Calculated Hourly < Actual Hourly income.

[JPat]

First time poster so please forgive any mistakes.
I play weekly at a friendly casino. Two deck / H17 / DAS / RSA / LS / Re-Split to 4 / BJ 3:2
I copied a basic EV calculation spread sheet and entered TC frequencies for 100 hands from Schlesinger's excellent book.
The resulting Per Hour calculation is $32.47.

Yet over the past 4 months of playing weekly (only 2-3 hours per week) I am averaging a little over $300/hour.
I understand that may average out over time to $200 or even $100, but...
Am I just having a great streak of good luck or am I calculating my spreadsheet incorrectly?
Please see spreadsheet example below.
Thank you!
Jack

[Freightman]

Nice rules - only thing you’re missing is S17. I’m just eyeballing your EV and spread. Looks to me that your EV calculation is really really low based on $25 min. What I see as missing is deck pen. Depending on what that % is will affect your EV though I see no way how that could be any less than your calculated $32.47 per hour.

Suppose I could calculate deck pen by evaluating TCfrequencies and working backwards - something I’m not prepared to do from my Starbucks satellite office.

PS - your EV calculation is really really really Low.

[Freightman]

Thinking more on this interesting point - I’ll be quoting Don on these comments though I’m surmising only and he can attest to the accuracy of my comments.

Don himself will comment on software simulation improvements over the years and its impact on a number of points. At some point either in BJA3 or in posts, he has commented on his ability to stay welcome by utilizing an increase of 1 unit (I think $100 in his case) per increase in true count (playing 1 hand) and maximizing at 2x6 units (think that was shoe v DD) at true 5.

This is far from optimal and would likely compare to the 1.25 units units per hour you have quoted (based on shoe which is where most of my sims lie) vs optimal ramping better calculated through computer software and likely influenced by Kelly Criterion which you have shown and would result in a very significant increase in EV over what you’ve stated.

If you would quote a page number from BJA3, then additional comments can be given. Also, advise your count system.

[Archvaldor]

You can't read anything into such a small sample size. 4 months of casual play means almost nothing.

[DSchles]

I'm surprised that Freightman was able to comment on your spread (apparently from the allegedly attached spreadsheet). I can't open anything below your post. Wonder why.

In any event, You can see from BJA3, page 250, that the possible hourly EVs range from 15.36 to 111.30, based on penetration and bet spread, so without seeing those, it's hard for me to comment other than to say that, yes, at $300 an hour, you've been the beneficiary of some good luck. We can quantify that in SDs above the mean, but not until you furnish the missing information (or tell me why I can't see the spreadsheet!). Note that my Summary charts don't allow for RSA, so all the numbers would increase by a little more.

Don

[DSchles]

Don himself will comment on software simulation improvements over the years and its impact on a number of points. At some point either in BJA3 or in posts, he has commented on his ability to stay welcome by utilizing an increase of 1 unit (I think $100 in his case) per increase in true count (playing 1 hand) and maximizing at 2x6 units (think that was shoe v DD) at true 5.

Yes, I never played two hands at DD. Rules were so good at the old Riviera and Maxim (S17 plus all of the above except RSA) that I thought it prudent not to be a pig. But no, I never increased by only one unit at a time. For DD, I most often used a 1-6 or occasionally 1-8 spread, so probably ramped, beginning at +1, to 2 units, then 4 at +2, 6 at +3, and 8 at >=+4. And yes, for 6-deck, 2 x 6 units at >=+5.

Don

[DSchles]

You can't read anything into such a small sample size. 4 months of casual play means almost nothing.

I agree with that.

Don

[Freightman]

I'm surprised that Freightman was able to comment on your spread (apparently from the allegedly attached spreadsheet). I can't open anything below your post. Wonder why.

In any event, You can see from BJA3, page 250, that the possible hourly EVs range from 15.36 to 111.30, based on penetration and bet spread, so without seeing those, it's hard for me to comment other than to say that, yes, at $300 an hour, you've been the beneficiary of some good luck. We can quantify that in SDs above the mean, but not until you furnish the missing information (or tell me why I can't see the spreadsheet!). Note that my Summary charts don't allow for RSA, so all the numbers would increase by a little more.

