Keeping blackjack records.

[bec88]

Ever since me and my wife have been to casinos, i keep an accurate record in excel of how much money we have lost/won on any trip to the casino, so i know exactly how far in the hole we are gambling in our lifetime.

We both recently started to count cards in the casino after lots of practice, we are both transitioning into an attitude where we are getting more reluctant to play any -EV game unless we deem the pure entertainment value is worth it.... I have been keeping an excel spreadsheet with information on my sessions and i wanted to get ideas on if i am keeping the correct information for future reference..

So far i keep the...

• Date
• Casino
• Game and HE%
• Penetration
• Bet spread used
• Hours played
• Hands per hour (i estimate..60hph for full table, 80 for 2-3 total players etc etc..)
• Bankroll amount at start of session
• Bankroll amount after session
• Session Win/Loss
• Expected Win
• Hourly EV (based on sims of the game and hph etc)
• Hourly AV (actual session win or loss / hours played)
• SD of the particular session.
• Any specific notes on the session i deem important to log.

I cannot think of much more information to keep, wanted to see if this is good enough records, that way if i am missing something vital i can add it now and not have many more sessions logged without needed info....
One thing i was wondering. If i employ different spreads of a different night or my wife plays a different spread... assuming i put in the correct expected win that matches that exact spread then i can still accurately total my expected win and see where my actual bankroll is to see how much ahead or behind i am. Ive been playing a 1-8 spread so far, my wife has started to take any "gambling" money we might have taken on a weekend night and started to count cards with it...( hard not to when you know how right? lol) I figure it would just be easier to add those funds to our dedicated blackjack bankroll and actually fund her a session bankroll she would need for a night... As long as i keep track of how much expected win that bankroll is expected to generate in a given time.

Thanks for any input..

[ZeeBabar]

Some people, full time professionals, see value in keeping these records, some even more detailed including shifts played, dealers names and description etc. Othe other hand, some have done it for a while, evaluated how much or how little a particular factor has been used and discarded the information they have not used.

I started keeping records but much less detailed (amount won/lost, name of casino, DD or 6D, length of session). Because I am a part time recreational player, it's enough for me.

[gutshot]

I keep similar data for my records but go a step further and make charts to better see what is going on with the data I have collected. You can make endless charts but just a few I find useful is a chart of my bankroll after each session, amount won/lost per session, etc. I have also plotted total won/lost per rules type in conjunction with house edge and penetration. For instance, the best game (lowest HE) I play is also my biggest losing game mainly due to having the worst penetration and bad run of cards.

[Stealth]

I am amazed at the discreet, detailed, anal, analysis conducted by blackjack players in order to plan their play to 3 decimal places and then execute their play at an advantage. Tools, like CVCX and CVData are invaluable in these efforts. Then I see a disconnect by not doing the work required to track your play at sufficient detail to compare it to you planned results. If any manufacturer did this in his production process he would have a disaster. I contend detailed feedback on your actual play is equally important to planning your play but hear of many that only track hours played and win/loss and then estimate what their EV should have been.

I track the following:

Player

Play Type

Dates

Start Time

Session Actual

Game Rule Pen

$Unit

Bet Ramp

Wong Out %

Rounds

# Spots

Casino

With this I can compute a myriad of metrics that provide insight into my play and how I am managing it. Here is a summary chart of actual play that illustrates my point.


StealthDD.JPG


This is one of many reports and charts that I track. I realize this is not necessary for the casual advantage player but I am hard pressed to believe it is not an important aspect of the consummate AP.

Other opinions are always welcome.

PS: With this type of record keeping I can follow my journey as shown below.

StealthDDHistory.JPG

[NotEnoughHeat]

PS: With this type of record keeping I can follow my journey as shown below.
StealthDDHistory.JPG

That's a very nice graph. I really only have one graph I put any weight in (cumulative win) since the others either don't represent anything meaningful or redundant. I really like the SD ranges. I'm going to steal this idea. I'm guessing one way you could compute them is just plot the upper and lower ranges cumulative to that point. However, I'd imagine your set up is a bit more complicated because if you have a mix of different ramps I'm betting you aren't just calculating each SD range by hand.

Anyways, I hope my record keeping reaches this level one day.

[Stealth]

I really only have one graph I put any weight in (cumulative win) since the others either don't represent anything meaningful or redundant. I really like the SD ranges. I'm going to steal this idea.

This graph is developed by maintaining running, cumulative values for EV, Actual, SD and then graphing. Tracking only win/loss is just that and offers little insight into your game.

Metrics like ZScore are much more instructive, but you need to employ sampling frequency for a targeted goal for them to be meaningful.
Zscores.JPG

Since ZSores represent the distance in SD you are away from your expectation they serve to normalize the variants of the game and focus on how you are performing when compared to your expectation. When the ZScore is develop on your play continuously it begins to point to the veracity of your ongoing game. The blue line is cumulative ZScores for the data in my previous post. The yellow line is my "confidence" factor which is related to my percentage of rounds played versus the N0 for 2 standard deviations. In other words, I have a confidence that my rounds played are results are ??% of those needed to achieve N0@2SD (my long term). When looked at in conjunction with cumulative ZScore you begin to see when the veracity of the cumulative numbers take on a more definitive statistical meaning.

If you really want to dig deep you can begin a search for "abnormal" variance which I find many players obscure by just attributing big swings to normal variance. Study's of these things (and there potential causes) over time, begins to define and illuminate your AP skills and most importantly areas in need of improvement.

Good luck and PM me if I can help.

[ZeeBabar]

Just like the I-18 and such, at some point, the gain is not worth the time and effort. It's why, I believe, in the evolution of an AP, they go through the the type of details that Stealth mentions but at some point most discard the details. It can go from a job to an obsession, a workaholic.

in my opinion, your EV should be calculated based on total time spent on BJ (keeping records, reading resources, research and playing). It's easy to say I am earning $50 per hour but if the real time you are spending recording details, research and forums and such, the EV is really $10'per hour!!!

[Norm]

Been thinking about developing an app to track and display such stuff. Just have no time.

[Three]

Just like the I-18 and such, at some point, the gain is not worth the time and effort

How much effort is there in memorizing a more indices? It is the easiest thing you will ever have to do as an AP.

[Bigdaddy]

How much effort is there in memorizing a more indices? It is the easiest thing you will ever have to do as an AP.

Yes - especially if you put the indexes into groups (e.g., <0 group, +2 group, +4 group, +7 group, +10 group). You could easily capture the top 60 indexes. And the math does show that there is only an infinitesimal loss in EV due to index grouping.

[ZenMaster_Flash]

" ... infinitesimal loss in EV due to index grouping."

I cannot overstate how much I stridently disagree with this statement.

I pictured Dr. Griffin spinning in his grave, and it was not a pretty sight.

[Norm]

Griffin's work mostly dealt with single deck.

[Bigdaddy]

I cannot overstate how much I stridently disagree with this statement.

I pictured Dr. Griffin spinning in his grave, and it was not a pretty sight.

My statement is based on Arnold Snyder's analysis of this subject:

http://blackjackforumonline.com/content/hilolite.htm