Calculating House Edge

Unleash

Member
I got a question for your guys...
when using the basic strategy engine on this website for these rules

6 decks, H17, DAS, Early Surrender, No Peek

http://www.blackjackinfo.com/bjbse.php?numdecks=6+decks&soft17=h17&dbl=all&das=yes&surr=es&peek=no

I get a house average of 0.14 however, when I calculate it using Arnold Snyder's calculation I get 0.00% so even...

http://www.blackjackforumonline.com/content/Calculating_the_House_Edge_for_Any_Blackjack_Rules_Set.htm

Six deck -0.54
Hit Soft Seventeen -0.21
No Hole Card -0.11
Double after Split +0.14
Early Surrender +0.72

Any Ideas on the differences between the two calculations??
I'm not sure why for BJinfo calculator they calculate early surrender to be 0.63% advantage for player instead of 0.72...

Anyways wondering your thoughts on the difference...
 

MangoJ

Well-Known Member
Assigning certain EVs to single rules is only a rule of thumb. You can do it for estimation purposes, but this is not an exact science.
The reason is, rule EVs are not independent.

The gain of EV of rules certainly depends on all other parameters of play. It is easy to see that all hands have different EVs against a S17 dealer compared to a H17 dealer. Therefore, a surrender option cannot increase the EV by the exact same amount in both situations, and hence - summing up fixed EV gains from surrender and H17 can only yield an estimate.

I'm not a fan of summing up those EVs for game selection - as this estimate can easily get wrong, and you don't want to commit larger sums on estimates.
Better stick to the relevant parameters of your games: penetration, spread, proximity, heat.
 

bj bob

Well-Known Member
Unleash said:
I got a question for your guys...
when using the basic strategy engine on this website for these rules

6 decks, H17, DAS, Early Surrender, No Peek

http://www.blackjackinfo.com/bjbse.php?numdecks=6+decks&soft17=h17&dbl=all&das=yes&surr=es&peek=no

I get a house average of 0.14 however, when I calculate it using Arnold Snyder's calculation I get 0.00% so even...

http://www.blackjackforumonline.com/content/Calculating_the_House_Edge_for_Any_Blackjack_Rules_Set.htm

Six deck -0.54
Hit Soft Seventeen -0.21
No Hole Card -0.11
Double after Split +0.14
Early Surrender +0.72

Any Ideas on the differences between the two calculations??
I'm not sure why for BJinfo calculator they calculate early surrender to be 0.63% advantage for player instead of 0.72...

Anyways wondering your thoughts on the difference...
I would certainly see if Snyder's calculation considered OBO.
 

iCountNTrack

Well-Known Member
Unleash said:
I got a question for your guys...
when using the basic strategy engine on this website for these rules

6 decks, H17, DAS, Early Surrender, No Peek

http://www.blackjackinfo.com/bjbse.php?numdecks=6+decks&soft17=h17&dbl=all&das=yes&surr=es&peek=no

I get a house average of 0.14 however, when I calculate it using Arnold Snyder's calculation I get 0.00% so even...

http://www.blackjackforumonline.com/content/Calculating_the_House_Edge_for_Any_Blackjack_Rules_Set.htm

Six deck -0.54
Hit Soft Seventeen -0.21
No Hole Card -0.11
Double after Split +0.14
Early Surrender +0.72

Any Ideas on the differences between the two calculations??
I'm not sure why for BJinfo calculator they calculate early surrender to be 0.63% advantage for player instead of 0.72...

Anyways wondering your thoughts on the difference...
Even though i hate to admit it, I have to agree with Snyder's on this one. A CDCA calculation shows that the EV for this game is -0.0032%. I think there is something not right with ES I will touchbase with Ken.
 

Unleash

Member
bj bob said:
I would certainly see if Snyder's calculation considered OBO.
( I believe) Synder's calculation does consider OBO well in this case ENHC "European No Hole Card" that's why it's a 0.11 disadvantage to the player in the calculation. I double checked this with Standford's book and it gives the same number for ENHC rule.
 

Unleash

Member
iCountNTrack said:
Even though i hate to admit it, I have to agree with Snyder's on this one. A CDCA calculation shows that the EV for this game is -0.0032%. I think there is something not right with ES I will touchbase with Ken.
Thanks for your help!
Just making sure my calculations were correct!
Would playing by the strategy given on the engine be correct then?
Or are there errors because of the different EV?
 

Unleash

Member
MangoJ said:
Assigning certain EVs to single rules is only a rule of thumb. You can do it for estimation purposes, but this is not an exact science.
The reason is, rule EVs are not independent.

The gain of EV of rules certainly depends on all other parameters of play. It is easy to see that all hands have different EVs against a S17 dealer compared to a H17 dealer. Therefore, a surrender option cannot increase the EV by the exact same amount in both situations, and hence - summing up fixed EV gains from surrender and H17 can only yield an estimate.

I'm not a fan of summing up those EVs for game selection - as this estimate can easily get wrong, and you don't want to commit larger sums on estimates.
Better stick to the relevant parameters of your games: penetration, spread, proximity, heat.
I understand but i'm using this casino strictly for comps because they have use a continuous shuffler (shuffling after every 2 decks)and i'm trying to minimize my losses...Thanks for your help! Greatly appreciated!
 

Unleash

Member
iCountNTrack said:
Even though i hate to admit it, I have to agree with Snyder's on this one. A CDCA calculation shows that the EV for this game is -0.0032%. I think there is something not right with ES I will touchbase with Ken.
I actually have one more question...if the casino i'm using these set of rules with uses a continuous shuffler does that affect the EV? and if it does for me or the casino??

Thanks
 

alwayssplitaces

Well-Known Member
The continuous shuffler helps the BS player. But if they shuffle after 2 decks, that's an effective 33% penetration. Counting would turn this into a +EV game since the house edge is so low, even with the CSM.
 
Top