18 months to really sort out what was going on. Mathematically (and computationly) it had to be defined as including rounds for which the bet is zero. But just because the bet is zero, it doesn't mean you are not playing the game - you are playing the game by accumulating information via the count
Of course the real world is different - do you include the table that you watched for five minutes and walked away because the count went south, in the definition of N0? The best I could come up with was to define watched hands to contribute to N0 and at the same time keep statistics of percentage of hands actually played.
In fact if you are comparing counting systems, this is precisely what you have to do - I may not have said this before, but in truth the factor which has the greatest effect on any shoe game is Wonging, no matter what system you use.
I believe that the game we call Blackjack is actually two games - 1 or 2 decks and shoes. That is, wongable and non wongable games. Single or double deck games seem to be peculiar to the US, perhaps that is the result of competition - but the rest of the world as much as I see it seems dominated by shoes. If one only has access to shoe games the following question seems rather obvious - if you only count and only bet on extreme counts, you win very little - if you play all rounds you need enormous betting spreads and with minimum bet limits a large bankroll. Of course the answer lies in between - is there is an actual optimal Wong point ?
Yes there is, and mostly it is around that Hi-Lo TC=+2 (~+0.5%)
(Red-7 anyone ;-)
Like many things in nature, when a dimensionless number like N0 appears, it can give an order of magnitude estimate of other things as well. A few years ago, I ran some simulations to determine the average number of rounds for a player to ruin for a fixed bettor using a fixed Kelly=1 bankroll and it turned out to be on average the same as N0.
My first reaction was surprise - isn't a lower N0 or higher SCORE a good thing? Of course N0 is not dimensionless - it is a measure of time, he number of rounds played etc... so there is a dark side to a good game - good games must have high EV, but this comes with high SD as well. Bankroll management is critical in these situations - before O/U-13 disappeared from this universe- the opportunity was enormous, but it was a real bumpy ride.
Cheers,
Brett.
Guys,
I have to say, it has been 13 years since the idea of minimizing N0 came to me, and it took perhaps 18 months to really sort out what was going on. Mathematically (and computationly) it had to be defined as including rounds for which the bet is zero. But just because the bet is zero, it doesn't mean you are not playing the game - you are playing the game by accumulating information via the count
Of course the real world is different - do you include the table that you watched for five minutes and walked away because the count went south, in the definition of N0? The best I could come up with was to define watched hands to contribute to N0 and at the same time keep statistics of percentage of hands actually played.
In fact if you are comparing counting systems, this is precisely what you have to do - I may not have said this before, but in truth the factor which has the greatest effect on any shoe game is Wonging, no matter what system you use.
I believe that the game we call Blackjack is actually two games - 1 or 2 decks and shoes. That is, wongable and non wongable games. Single or double deck games seem to be peculiar to the US, perhaps that is the result of competition - but the rest of the world as much as I see it seems dominated by shoes. If one only has access to shoe games the following question seems rather obvious - if you only count and only bet on extreme counts, you win very little - if you play all rounds you need enormous betting spreads and with minimum bet limits a large bankroll. Of course the answer lies in between - is there is an actual optimal Wong point ?
Yes there is, and mostly it is around that Hi-Lo TC=+2 (~+0.5%)
(Red-7 anyone ;-)
Like many things in nature, when a dimensionless number like N0 appears, it can give an order of magnitude estimate of other things as well. A few years ago, I ran some simulations to determine the average number of rounds for a player to ruin for a fixed bettor using a fixed Kelly=1 bankroll and it turned out to be on average the same as N0.
My first reaction was surprise - isn't a lower N0 or higher SCORE a good thing? Of course N0 is not dimensionless - it is a measure of time, he number of rounds played etc... so there is a dark side to a good game - good games must have high EV, but this comes with high SD as well. Bankroll management is critical in these situations - before O/U-13 disappeared from this universe- the opportunity was enormous, but it was a real bumpy ride.
Cheers,
Brett.
Talk about an epiphany lol. All it took was you and The Wise One telling me I'm nuts. As my wife sometimes says I now "gots it" lol. (Long story, once at work some woman was picking a corn or maybe "toe jam" and eventually announced to all "I got's it" lmao.)QFIT said:
This revelation seems to imply to me that the $0.29 "win/Hand" also would never change since, maybe, it really means "avg $win per hour"? Would you say that might be a more accurate description of what that field represents?
Either way, can we agree N0 would occur after approximately 8379 physical hands?
Here's my last area, I promise, of asking stuff. What does the $29.27 "Average bet" mean? Does it mean per physical "hand" played or some "avg initial bet per hour?" Also does it mean per "round" maybe rather than "hand" ie just for the one spot before spreading?
Is my total initial avg bet per physical round $29.27*8379 rounds=$245,259 or so after 8379 physical rounds?
OK - I lied, what is the definition of "W/L%? Would there be something wrong in saying it is $0.55133 per round/$29.27 per round= 1.884% and that that is the win rate per physical round?
Thanks for your expalnation of "Departure Point" - another thing I've never exactly understood what it may mean.bjcount said:
I guess I've thought, whatever it may mean, the effect of it is already included in the frequencies, adv at each TC and SD at each TC. That's all I usually use to try to figure my crap out. Whatever assumptions are made, that's what happens for those assumptions kind of thing.
So, now that I've been straightened out on what N0 means lol, it seems to me, in the OP's orig sim, whatever it was assuming, that N0 will occur after "seeing 15930 rounds". And, by extension, mean it will also occur after 8379 physical rounds played no matter how long it may take you to play those 8379 physical rounds.
That's sort of what I was alluding to earlier when I said i thought the N0number of hands did not always represent the number of physiacl rounds.
Doesn't mean, just becasue I may think it, makes it right lol. As already has been more than amply demonstrated.
I love it. Beat me again. Harder. lmao.
What do you think of N0 being 8379 physiacl rounds in this case?
Brett_Harris
Member
Perhaps to go back to the old quantum mechanics days, the rounds watched can be called 'virtual' hands as opposed to the 'real' hands :joker:
What a great post and just the fact you chimed in makes all my stupid questions worthwhile, if they had something at all to do with it.Brett_Harris said:
I had no idea N0 was a concept invented by you. That's how ignorant I am.
Would it also be true that the avg number of rounds, seen or played, for a fixed-Kelly=1 guy to double his roll would also be the same as N0?
sagefr0g
Well-Known Member
here's a neat link, Kasi....Kasi said:
http://www.bjmath.com/bjmath/Betsize/theory.htm (Archive copy)
no wonder it took him i think he said 19months or so.
sagefr0g
Well-Known Member
it's good to hear a maths guy like you say that.Brett_Harris said:
it's been my similar qualitative sense of the goodness and badness factors of some games, like for instance, penetration.
don't get me wrong, i like to see excellent pen, thing is though, say in this really world we live in, you don't see excellent pen all that often.
well, as good as it is, like you say, there can be a dark side, like it may be the case say those great tc's just don't come, or maybe they do but you still lose. either way, your betting up into or playing through a longer lasting situation and if things don't pan out, you can take quite a hit.
and that risk is for what may be this rare situation that is only a fleeting moment in time, sort of thing.
reminds me of the time i took a scavenger play opportunity on a table mate's double down on a black chip bet that he couldn't cover. darned if i didn't lose that bet and it was 'over betting' for me as well and haven't had another similar opportunity in a couple of years now.:eyepatch::whip: