Wednesday 27 June 2007

An Interesting Gamble

The following ad caught my eye - "Bid to Buy an Apartment for less than 5% of it's value!". It turns out that this is a sort of a lottery (although it appears that it probably is classified as a property auction, as there doesn't appear to be a lottery permit number on the site) with the entry (bid) fee costing $5 and the prize value being a Queensland apartment "valued" at $350,000. You have to pay $5 for each "bid" you enter in a reverse auction for the property, with a maximum of 20 bids allowed per bidder each week. The winning bid will be the lowest bid (below $17,000) that is unique at the time the auction closes - either a set date (although the site doesn't display that anywhere I could find) or when the required number of bids has been received (also that's also not listed anywhere that I could see).

In lotteries the total ticket sales generally add up to around twice the total prize pool value, so I'd guess the required value of bid fees is around $750,000. This would mean 150,000 bids. As bid amounts include dollars and cents, this means that a maximum bid of $1,500.00 would allow each bid to be unique, in which case the person who bid $0.01 would win the prize. In reality most people will be trying to bid as low as possible, so low bids are unlikely to be unique when the auction closes. High bids have more chance of being unique, but are unlikely to win.

Aside from the question of what the actual value of the prize may be, it's interesting to think of what the "best" strategy would be to employ in this contest. When you place you first bid you are emailed if it is
a) not unique
b) unique, but not the lowest unique bid
c) currently the lowest unique bid

In case a) you've obviously not won, but a lot of contestants would then proceed to place a higher bid (the email will tell you if the current winning unique bid is higher or lower than your bid). This will reduce the number of bids placed at low prices once the current "winning" bid has passed that price point.

In case b) you have to wait and see if all the lower, unique bids become "not unique" by the closing time of the auction.

In case c) you have to wait and see if anyone else bids this same amount before the auction closes.

The promoter states that if there are no unique bids when the auction closes the lowest price with two bids will be used to determine the winner - the person who placed the first bid at that price. However, with a price limit of $17,000 and $5 fee per 1c bid, this would mean that $8.5 worth of bid fees had been paid for a $350,000 prize! The promoter must be dreaming that this ends up being the case, but I doubt that that many entries will occur. In which case you'd get the best information and have the best chance of winning if you placed your bids close to the end of the auction. If I was going to enter this contest I'd contact the promoter to find out what the "stated date" for the auction to end actually is.

Assuming that the value of all bid fees ends up being around twice the prize value (like most lotteries), we can deduce that a bid of around $1,500.00 has a good chance of being unique. SO one strategy would be to enter an odd bid value around this number, say $1,384.73, and if you get notified that this is unique but no the winning bid you could enter a second bid for half this amount - say, $692.36. Assuming that all your bids came back as unique (but not the current "winning" bid) you would repeat this process until you bit a price that was not unique. If the current "winning" bid was less that this, you could then use of the remainder of the 20 bids you're allowed to place each week in small steps below this price, hoping to hit a unique value that had a good chance of winning the auction.

An alternate strategy would be to make a first bid at a much lower price point, say $236.57. The email notification would then tell you if the current winning bid is higher or lower than this price. If you made this initial bid too low there is a high probability that the bid won't be unique.

Although this contest is a bit more interesting than your run of the mill lottery, I don't think I'll enter this first auction. (I don't gamble much anyhow - and recently losing $3,000 on forex day trading has used up all my "play" money for the time being!). On the one hand no-one will know much about how many entries this auction will eventually attract, so you could be lucky and the contest doesn't attract many entries, boosting your chance of winning. On the other hand, in future auctions you could estimate from the previous contests winning bids what amount the winning bid is likely to be (say, $300-$400 or whatever), and roughly how many entries there were. If the number of contestants in future auctions doesn't increase, this extra info would help you target your bids at the prices most likely to win.

I may check back in a couple of months to see what the winning bids were for the first couple of auctions - assuming this contest is a success and the promoter stays in business.

ps. I doubt that the emails sent out to bidders will be encrypted. So a hacker that knew how to intercept emails could monitor the autoresponder emails for bids that have been entered, and therefore track what the current "winning" bid is, and ignore bid values above this price that have already been entered. Combine this info with the date the auction ends and the hacker would be in a good position to enter 20 bids that have a very high chance of winning, just before the contest closes...

I don't know how to do this sort of email hacking, but it's another reason I probably won't enter this contest - I'd hate to find out that the eventual winner was some pimply computer nerd ;)

Enough Wealth

No comments: