Monday 5 January 2015

The more coin tosses we do, the closer we get to 50/50

So in the previous posts, we looked at how we can generate coin tosses in excel using a random number generator.

One of the things we noted is that although coin tosses have a 50/50 chance of being heads or tails, it doesn't work out that way in reality for small samples.  The following shows the distribution for one of our excel simulations (39 coin tosses)



This is what happens as we increase the number of tosses, here is 1000 tosses



And 5000 tosses




10,000 tosses


And 20,000 tosses



As we can see, the more tosses we do the closer we get to 50%, but ultimately it's still random, so we can still get a decent difference.

This tells us a few things:

  • Small samples aren't sufficient to predict trends or check randomness
  • The more samples we do the more likely we will get something that fits the probability
  • Random doesn't look random
  • Our coin tosses over a large number of samples fits the probability
In future posts we will check some of these simulations for betting patterns such as martingale, check probabilities of roulette tables etc etc

And of course, you can check all of this yourself using the techniques described in the earlier post.  You don't need to take my word for it.

No comments:

Post a Comment