Wolverine: Actual results calculation in CV suite

Wolverine · 06-19-2005, 08:31 PM

Norm or others,
I just finished playing 604 hands of simulated blackjack on CVBJ. My correct play rating fell at 98% (index play) (and they were corrected, so actual playing percentage was 100%) and betting was perfect thanks to the fact that it can be changed if there is an error. The game was standard MGM DD S17 rules: 2D, DAS, SPA1, 60% pen, S17. Play all. The betting ramp was: 1 unit for TC 1 or below, 2 units at TC=2, 3 units at TC=3, 4 units at TC=4 and 6 units at TC>=5. {yes, I know the game doesn't SCORE very well--but I am not a professional. I'm trying to keep my game EVEN with them and enjoy a free vacation, eat well, and not get creamed. If I win a little money, great. If I lose a little, fine. If I lose bunches, I need to find another hobby.}

I happened to use units of $100 for this action. I managed to lose $4,850 in those 604 hands. Using an old sim of a similar (but not exact) situation (betting ramp was more aggressive in the sim vs. actual play), the win rate per hand was 2.777 (and with a variable of 170 hands/hour [one on one play] a results of $/hr of 277.72). The Std Dev per hand was 3.263 and the Std Dev per hour (again, 170 hands) was 42.543.

When I punch in these numbers to the CVCX "actual results" software window, I get a chance of results of 0%.

Now, I'm mathematically challenged enough. But when I found out that the results I just witnessed with my own eyes (playing perfectly) are absolutely not possible, I get concerned. Obviously, I need help getting the information entered correctly to find out how bad these results really are. And I'm sure that will spawn a whole new set of questions.

I'm off to run an IDENTICAL sim to what was just played to get ChapterX data to enter. But I'd like some expert input before I continue to spin my wheels. Thanks.

Wolverine · 06-19-2005, 09:18 PM

Okay, ran a quick 250M hand sim to get numbers. I think I understand better now what is going on.

I got a win rate of $153.14/hour. The Std Dev is reported as 24.40 per hour but that is in UNITS, not $. Since the win rate is based upon a $100 unit, the Std dev must be multiplied by $100 to equal $2440. {Alternatively, the win rate could be reduced to 1.5314 units per hour, achieving the same result.}

Now, given that data, placed into the "actual results" calculator, it gives the possibility of this outcome to be 16.78% or worse. Or about a 1 in 6 chance? In other words, pretty likely.

Am I on the right track here? Using basic statistics then, my expected range of results for this 6 hours of play would be from losing about $6,400 to winning $8,240. That was figured with an expected win rate of $153.14 x 6 hours = $918.84. The $2440 Std dev x 3 (to cover the most likely results) = $7,320. So $920-$7,320 = -$6,400 and $920 + $7,320 = $8,240.

By the way, the N0 was 25,381. That figures to be 16 hours of play per trip for 16 trips. Even for a modest player, that seems like the "long run" could be achieved in just a few years (4 or 5)!