The point is that an excess of low cards early means an expectation of an overall losing shoe.
The pure opposit is true.To see low cards early is the dream of any card counter as this allows him to attack, the remainder of the shoe being rich in high cards.
Francis Salmon
I thought of you, Moose, when I saw the article about the pending 'retirement' of Mr. Lucky. So I guess the Fritz will be back on your list of places to play.
As they say in Maine:
"Never taunt a Moose!"
@^D
Cyclops
I ran sims in the following conditions:
6-decks, ENHC, 80% penetration, 1 player at table.
1. Wonging with $50-$250 spread (1-5) returns $29 per shoe with a std deviation of $627.
2. Playing all with $10-$250 (1-25) spread returns $18 per shoe with a std deviation of $432.
Based on this both strategies look equivalent (the ration EV/stdev is approximately the same, so Kelly-optimal betting should return the same in both cases).
is almost insufficient when giving sim comparisons in the present day because there are several different models. The backcounting figures given in the bj attack charts are based upon stationary Wong in/Wong out at +1. This is very inefficient. A second model, the white rabbit, starts fresh shoes and Wongs out and departs table at -1. A third model, Mr Pefect, Wongs in at +1, Wongs out but remains stationary at the current table counting until about -1TC, and then departs to find a new table at the optimal departure point, approx -1TC. And, of course, the play all model stays and plays at the same table always.
The play all model does worst, of course. And of course, Mr Perfect does best. The suprise to most is that the White Rabbit, who does not backcount and continues to play all the way down to -1 TC, wins far more than the stationary Wong in/Wong out model and wins nearly as much as Mr Perfect.
Since your two models come out with similar SCORES, it has to be assumed you are using the least efficient and least effective Wonging model. When using a superior Wonging model you can't use "per shoe" stats as the modern Wonger does not observe all hands of the shoe but leaves at the optimal departure point. The play all model may have a higher win rate "per shoe" but will play far fewer shoes. The Wonger plays many more shoes since he'll be departing during the 1st, 2nd, 3rd, and 4th decks, but the shoes will be made up of fewer rounds. The comparison charts in the 2nd edition of BJ Attack show both instantaneous table changes and also a comparison that assesses a 6 round time penalty per table change, a more fair method.
In fact I was using KO count, with "Wonging" meaning entering/leaving at the Key count (when EV becomes >0) and remaining at the same table for the whole shoe. I did not realize this made Wonging less effective than leaving the table at the optimal time (however when you go to another table you may have to wait for a new shoe to start, thus wasting some time).
(however when you go to another table you may have to wait for a new shoe to start, thus wasting some time).
This is why I advocate active eyes while at the current shoe. As the count approaches the departure point you scan the pit looking for shuffling dealers. The BJ Attack comparisons take this into account by simply charging an average 6 rounds expectation per table change.
Whatever! Anyone who has seen any horror movie knows that nothing can keep evil chained up for long.
Maybe the museum curator will be interested in a purchase...
M.
Using the optimal exit point provided in KOBJ p118-120, the rabbit plays on average 33% more shoes than a stationary player.
So in an hour of play he makes 33% more money, and improves his Sharpe ratio (EV/stdev) by +11%.
Using the optimal exit point provided in KOBJ p118-120, the rabbit plays on average 33% more shoes than a stationary player.
A "shoe" is not a thing that can be considered any valid measure. They will differ in # of rounds, composition, etc. The rabbit makes 33% more if the Rabbit's shoe, which consist of fewer rounds, has the exact same expectation as the stationary player. This can't be the case. The stationary back-counter gets credit for every round in a shoe where the count goes over +1. The rabbit departs shoes which will often turn back and have counts higher than +1 later. So a "shoe" must be worth more to the stationary model (WI/WO). The rabbit gets more shoes, each shoe being of less value.
If you get per round statistics such analysis would be much easier.