Don

Spread was $25-$200, properly ramped, $200 at true 4 flat betting $200 thereafter. Frequencies per true count looked off and average bet of $56.75 simply makes no sense. I’ll have a peek at page 250 and may have additional comments.

[Freightman]

I’ll have a peek at page 250 and may have additional comments.

Utilizing my pet theories of QTC enshrined within the vaunted FBM ASC, I would expect returns well in excess of those stated, though one would need to subscribe to my philosophies of maximizing revenue over n sessions. I would suspect that DD volatility would be such as to make this a very worthwhile trip.

For this rule set coupled with good penetration and blessed with tolerance, I would be happy to fly to this location incurring those additional cross border expenses including medical insurance.

[DSchles]

Spread was $25-$200, properly ramped, $200 at true 4 flat betting $200 thereafter. Frequencies per true count looked off and average bet of $56.75 simply makes no sense. I’ll have a peek at page 250 and may have additional comments.

No, the average bet is perfectly fine. When you ramp that way, the average bet is usually just a little more than two units per hour. CVCX gives $55. But, the hourly win rate is obviously completely off. I get $89. And hourly SD is $916.

So, at three hours per week, for four months (17 weeks), that's a total of 51 hours of play. Expected win would be 51 x $89 = $4,539, with SD of sqrt (50) x $916 = $6,477.

Now our OP says he averaged about $300 per hour, so let's say he won 51 x $300 = $15,300, which is $10,761 about expected win, or 10,761/6,477 = 1.66 SDs to the right of expectation. This happens with probability of 4.85%. So, as I said, far from impossible, but some very good luck.

Don

[JPat]

Spread was $25-$200, properly ramped, $200 at true 4 flat betting $200 thereafter. Frequencies per true count looked off and average bet of $56.75 simply makes no sense. I’ll have a peek at page 250 and may have additional comments.

Thank you all for responding. Not sure how to respond without including previous quote.
I use the simple Hi-Low system.
The PEN is 75% and Freightman listed my bet spread correctly.
I'm not sure why the attachment doesn't open for everyone.
It is a pic of the spreadsheet. If anyone can tell me how to attach the actual excel sheet I will do that.

For advantage I started with -.263% at 0 TC and moved .05% at a time.
The spreadsheet uses this formula: (bet size * frequency over 100 hands) * advantage = Gain per TC.
I added all of those together to get Total Gain of $42.4246

For Hands Played I added all the frequencies together to get 97.993 (I didn't go below -2 TC and assume that accounted for the difference between this number and 100.)
I multiplied Frequency * Bet Size to get total bet per TC and then added all of those together to get Total Bet of $5856.600

Average Bet per Hand: I divided Total Bet by Hands Played to get : $59.7655.
Gain Per Hand: I divided Total Gain by Hands Played to get: 0.4329.
Hourly Win: I multiplied Gain Per Hand by #Hands per Hour (75 in my Casino) to get $32.470.

I may be way off in how I set up the calculations.
On top of all of this, my Casino does have a couple tables in the High Limit room with all the same rules AND dealer stands on S17.
Minimum bet is $50 rather than $25 so I don't play as often. Also, a lot less people playing so I feel like my play is more closely watched.
Would you recommend that game over the outside tables I am playing?

Thank you again!
Jack

[DSchles]

For advantage I started with -.263% at 0 TC and moved .05% at a time.

It would make me feel a great deal better if you meant to write 0.50%. In any event, you can't even do it this way because, as index numbers kick in, especially for 16 vs. T (0) and insurance (+3), the gain per true count is greater than 0.50%.

Finally, we all assumed 100 rounds per hour, so if you used 75, that obviously affects the calculation. Regardless, your number is much too low. Your gain per hand should be around $1.10.

Don

[JPat]

Yes, my bad. Meant 0.5%. Important in a discussion like this.
I only used 75 hands per hour for the final calculation of Hourly Win.
Appreciate your help and I agree re Gain per Hand.
If anyone reading my post above can see where I am going wrong, it would be greatly appreciated.
Or, let me know the formula you use to calculate Gain per Hand.

Side Note: I am pretty disciplined about setting a loss floor and leaving as well as reaching a net win total (say $1000 and walking away rather than dipping back down).
Probably explains why my normal playing time averages about 90 minutes.
Could that contribute to why my supposed hourly win is higher than expected?
Jack