CVCX/CVData Suggestions List
- Thread starter MJ1
- Start date
That is a given. Everything users enter is an estimate. We estimate PEN, # players at table, rounds dealt per hour, playing errors, betting errors, dealer errors, etc.QFIT said:
One thing I learned about simulations is that they help us to construct a model based upon many variables. Unfortunately, output can be quite sensitive to some of these variables like PEN. But we do the best that we can. An estimate is certainly better than nothing at all.
Last edited:
sagefr0g
Well-Known Member
the method is basically:QFIT said:
raised bet if previous hand was:
seven or more cards between you & the dealer
or
double down (if a hard double down, not a soft double down)
or
split hand (but not split aces)
or
push ( excepting four card pushs between you and the dealer)
minimum bet if previous hand was:
four cards between you & the dealer
or
aces were split
or
ordinary win or ordinary loss
the kind of data of interest would be of the type in this link:
http://www.blackjackinfo.com/bb/showpost.php?p=134835&postcount=44
fwb
Well-Known Member
Another suggestion: Let us specify process priority, maybe right under where # of threads are set. I like doing other things when my sims are running and don't care if they have to give up 1% of the processor for web browsing. I found myself constantly going into task manager and setting CVdata priority to Low, though I found a task manager utility which can do this automatically now.
Also: with the setup I'm currently working with, there appears to be a bug with the max # of players allowed at a table. It will not let me increase from 4 to 5 players. One player plays up to 3 hands while the rest only play 1. Thats only 6 hands max, it should let me add one more player.
Actually, I came across the solution while typing this. Player 5 was left with a strategy which used multiple hands from a previous sim which is why it wouldn't let me add him. I had to reduce all players to 1 hand, add player 5, reduce his max hands back to 1, and then change everything back. Maybe this isn't as user-friendly as it should be?
Also: with the setup I'm currently working with, there appears to be a bug with the max # of players allowed at a table. It will not let me increase from 4 to 5 players. One player plays up to 3 hands while the rest only play 1. Thats only 6 hands max, it should let me add one more player.
Actually, I came across the solution while typing this. Player 5 was left with a strategy which used multiple hands from a previous sim which is why it wouldn't let me add him. I had to reduce all players to 1 hand, add player 5, reduce his max hands back to 1, and then change everything back. Maybe this isn't as user-friendly as it should be?
iCountNTrack
Well-Known Member
Hit-Card (next card) simulation feature, very similar to the hole-carding one which is brilliant.
iCountNTrack
Well-Known Member
the simplest approach i can think off is to have a strategy table for each type of next card ace through 10 for a total of 10 tables. Each table will simply have the dealer upcard and the player's cards, each entry in the table will have the optimum player's play (hit, stand, double, split).QFIT said:
iCountNTrack
Well-Known Member
It could be enormous, my guess the ev should be over 30%.QFIT said:Now, what is your gain if you know both the hole card and next card?
100% next card carding (~17% ev) is more powerful than 100% hole carding (~11% ev), i guess because there is a large deterministic component with next carding like for instance never busting your hand.
aslan
Well-Known Member
I know CVDATA will allow one to set the point at which to begin increasing the bet to take advantage of a positive count, but I was not able to make it use a different strategy in reverse. That is, if you want to simulate oppositional betting (OB), you can tell CVDATA to bet one unit until a specified point, say, +7 for example, and then to increase all at once to max bet, but when the count begins to tank, one practicing OB as a masking technique might want to continue max bet down either to a set point other than the +7, say 0, or to continue at max bet beyond 0 until the the count stops falling (not that you can predict a run of big cards, but at least you can take advantage of any streak of big cards that might occur.) I am not saying you can make money trying to be in on "streaks," but you can evaluate this as a masking technique to see how much advantage you are sacrificing for the sake of masking your play. Also, if you were betting one unit until +7 as above, you might want to incrementally reduce your bet, 9 units, 8 units, 7 units, etc. as the count dropped, say to 0. As I recall, CVDATA applies the same strategy going down in the count as it does going up in the count. Please correct me if I am wrong about this.
Also, does CVDATA allow one to insert little masking flourishes from time to time, such as, when the count goes down, occasionally (or a set number of times) increase the bet one unit, and if one wins, keep the bet at two units until the count reverses. It's not to make money, just to mask one's play. Hopefully these masking techniques will come close to evening out in the long run. The purpose again would simply be to evaluate the sacrifice of advantage for the purpose of occasional masking, especially in play-all or wong-out-only situations.
Also, does CVDATA allow one to insert little masking flourishes from time to time, such as, when the count goes down, occasionally (or a set number of times) increase the bet one unit, and if one wins, keep the bet at two units until the count reverses. It's not to make money, just to mask one's play. Hopefully these masking techniques will come close to evening out in the long run. The purpose again would simply be to evaluate the sacrifice of advantage for the purpose of occasional masking, especially in play-all or wong-out-only situations.