eBay investigations part 2...

So I've put the updated eBay feedback script here on my other blog Phill's Tech Stuff, and the first lot of results are in...

For the first 40 user ids that the script has given me data for, the %feedback and $feedback results have been close:

• average %feedback = 98.94
• average $feedback = 99.23
  

These initial results are actually contrary to my original thoughts that there would be people selling cheap items more honestly than expensive items! But we'll see what results we have when there is more data in. So far the user with the greatest difference between these variables has %feedback of 80 and $feedback of 86.55.

I've made this script up to generate a long list of user ids and I'm just going through them one by one. I'll update this once I have a significant number of them done, say around 1000, or if I find someone scamming that can prove my original hypothesis.

Where does Spam come from?

I've wondered for a while why I get Spam. There are a few hypotheses:

• I'm the admin of some websites that receive thousands of hits per month
  
• One of my addresses is comprised of my first initial and my surname, both of which are very common
• In my time on the net I've signed up to many email lists, web sites and forums. I've also had to give my email address to download software and do other internet activities

There are of course many other ways you can get Spam, and it is probably more that I have done all of these things rather than just one.

I'm going to do an experiment to test out the three spam-getting methods above. I'll get a domain, run a mail server on it and create some email accounts. Some email addresses will be cryptic ones that you'd never guess like akjalfdslkjasdlkjf324@mydomain.com, while others will be constructed from a list of baby names. I'll leave the accounts open and receiving mail for a few months and report back on my results.

eBay investigations

I've posted up to my other blog a script that gives the total $ amount of positive, negative and neutral feedback for a particular user. For a while I've had the hypothesis that there may be people rorting eBay by selling many cheap items honestly and a very few expensive items dishonestly, resulting in them being able to maintain a high feedback rating but getting a lot of dishonest money. For example:

• I sell 99 items for $1 each and get positive feedback on all of them. I then sell 1 item for $100 but never deliver it, and get negative feedback. Despite only delivering on $99 out of $199 of items I would end up with a 99% feedback rating!

Does this sort of deception occur on eBay? I'm going to find out. I'll search through 100 eBay usernames and provide two variables as results:

• %feedback, which is the percentage of positive feedback for a user
• $feedback, which is the percentage of dollars of positive feedback out of all dollars a user has transactions for

For the example a few paragraphs above, I would have a %feedback of 99 and a $feedback of 49.7

Watch this space!