Wednesday, June 09, 2004

On Casino Gambling

(Note: I composed the majority of this e-mail during a non-applicable product demo in Reno on Thursday, May 27, with additional updated edits done on Friday May 28, and now final edits on Monday, May 31 before my final 'send'. It was also composed entirely on my Pocket PC, so it may be a bit disjointed.)

I've found myself attending a product demo that doesn't apply to my work, so it's a good time for me to log my observations of gambling in Reno of the last few days.

I'd heard it beforehand, and it seems obvious, but the most important thing to make a gambling excursion sucessful has to be good money management. "Sucessful" to me, from an economic perspective, is finishing up with the cost (funds lost) being less than or equal to the entertainment value.

Foremost in this management is kowing my limits. Before starting this trip, I'd set a maximum that I was willing to lose per day, and that would be the only cash I'd have on-hand for the day. "Borrowing" from the next day is straight out.

Secondly, every time I've left a slot machine, the craps table, or the roulette table, I've entered my total wagers and results into Pocket Excel. As a result, I can tell you that my final results for each day were

Sun -$4.48
Mon -$5.75
Tues +$48.99
Weds -$37.8
Thurs +$139.25
Fri -$70
Sun +$0

So, after all is said and done, I'm up about $70 for the week. The big win on Tuesday was a $50 payoff on a $1 slot wager. The loss on Wednesday was primarily due to $40 in wagers on slot machines that didn't pay off. Thursday I discovered a low limit single-zero roulette table and had a streak of good luck. At this moment [written Thursday], I've got one more round at that roulette table I'd like to try, $70 out-of-pocket max, so I know I'll be leaving here ahead. [Result was a loss of $70, followed by a round of coming out flat even at an indian casino on Sunday].

Third, having a grasp of probabilities has been imensely helpful. Before coming down here, I'd researched as much as I could, and come up with a way to play craps that gives the house only a 0.5% edge. Only blackjack (played perfectly) or baccarat gives better odds. Baccarat has higher table minimums than I'm willing to risk, and I know myself well enough that I know I won't end up playing blackjack perfectly. More on why I didn't continue playing craps is in point 4 below.

Wednesday morning I sat at the roulette table with someone who was apparently very bad at money management, not to mention the concept of probabilities. She had no concept of why not to place $1 on each of the 38 numbers for a $36 payoff, had the classic gamblers' fallacy that some colors were "due", and remarked that she'd lost $800 at a casino down the road, and had stopped by her house on the way here to get her credit card.

(Side thought: are people who believe in an active God more likely to ignore probabilities, thinking that some cosmic force is at work in at least some miniscule way?)

The urge to assign irrational meaning to odds is strong. It takes a conscious effort to avoid thinking that one number on the wheel is more likely to come up, or that a slot machine is "due" for a payout.

My fourth point: flexibility and knowing my personal limitations. My original plan was to focus on craps most of the time. Once I got to the craps table, I found that I was mentally unable to follow the stategy I'd come up with: bet don't pass, then put odds on don't pass and bet don't come each following roll, and when my don't come point is 4 or 10, put don't come odds. Yes, it was complicated, and I found myself down $16 on Monday morning as a result of mistakes I made implementing this strategy. Solution: find something easier.

Combining point 3 (probabilities) and point 4 (flexibility), I found myself considering roulette. If you feel like running the numbers yourself, there's really only two wagers I considered: even bets (odds 18:38, pays 1:1) or straight numbers (odds 1:38, pays 1:35). I ran a bunch of different scenarios and wagering strategies a few thousand times in Excel, and as my prior research showed, any cumulative strategy ultimately lost, either because the odds were against it, or because I 'd have to take ludicrous chances with my bankroll (like betting $128 to chase a $2 bet).

Now is a good time to bring up my fifth point, which I touched on in my introductory paragraph: treat losses as the cost of the entertainment of gambling, and consider how much you're willing to spend on that entertainment. Particularly on some of the penny video slots, I've intentionally strung out the game on a $1 bankroll to 20 minutes. Only in the heyday of my college pinball days could I make a game dollar last as long.

This brings me to my sixth and final strategy, and the most difficult: quit while you're even or ahead. Slot machines in Reno have a payout of about 95% to 98%. At first glance, one would think that on average, most people walk away close to even. The machines, however, are psychologically designed to entice the player to re-bet their winnings -- winnings are added to the machine's credits, and not dispensed as coins (or worse, a negotiable voucher) until the player cashes out. If you keep pumping your winnings back into a 95% machine, that 5% loss will chew through your bankroll pretty quick. My strategy here is to keep track of the number of credits that I'd won, and when CreditsOnMachine = CreditsWon, cash out.

I guess I've got one more slot machine strategy that I hadn't heard anywhere, and one that seems counterintuitive to the mathematically challenged: play machines with low payoffs. I've seen machines that pay maximums of 1:200 ($10 on a nickel bet), and machines that pay maximums of 1:50,000. Because I know that the probability of winng the BigPrize is close to zero, then it seems obvious to me that to keep the ~97% payout rate, the lower maximum machines would pay out smaller amounts more often.

At the roulette table, the time to cash out is as soon as you've won a payoff. Unfortunately, you don't know which payoff that should be. My $140 win on Thursday came after quite a few 1:35 payoffs, although I was up as high as $180 at one point. My $70 loss on Friday was the result of chasing after a second 1:35 payoff that never came.

So, in conclusion, on the leisure side, it's been a fun week of doing geeky math analysis and applying them to the real world. It's also been a very educational week professionally (InfoPath is bad-ass). As little time as I'd actually spent gambling, I can't imagine how anyone could spend more than a few hours here (much less make a vacation out of it) without losing a lot of money.

No comments: