Weekend Warriors

sagefr0g

Well-Known Member
#81
k_c said:
Went through another shoe with my program. This time I just recorded the shoe comp before each deal so any count can be used. I made a spreadsheet of the shoe including pre-deal overall basic strategy EVs. This detail as well as manually having to record data makes recording data slower so I am only considering one shoe at a time. If I incoporated automatic playing and logging into my program I could generate more data and it also would be much easier. I don't have time to do that right now. The average bet is the average of the bet for that round so splits do not count as an extra hand. I used the composition to compute bet spreads and results using Hi-Lo but the spreadsheet could be expanded to include any number of addtional counts. Basic strategy is always used since to compare counts using the same set of compositions requires the same strategy be used.

Results:
Flat bettor: -3.5 units

Hi-Lo
TC <1 1 unit
TC +1 to <+3 6 units
TC +3 to <+4 13 units
TC +4 to <+5 20 units
TC >=+5 24 units
Using above spread: +25 units

Multiply by 25 to convert to $25 units.
ok k_c i think what i'm getting here for our score board is now you've played
a total of 92 hands and your now up $37.50 . sound right?
by the way kewl spread sheet. i'm still digesting that lol. and yeah if you do incorporate automatic play and logging well i think i could part with some dough-re-me lol.
 

sagefr0g

Well-Known Member
#82
QFIT said:
CVCX average bet is based on initial bets. CVBJ counts split hands as separate. CVCX is used for statistical analysis and stats are based on rounds as this is the correct method for calculations. CVBJ is a practice tool and based more on hands. CVBJ is not really designed for any sort of statistical analysis since you would have to play for years to amass enough data to be meaningful.
that's what we were wanting to know. thank you QFIT!
 

k_c

Well-Known Member
#83
sagefr0g said:
ok k_c i think what i'm getting here for our score board is now you've played
a total of 92 hands and your now up $37.50 . sound right?
by the way kewl spread sheet. i'm still digesting that lol. and yeah if you do incorporate automatic play and logging well i think i could part with some dough-re-me lol.
No, 91 hands (rounds), up 30 units. (shoe 1, 47 rounds, +5 units; shoe 2, 44 rounds, +25 units)
30 units @ $25/unit = +$750.

I don't think short term results mean that much. I think the thing to look at is how well your counting system correlates with actual EV.
 

sagefr0g

Well-Known Member
#84
k_c said:
No, 91 hands (rounds), up 30 units. (shoe 1, 47 rounds, +5 units; shoe 2, 44 rounds, +25 units)
30 units @ $25/unit = +$750.

I don't think short term results mean that much. I think the thing to look at is how well your counting system correlates with actual EV.
lol sorry i confused the flat bet results with yours. :eek:
right anything can happen in the short term. weird how anything become something after a while lol.
ok so how do you make your counting system correlate with actual EV? :confused:
 

Kasi

Well-Known Member
#85
sagefr0g said:
that's what we were wanting to know. thank you QFIT!
Yes thank you QFIT.

I suppose what it means if you are using CVBJ that your total dollars wagered might be about 12% higher than the avg bet in the sim. (Just assuming 10% doubles and 2 % splits). So if your avg bet is running higher than the $108 it actually makes sense.

I suppose it also means that if splits occur 2% of the time and CVBJ says you've played 2000 hands maybe sticking 1960 hands in the sheet might be more technically correct.

And of course I really have no idea how often splits and doubles would be expected to occur. That 2% is probably high - that might be any potentially splittable hand and we don't split all splits lol.
 

sagefr0g

Well-Known Member
#86
Kasi said:
Yes thank you QFIT.

I suppose what it means if you are using CVBJ that your total dollars wagered might be about 12% higher than the avg bet in the sim. (Just assuming 10% doubles and 2 % splits). So if your avg bet is running higher than the $108 it actually makes sense.
i'm with ya so far. i think.
Kasi said:
I suppose it also means that if splits occur 2% of the time and CVBJ says you've played 2000 hands maybe sticking 1960 hands in the sheet might be more technically correct.
ok so in my case i could count up in my logs times that i've split and just subtract those from the cvbj total hands inorder to get a more technically correct number of initial hands.
so but if you used the percent of the time splits were expected that would give one a 'fuzzier' lol estimate of the number of hands to use for the bjinfoteam spread sheet.
Kasi said:
And of course I really have no idea how often splits and doubles would be expected to occur. That 2% is probably high - that might be any potentially splittable hand and we don't split all splits lol.
ok bear with me i'm young and easily confused. :confused:
the double down aspect of this issue is differant than the split issue. here we are introducing a larger than initial bet. so are you thinking that aspect is significant as well? i believe i see double downs doesn't introduce more hands but i guess from the aspect of determining or comparing your initial average bet it would make you believe it was higher than it really was. and the bets made on the split hands would do that as well.
so we need to take our percieved average bet with a grain of salt as well lol.
realize that it's gonna be circa 10% more or less higher than the sim's initial average bet. so but i guess we know that will naturally fluctuate anyway but we ought to find our average bet to be in the same ball park as the sim's initial average bet.
here's a link of interest from the 'good ole book':
http://www.blackjackincolor.com/truecount6.htm
 
Last edited:

k_c

Well-Known Member
#87
sagefr0g said:
lol sorry i confused the flat bet results with yours. :eek:
right anything can happen in the short term. weird how anything become something after a while lol.
ok so how do you make your counting system correlate with actual EV? :confused:
You can't do anything to make it correlate. It either is effective in predicting EV or it isn't. One shoe isn't conclusive but it's a start and maybe something to base designing something else on. With exact calculations you can pretty quickly get a decent idea of how effective a count is. I designed the spreadsheet so anyone can use the composition to add any other count. Any count that was added could then be compared to Hi-Lo or any other count that was entered to see how it performed for that one shoe or for more shoes if logging was added. Canceler could probably adapt his program to do the same thing as far as recording results for different counts without the capacity to record actual EV. He would have a larger sampling of shoes though.
 

Kasi

Well-Known Member
#88
sagefr0g said:
ok so in my case i could count up in my logs times that i've split and just subtract those from the cvbj total hands inorder to get a more technically correct number of initial hands...

the double down aspect of this issue is differant than the split issue. here we are introducing a larger than initial bet. so are you thinking that aspect is significant as well? i believe i see double downs doesn't introduce more hands but i guess from the aspect of determining or comparing your initial average bet it would make you believe it was higher than it really was. and the bets made on the split hands would do that as well.
so we need to take our percieved average bet with a grain of salt as well lol.
realize that it's gonna be circa 10% more or less higher than the sim's initial average bet. so but i guess we know that will naturally fluctuate anyway but we ought to find our average bet to be in the same ball park as the sim's initial average bet.
Well, for splits, say, if your first 2 hands got split 4 times CVBJ would report 8 hands played I assume? The sim assumed you played 2 rounds or versus 2 dealer upcards. So just knowing number of split hands wouldn't necessarily be all you'd need to know.

For doubles, I guess you should just keep in mind that your perceived avg bet, if based on total dollars wagered, will always be overstated compared to the sim.

It's kind of like alot of players, even BS players, knowing the HA is 0.5%, will say I wagered a total of $100,000 so my EV is -$500. If doubles were 10% in that game, then their EV is actually closer to -$450.

The sim win% is not intended to be applied against total dollars wagered to determine EV. The sim already has figured out all the doubling and splitting stuff and expressed everything in terms of avg initial bet. And, to the sim, a "hand" played actually means you played against 1 dealer up card even if you split 4 times and doubled each of those hands even though CVBJ would tell you you've wagered 8 units and played 4 hands.

So, if Canceller's "avg bet", if defined as total $'s wagered/by num of hands played, is running 10% or so higher than the sim predicted "avg bet", chances are he's "right-on" betting-wise lol.

Now how anybody keeps track of the number of initial dealer upcards they have played against is up to them. The more they actually they know that, the closer the sim will reflect EV.

But, like in real life, reasonable guesses are acceptable. That's all one can do if playing without the benefit of knowing exactly dollars wagered etc.

So 10,000 hands later, if that's what CVBJ tells you, plug in 9800 or 9900 hands and see what diff it makes lol.

Obviously splits and doubles are rule-dependent. D9 or NDAS etc.

So I guess the best thing about all this minutiae? has been it's useful to realize what a sim is telling you and what it's assumptions are and even what a "hand" might mean lol.

You're just burdened by all the info CVBJ gives you and what it means lol.

