Weekend Warriors

sagefr0g

Well-Known Member
#1
Officials Rules:
Rules: -.40
6D,NRSA,DAS,S17,DA2,

Conditions:
Play-all,80%pen,Heads-up,1-24spread,(25unit$)

for hi-lo bet as follows:
tc <= 0 $25
tc = 1 $150
tc = 2 $325
tc = 3 $500
tc >=4 $600

KO players might consider the betting suggested in this link:
http://www.blackjackinfo.com/bb/showpost.php?p=82241&postcount=11

Highlights of game - EV $1.06 per hand, SD $224 per hand, Avg bet $108, EV 0.98%, ROR 1.43%, N0 a mere 44,200 hands away lol. SCORE 22.6.

all one has to do is count hi-lo or count your favorite count and bet consistently and just play basic strategy if you want or indexes if you want. anybody is welcome to play alot or a little BS.

any software that can cut to 80% for that game is all you need or just use physical cards.



post results in this thread
 

Attachments

Last edited:

sagefr0g

Well-Known Member
#2
kc's post moved

k_c's initial results
http://www.blackjackinfo.com/bb/showpost.php?p=82218&postcount=9

k_c said:
Had some time to kill because of holiday. Went through 1 shoe using my program with suggested rules/bet ramp and recorded results by hand for heads up play, 80% pen. My computer is slow so it takes longer than I would like, particularly for 6 decks. I guess a logging feature and ability to save practice rounds to a database would be the next thing to add to the program if I ever get to it.

Counting system used was Hi-Lo but any count could be used, balanced or unbalanced. Only 1 bet above 1 unit was encountered and it turned out player won with 10-10 vs dealer's 10-7. A hand like this is 50-50 that player or dealer will win. EVs are for using basic strategy and basic strategy was always used. No insurance was used and there were no instances where insurance should have been taken anyway. The only index plays that came up were for hitting rather than standing in negative count situations and that happened maybe once or twice but basic strategy was used even though the index play called for something different.

Code:
6D, DAS, DOA, split to four hands, one card on aces, no surrender,
80% pen, play all (S17)

$25 until TC+1 is achieved. TC+0.9 is still $25. Then $150 at +1 until 2,
$325 at +3, $500 
at +4 and $600 at +4 or greater.

<1 1 unit
+1 to <+3 6 units
+3 to <+4 13 units
+4 20 units
>+4 24 units

Initial HA .4128%

TC   Actual EV   Bet   Result
0.0  -.4128      1     +2
+0.3 -.1905      1     +1.5
-0.3 -.5426      1     +1
-0.4 -.6063      1     -1
-0.4 -.6181      1     +1
-0.7 -.7470      1     -1
-1.3 -.9983      1     -1
-1.1 -.9706      1     +1
-1.2 -.9863      1     0
-0.8 -.6952      1     +2
-0.8 -.6723      1     -1
-0.2 -.3500      1     0
-0.4 -.5551      1     +1
-0.9 -.7923      1     -1
-0.9 -.8130      1     -2
-0.5 -.3439      1     +1
-1.1 -.7646      1     -1
-0.9 -.6988      1     +2
-1.2 -.8387      1     +1
-1.0 -.7239      1     +1
-1.3 -.8265      1     -1
-0.3 -.1404      1     +2
-0.5 -.4012      1     -1
0.0  -.1951      1     -1
+1.1 +.4048      6     +6
+0.6 +.0531      1     -1
+0.6 +.2738      1     0
-0.6 -.3660      1     +1
0.0  -.0703      1     +1
0.0  -.0253      1     +1
+0.3 +.2485      1     -1
-0.4 -.1664      1     0
-1.1 -.5363      1     -1
-0.8 -.4400      1     -1
-1.6 -.8722      1     -2
-2.1 -1.1372     1     -1
-1.7 -.9598      1     -1
-3.1 -1.6576     1     -1
-3.3 -1.7538     1     +1
-3.9 -2.1087     1     -1
-1.1 -.7430      1     +1
-1.1 -.6559      1     +1.5
-1.8 -.7486      1     0
-1.3 -.6530      1     0
-1.4 -.8120      1     -1
-1.4 -1.1479     1     -1
-0.8 -.8285      1     -1

Total for shoe:  +5 units
k_c
.........
 
Last edited:

Kasi

Well-Known Member
#5
Tarento said:
What would your betting units be if the rules were Hit on Soft 17 and insurance was offered?
Not sure why you're asking but they could be the same or different. If the same, it might produce a little different win rate and ROR kind of thing.

Even in this game we could have chosen a 10-1 spread or really anything at all. It's not like it's fixed in stone just because of the rules. Lots and lots of ways to bet your money in the same game. Hopefully you chose a way or 2 that makes sense to you before you play.

