On randomness of winning streaks, and barcode tracking

The Drunkard’s Walk

Audiobook: part 2, chapter 3, ending at 44:35

discussion of randomness and performance and how we misperceive the cumulative effects of success


the difference between the odds of one specific individual in one fixed period of time hitting a winning streak based on purely random, coin toss decisions,


the odds of some ONE of many many players, over a BROAD stretch of time, having SOME period in which they have a winning streak based on coin toss decision making.

these are INCREDIBLY different factors:

Odds of the first happening:  Extremely, impossibly low.
Odds of the second happening:  Surprisingly high.


For TRADE, if we’re interested in, at the end of the show, showing charts and graphs of audience performance — and cumulative charts and graphs, not just that particular audience but ALL audiences in the run of the show to that point (and then maybe also charting that particular audience against the overall, cumulative performance of all audiences) — something like this might be interesting to include if we can figure out a way in which we can track and determine what it is to have a winning streak:

  • Does this mean always picking or buying tickets to shows that trend upward?
  • People who buy tickets early for subsequent rounds, future market, that end up with higher value whether or not you resell them?

How do we track this?

If there is a centralized place for selling tickets, even if we have multiple brokers we still have a place — the desk, the post — where the ticket has to be issued, then can we encode on the ticket what the purchase/sale price was?

(i.e. buying R3-R6 tickets during T2 at the (then) lower price, and USING THEM FOR ADMISSION to see the shows during Rounds 3 through 6 — when buying those tickets during T3-T6 would be higher than the T2 price — could be considered a “win” just as if the person had decided to sell those tickets at the higher prices instead.  They saved money, they invested in the show in advance, and they made a “profit” just in their use, since they were paying less for the future round shows than people who did not choose to invest in advance but instead bought on a per round, T(n) for R(n), basis.)

If this info can be encoded on the ticket, we probably want to find a way to encode it in a way that is not visible to everyone — for fear that if a buyer who is about to buy a ticket for $3 can see that the seller actually bought it for $1.25, they might have a negative feeling about their purchase.

Barcode system?

Could we have a barcode system set up so that information can be encoded on the ticket when purchased (and the ticket is printed?)  Then when it is used (for admission) or sold, the barcode is scanned and data is sent to a central computer:

  • original purchase price
  • selling price if sold
    • or that it was utilized to gain admission to the show
  • AND marking it against the current prices at that point for ticket (so as to know if it was a “win” or not?)

There would then need to be a system where, if a ticket is sold, the old one is scanned in (to retain the data), new information is added about the current transaction, and a new ticket (whose barcode contains the additional information) is printed and given to the buyer.

This could be a way of tracking overall performance of entire audience — and, we could potentially barcode PEOPLE as well so that we could look not just at overall data, but track individual performance in an anonymous manner.

Maybe they each get a trader ID that they have to scan every time there is a transaction?

Then that info (Trader ID number) also gets encoded in ticket barcode, so it has information about the buyer and seller for each transaction.  Doing this, we can actually gather data for how the entire audience, at an individual level, perform over the course of the show.  If we have 100 people, we can actually track 100 individual people’s:

  • purchases
  • sales
  • choices of what to see
  • whether they actually watch a performance in a given round, or sit one or more out
  • how heavily they buy/sell versus just buying for personal use
  • look for trends in decision making about when they switch from one performance to another, marked against
    • price increases/decreases
    • runs on tickets (i.e. maybe the show they wanted to see sold out before they could get a ticket)
    • whether and when they start playing more complex elements (futures, shorting etc)

As I said before, of primary concern would of course be anonymity — we could track people as individual traders, but maintain their anonymity by never attaching their real world identity to their trading data.  All numbers.

One Response to “On randomness of winning streaks, and barcode tracking”
Check out what others are saying...
  1. […] look at my voice note about barcodes […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: