Monday 2 November 2009

Measuring Utility

How useful it is to measure utility! (pun intended).

After having spent at least 40 attempts to recreate the graphs from David Hardman's post, I finally made it! Hooray!
































As you can read, the first one is for Certainty equivalence, and the second one for Probability equivalence.

As you can see, both graphs are fairly similar, only with slight differences. On the vertical scale Y you will find the utility values given, whilst on the X scale there are 2 different elements represented.
On the top graph (certainty equivalence), the x axis represents the economical prize (for sure) I'd settle down between obtaining that prize or a chance between £1000 or £0 (with different values for the winnable amounts).
On the bottom graph, probability equivalence, the represented value is the percentage in which I would be indifferent between winning a certain amount for sure or having that said percentage between winning £1000 or £0.

Reading the graphs themselves, you can see that I'm not really a gambler, prefering more than 50% of what I can win before I settle down for the cash prize (I'd rather take £600 than being able to win £1000 with a 50% chance). This is consistent with the fact that I'd chose the same amount in cash (£600) unless I was given a high chance (60%) of winning £1000.

In general, I prefer to have a bit more cash than 50% chance, and at the same time, I still prefer the cash than to get the "big prize", unless a high (HIGH) percentage (I would not risk £800 unless I was given a 70% chance of getting £1000!)

I hope this wasn't "too long; didn't read" for anyone reading this! If you have any questions/comments, post a comment (duh!) in the comment field!

1 comment:

  1. Allowing for the fact that these two graphs are only based on a few data points, they are both somewhat different from a typical utility curve. In what way are they different and what does this suggest about your attitude to risk?

    ReplyDelete