Scratch Off Game Strategies?

First, all lotteries are guaranteed money loser over the long run, medium run and even over reasonable short runs, that much is a given but it does appear that unlike the lucky number astrology stuff of the draw lotteries, there’s some actual possibilities for strategizing with scratch off games so I was wondering what strategies would work best.

Unlike the odds of a draw lottery, the odds of scratch off games change as tickets are sold. Lottery authorities often publish the number of tickets printed, the number of prizes in each category printed and the number of prizes in each category claimed to date on a daily basis, Texas does so in the form of a CSV file for download. They obviously don’t have a way of collecting the number of sales to date as the tickets are bearer instruments and not sold through a computer terminal but a reasonable estimate of the number of tickets sold and hence the number of tickets remaining by using the total number of prizes claimed versus the total number of prizes printed as the ratio of tickets sold. This means that scratch off games can be timed, with tickets purchased when the odds are in your favor or at least less against you. But what do you decide upon, expectations? (That’s when the probability of each prize level is multiplied with the value of the prize and then summed up over all the prizes and is a common Engineering technique for estimating returns on investments). The problem with expectations is that they require a lot of statistically independent chances before they average out if the probabilities are low hence you would literally have to play millions of dollars before the winnings play out as expected. Do you look for games with more prizes? more low value prizes? more high value prizes? underclaimed high value prizes? Probabilities of winning at least one prize of a certain value or higher?

Using the ratio of prizes claimed to prizes printed to estimate the number of tickets sold would be able to detect if given prize categories had been under claimed or over claimed when compared to the mean and the higher value prizes would show the most dramatic variation in odds. For example, the "Super Set For Life" game in Texas is currently 89% sold based on the ratio of prizes claimed versus prizes printed but only 3 out the 4 top prizes have been claimed which equates to 75% of that prize category. Another interesting aspect of that game is that if you took the full value of the top prize which is $500,000 a year for 15 years or 7.5 million, then the expectations show a $1.01 return on $1.00 invested. Of course, the prize is actually an annuity probably worth only 63% of the total advertised value which equates with a $0.85 return on the dollar but this is much better then the original $0.68 on the dollar expectation when the game was first printed. This would seem to indicate that if you could corral all the remaining unsold tickets together and just buy them one at a time until you’ve won the big prize hopefully by about the half way point, you might make money. Of course, you would have to buy 724,059 tickets before being 95% certain of obtaining that last 7.5 million dollar top prize but you would have a 50/50 chance at 381,084 tickets which would be closer to the median of what you would have to buy. Of course, the top prize alone would not meet the 7.6 million dollar median that you would expect to require to obtain the top prize but together with all the smaller prizes, perhaps. Mind you, your scratching arm would be very very tired.

On the Internet, there seems to be a lot of people espousing attempting to exploit manufacturing processes such as purchasing tickets at the end of the roll or changing rolls once a certain amount of prizes has been won. The former seems somewhat suspect since it’s nothing for a computer to randomize the prize placement throughout a typical roll of 50 or 25 tickets at the point of printing but the latter is interesting since there is a guarantee of minimum prize amount per roll that’s published to assure retailers they are not getting rolls completely devoid of prizes and it’s likely that this would be achieved by evenly dividing low tiered prizes between the rolls and then randomizing their positions within the roll. Having said that, since I have scanned images of my winning tickets, I can see that my winning tickets have all been in the first quarter or the last quarter of a roll and none of them has been in the middle half but then I can hardly afford enough tickets to have a statistically relevant sample. Maybe if you have sufficient records to determine where in the roll your winning tickets were, you could email them to me so that we could get a reasonably relevant sample (I’d only be interested in Texas games for the time being).

So what strategies do you think might marginally improve your odds of winning? Aside from not playing at all.
Duh, I said these were astronomical odds and improving them marginally still leaves them astronomical. I just wanted to hear some strategies that people use. It’s somewhat simple minded to just be saying that lotteries are bad odds, most gambling is and that’s a given.

FIFTY PERCENT!!!

That’s the odds you’re trying to "marginally improve".

Scratch tickets, and all other lottery games start with a 50% take for the lottery companies. In other words, for every $100,000 lottery companies take in from sales, they only pay out around $50,000 to winning tickets!

To put this into perspective, the WORST casino game you can play has around a 10 to 15% advantage over the player. Lottery = 50%!!!

Even if you COULD "marginally improve" your odds, you STILL wouldn’t be able to overcome that great of an advantage. Let’s say you had an incredible system that could give you a 10% edge over all other ’scratchers’. You’d still be trying to beat a 40% advantage!

Simply put, ANY form of lottery is the worst bet you can ever make!