Like this spread wouldn't make sense in the real world probably because bets are probably jumping around too much. Maybe even the game itself doesn't make sense as the best way to bet $100K lol.

So if you want to play that spread with H17 and insurance or insurance and other index plays with play money just for the experience of counting and betting and being able to see how your results compare to expected, go right ahead. Just let us know if you post results that you were doing a slightly different thing. Do you use Hi-Lo?

Love to have you give it a shot.
 

k_c

Well-Known Member
#7
Conditional expectations in blackjack

Since you went to the trouble of moving my post I thought I'd give some observations on the shoe I dealt programatically. Nothing really earth shattering but hopefully scientific.

Average true count for shoe's 47 hands: -.87
Average change in true count (46 times): -.035
Above shows that on average more high cards were dealt than low cards.

Player's average actual EV for shoe's 47 hands (flat bet): -.6295
Player's average actual EV for shoe's 47 hands (using given bet spread): -.5692
Player would expect to do better than above because as it turned out more high cards than expected were dealt (causing actual EV to tend to decline.) Player's EV is lower than original HA but since more high cards were actually dealt he would expect to do better than the original HA. (The original HA is replaced by the actual EV for hands after the first hand.) If it turned out that a lot more low cards than expected were dealt then player could possibly have had the opportunity to raise the bet later on in the shoe and take his chances that he would win at that point but he probably would have done worse than original HA (or actual EV) prior to raising his bet.

Kinda shows the fickle nature of blackjack. Can't expect to have your cake and eat it too (usually, but if it happens you gotta like it :grin: ).

k_c
 
Last edited:

Canceler

Well-Known Member
#8
Here are the results of 10 shoes played with KO Preferred using the bet ramp sagefr0g provided.

A couple comments: One shoe went up well over the pivot point, and stayed there a long time. That helped a lot. On the other hand, one shoe ended up exactly at the IRC. Since I would expect the RC to rise over 19 points during an average shoe, I thought that was fairly strange.
 

Attachments

Last edited:

Canceler

Well-Known Member
#10
zengrifter said:
I'm not getting something - why only ten shoes? zg
A better question is, why are we doing this at all? I'm not sure. It's like a manually played sim. Maybe Kasi or sagefr0g can explain it.

But it's sorta interesting & kinda fun.
 
#13
k_c said:
How many meaningless samples does it take to identify a trend? Depends on the method.
You mean: How many meaningless samples does it take to identify a MEANINGLESS trend? Depends on the MEANINGLESS method?

Attention JACK-BOOTED MODERATORS - please move this thread to VOODOO. zg
 

k_c

Well-Known Member
#14
zengrifter said:
You mean: How many meaningless samples does it take to identify a MEANINGLESS trend? Depends on the MEANINGLESS method?

Attention JACK-BOOTED MODERATORS - please move this thread to VOODOO. zg
No. I mean a 1 round sim is probably meaningless...A 2 round sim is probably meaningless...A billion round sim may or may not be meaningless depending on method. Valid sims are just the summation of a finite number of meaningless (in the long run) data. Exact calculations in conjunction with sims can reduce the amount of trials needed for valid results.

ZG, I'm afraid you have a voodoo gene lodged deep within. The first thing you do when you don't understand is to play the voodoo card.:eek:

k_c
 

sagefr0g

Well-Known Member
#16
let them eat cake lol

k_c said:
Since you went to the trouble of moving my post I thought I'd give some observations on the shoe I dealt programatically. Nothing really earth shattering but hopefully scientific.

Average true count for shoe's 47 hands: -.87
Average change in true count (46 times): -.035
Above shows that on average more high cards were dealt than low cards.

Player's average actual EV for shoe's 47 hands (flat bet): -.6295
Player's average actual EV for shoe's 47 hands (using given bet spread): -.5692
Player would expect to do better than above because as it turned out more high cards than expected were dealt (causing actual EV to tend to decline.) Player's EV is lower than original HA but since more high cards were actually dealt he would expect to do better than the original HA. (The original HA is replaced by the actual EV for hands after the first hand.) If it turned out that a lot more low cards than expected were dealt then player could possibly have had the opportunity to raise the bet later on in the shoe and take his chances that he would win at that point but he probably would have done worse than original HA (or actual EV) prior to raising his bet.

Kinda shows the fickle nature of blackjack. Can't expect to have your cake and eat it too (usually, but if it happens you gotta like it :grin: ).

k_c
interesting, ok. i saw where you just had the one raised bet and won that. lol
lucky? and i think you came out +5 units? so that +5 you got left from the +6 you won on the raised bet to where out of the other 46 bets you managed to lose one unit of those six you won on the big bet? lol i guess you could look at it that way. i guess about a 1.1 unit average bet. so is like where your result is +1.5 that was a snapper? and if your result was +2 or -2 maybe a double down or split won or lost? then a +1 or -1 result just a normal hand?
well but 47 hands so using Kasi's spread sheet :
http://www.blackjackinfo.com/bb/showpost.php?p=82245&postcount=12
and make your unit = $25 to where your up five units.... up $125.....
 