Which is why I'm just gonna use the BJ Trainer here next week when I get back and only know how many dealer upcards I played against and won't have a clue of totall dollars wagered or how many splits I had etc. I'll just know I'm really really unlucky aka I probably stink but I'd rather blame it on just my bad luck :)
 

sagefr0g

Well-Known Member
#90
Weekend Warriors:

Rules:6D,S17,DA2,DAS,/HA-.40

Conditions:Headsup,Playall,80%,Pen,100KBR

Warriors:

Dollars....Hands played

Sagefrog////+28,248$....1925
Canceler////+461$.........4269
Ghostrider//+10,838$.....7036
Aslan//////+425$...........325
K_C///////+750$............91
Kasi///////n/a...............n/a

____________________________

Total:+40,722$........13,646
 

Attachments

Last edited:

jack.jackson

Well-Known Member
#92
k_c said:
No, 91 hands (rounds), up 30 units. (shoe 1, 47 rounds, +5 units; shoe 2, 44 rounds, +25 units)
30 units @ $25/unit = +$750.

I don't think short term results mean that much. I think the thing to look at is how well your counting system correlates with actual EV.
I tell ya what kc. If you make it to 40 units, Ill grant ya a nickname.
 

jack.jackson

Well-Known Member
#93
Weekend Warriors:

Rules:6D,S17,DA2,DAS,/HA-.40

Conditions:Headsup,Playall,80%,Pen,100KBR

Warriors:

Dollars.........Hands played

Sagefrog//+28,248$....1925
Canceler//+461$.........4269
Ghostrider+21,275$.....10,036
Aslan////+425$...........325
K_C/////+750$............91
Kasi/////n/a...............n/a

____________________________

Total:+51,159.......16,646
 

aslan

Well-Known Member
#94
jack said:
Weekend Warriors:

Rules:6D,S17,DA2,DAS,/HA-.40

Conditions:Headsup,Playall,80%,Pen,100KBR

Warriors:

Dollars.........Hands played

Sagefrog//+28,248$....1925
Canceler//+461$.........4269
Ghostrider+21,275$.....10,036
Aslan////+425$...........325
K_C/////+750$............91
Kasi/////n/a...............n/a

____________________________

Total:+51,159.......16,646
Ghostrider,
I thought this was supposed to be "weekend" warrior, not "24 hours a day" warrior! Lol You need to get some sleep! Poor froggy! He'll probably be frogLEGS by morning! Lol Go get him, O gallant galloper!
 

jack.jackson

Well-Known Member
#95
aslan said:
Ghostrider,
I thought this was supposed to be "weekend" warrior, not "24 hours a day" warrior! Lol You need to get some sleep! Poor froggy! He'll probably be frogLEGS by morning! Lol Go get him, O gallant galloper!
Iwas about to say the same thing to you....lol

I seen you up all nite last night, into the wee hours of the morning:laugh:
 

sagefr0g

Well-Known Member
#96
aslan said:
Ghostrider,
I thought this was supposed to be "weekend" warrior, not "24 hours a day" warrior! Lol You need to get some sleep! Poor froggy! He'll probably be frogLEGS by morning! Lol Go get him, O gallant galloper!
yep history repeating it's self. allways happens.......darn that ghostrider! :whip:
 

Attachments

sagefr0g

Well-Known Member
#99
Quote:
Originally Posted by Canceler
It's not just ghostrider that's gunning for you!

(Oops, we're supposed to be a TEAM here! )


jack said:
Holy sh:eek:t The Cancelers tryin to take second.
e too canceleree?
 

Attachments

Last edited:

sagefr0g

Well-Known Member
k_c said:
You can't do anything to make it correlate. It either is effective in predicting EV or it isn't. One shoe isn't conclusive but it's a start and maybe something to base designing something else on. With exact calculations you can pretty quickly get a decent idea of how effective a count is. I designed the spreadsheet so anyone can use the composition to add any other count. Any count that was added could then be compared to Hi-Lo or any other count that was entered to see how it performed for that one shoe or for more shoes if logging was added. Canceler could probably adapt his program to do the same thing as far as recording results for different counts without the capacity to record actual EV. He would have a larger sampling of shoes though.
ok k_c below i have an image of your spreadsheet. maybe it's a lot to ask. but well if you can it would be interesting to have an interpretation or sort of an explaination. sort of in line with what you were explaining above.
 

Attachments

Top