Simulating Video Poker - Part 2

This is a small piece of analysis I've done, now I have Office 7, it makes life a little easier.

I simulated 10,000 players each play 100 sessions of Video Poker. Each session was at $1 Video Poker (e.g. $5 per hand), and last 120 hands. I estimate it takes about 15 minutes to play 120 hands, so this represents 25 hours of play per player... the sort of long weekend of a Vegas VP addict, or a 3-4 visits by a tourist like me.

I then looked at the final bank roll of each player, and made a frequency graph of the results. This shows what the likely results of playing 25 hours of VP on each type of machine, and the chances of ending up at each range of bankrolls.

I intend to repeat this with Bankroll Based goals.

First, lets compare Bonus Poker - 8/5 versus 6/5

Machine Type	EV	Maxmium Bankroll	Minimum Bankroll	Average Final Bankroll	Standard Deviation	Percentage of players in Profit
8/5 Bonus Poker	0.9917	$18,005	-$5,690	-$468	$2,521	32%
6/5 Bonus Poker	0.9687	$15,890	-$6,775	-$1,866	$2,468	22%

Effectively, you are giving up $14 for every 120 hands played by playing 6/5 Bonus... and also giving up variance as well. That's $56 per hour, so hopefully the casino rental you are paying is worth it.

Here's a graph of the results. The x-axis is a players final bank roll, and the y-axis the number of players from 10,000 with that bankroll (+/- $100). You should note that over $10,000 there 26 players (0.26%) in the 85 who end up over $10k, while 6 players in 65 end up over $10k.



I'll work on presenting the results for other machines this week... there's football on right now, and this is taking too long to present :)

Nice analysis. I go a simpler route--you hit a full house every 90 hands, so once every 90 hands, you flush a $10 bill down the toilet, by getting paid $30 instead of $40.

Machine Type	EV	Maxmium Bankroll	Minimum Bankroll	Average Final Bankroll	Standard Deviation	Percentage of players in Profit	Percentage of player over $10k in profit
9/6 Jacks or Better/dat]	0.9840	$15,195	-$5,690	-$972	$2,363	26%	0.14%
Super Double Bonus	0.9969	$17,385	-$9,750	-$226	$3,384	43%	0.7%
Deuces Wild - Full Pay	1.0076	$18,635	-$6,055	$491	$2,805	51%	0.6%

Super Double Bonus has the widest variance of all the games, and is the game that trends most like a normal bell curve. The rest all are skewed to the left... a large number of players ending up around a mode value that's less than the average, with a long tail of player moving towards the positive side.

I just got the modal values for these simulations :

Poker	JoB	85BP	65BP	Deuces	SuperDouble
Mean	-$972	-$468	-$1,866	$491	-$226
Mode	-$2,000	-$1,400	-$2,800	-$800	-$600
Median	-$1670	-$1,155	-$2520	$43	-$580

This can explain a fair amount, for me at least. While the averages for the games are all where you expect them, for the games with lower variance, the importance of the high paying hands is much more on how close you get to the EV, and a lot of players will be batting less than that EV, and playing around the modal/median values (the left shift peaks in the graph above and below). This may result in disillusionment and annoyance with the AP grind. I can also see now the importance of hitting those big hands for any medium term play.



Note the up tick and either end of the graph are players go outside the range I've plotted (the percentage making over 10,000 and the percentage losing more than $6,500).

Yep, JOB is the most frustrating game in the world. You almost always lose, unless you hit a royal. The key is too lose less while waiting for the royal, and the grind sucks. I recently had the opportunity to play 15,000 hands of JOB, and did badly. The store also had 99.93% Super Aces. I'm almost sure I would have broke even at Super Aces, even with the increased variance, since I hit aces twice during that session.

Quote: teddys

Yep, JOB is the most frustrating game in the world. You almost always lose, unless you hit a royal. The key is too lose less while waiting for the royal, and the grind sucks. I recently had the opportunity to play 15,000 hands of JOB, and did badly. The store also had 99.93% Super Aces. I'm almost sure I would have broke even at Super Aces, even with the increased variance, since I hit aces twice during that session.

Holy crap. I don't play VP, but 15,000 hands sounds like quite a long time of playing any casino game.

Good way to get rated, I take it? I know a person shouldn't play for comps, but sitting down and playing video poker for hours and hours on end sounds like one of the more steady ways to do it.

I was playing with a partner, and had a 25x points multiplier so was playing with a +0.79% edge. I still hit the hard downward slope of the bell curve. Took about 6 hours.