Attachments

sagefr0g

Well-Known Member
#17
Canceler said:
A better question is, why are we doing this at all? I'm not sure. It's like a manually played sim. Maybe Kasi or sagefr0g can explain it.

But it's sorta interesting & kinda fun.
well it didn't have to be ten shoes. it could'a been one or one thousand. anyone can report however many they want.
but yes fun and interesting. a way to see maybe some of the things that are going on lol. you know like things going on behind your back so to speak. like maybe when your in a real casino playing, counting doing the AP thing and what ever happens and maybe you don't know (probably never could remmember what in the world transpired). you know your game plan or what ever you want to call it. so but like what we are doing here if you know just how many hands you played and how much money you won or lost there are some abstract maths things you can know about what's expected along standard deviation lines and expected value lines.
or too trip ROR maths, reaching a goal probability stuff.
this stuff for each individual player, or all of us as a team. the bjinfocomteam.
well too we can discuss these long or short haul plays as far as our counting experiences and the practice value of what we are doing. or just shoot the bull?
 

sagefr0g

Well-Known Member
#18
zengrifter said:
Oh, its manual! And whats the point of these tiny meaningless samplings? zg
whaaa? meaningless?
oh no, say it a'int so!
not you too zen? now i know you aren't voodoo! when you play a tiny sample say it ain't so. say it ain't voodoo and you play like a good little robot. lol
it is AP play, right? the real deal?
 

sagefr0g

Well-Known Member
#19
zengrifter said:
You mean: How many meaningless samples does it take to identify a MEANINGLESS trend? Depends on the MEANINGLESS method?

Attention JACK-BOOTED MODERATORS - please move this thread to VOODOO. zg
lol you gotta be kidding. you gonna have us banished to voodoo land for being good little robot counters?
geesh i thought the pit crews were lame. :flame:
 

Kasi

Well-Known Member
#20
zengrifter said:
You mean: How many meaningless samples does it take to identify a MEANINGLESS trend? Depends on the MEANINGLESS method?

Attention JACK-BOOTED MODERATORS - please move this thread to VOODOO. zg
Actually, if there is one thread that at least attempts to be not voodoo, this is it.

A completely meaningless large, let alone small, sample, to me anyway, is more like "I've played a really long time, thousands and thousands of hands, I'd guess, and I'm way ahead but I don't know how much but I've been counting and spreading 1-to something or sometimes 1 to something else but, I've won 12 sessions and lost 5 in a bunch of different games. Can anyone tell me my EV - I'm pretty sure I'm probably playing at an advantage?"
Is that card-counting to you?

See, here, even if you play an apparently meaningless small sample of 188 hands, you actually know what your EV is. You actually know if you lose x dollars in y hands, whether that's a 1 in 5 chance or a 1 - 1000 chance. You actually know what your N0 is and how long it will take to double roll. You know what ROR the bankroll you're betting to is and what your avg bet should be and what your overall win % is.

Eventually, perhaps, the MEANINGLESS small samples, depending as they do on the MEANINGLESS methods predicted by a sim, may eventually identify one of those many MEANINGLESS trends like the one where you will likely win money forever or maybe the MEANINGLESS trend where you know you how likely you'll be up by $10,000 at some point or the MEANINGLESS trend of answering "how come I keep losing these trip rolls?"

See, here, people are measured against a simmed plan(s) and try to do the best they can to follow that plan every hand they bet. All they have to do is count and bet as prescribed if and when they feel like betting consistently kind of thing.

As if they/anyone were on a team and somebody actually cared from time to time to how well their results compare to the EV of their plan. As if, as an individual, one might just want to attempt to compare his own results to his own sim and see what's involved with that. Maybe he might even learn/want to to do it for MEANINGLESS small samples just because he can.

Maybe 10 guys eventually play 2000 hands to a plan. We're already half-way to N0 for that plan. There's no time limit or limit on number of players choosing to participate or anything.

Broadly speaking, that's the general idea anyway lol.

Love to hear an opinion though from a JACK-BOOTED MODERATOR, not that I have a clue what that is, or from anyone else, if it makes even the slightest sense conceptually.

And, of course, just what strikes you as voodoo about it.

Give it a little while - it's early yet in the experiment - I think it'll make some sense to you eventually. There is work to be done that I haven't done yet without a doubt lol.
 
Top