Feel free delve into any questions.
* Filter results by eBay category.
* Outlier elimination such as excluding "broken" items (e.g. listings containing commonly used words such as "spares or repair", "cracked screen", "damaged").
* Country specific currency (e.g. for UK searches show the results in GBP, £).
* Predict a fair market value for a given search. Useful when trying to sell items second hand as an "arm's-length transaction".
I must believe there's any easy way to eliminate some "outliers" using mathematics, but I can't recall the function(s) to do so.
The median is one good way, as you already have. You can also use the interquartile mean: http://en.wikipedia.org/wiki/Interquartile_mean
Perhaps I need to filter it within 1 or 1.5 stdevs. Will experiment with this.
However, sometimes you can easily see there are two clusters of results. Not sure how to mathematically determine this. Any ideas?
Just pointing out that there is a decent sized market for this type of data.
Locksley, keep up the good work. You need to clean up the results. Removing the unrelated items is desirable. This has a good monetizing potential.
Yeah, I've come across Pricenomics but found their data inconsistent and the sample size too small with ebay results. They're doing good work nonetheless and their front end rapes mine haha. I ain't much of a designer.
This article explains it better than I ever will.
I'm running 8 unicorn workers per dyno at the moment and that's still within the 512mb memory limit because I'm not using ActiveRecord.
EDIT - also it looks like you're submitting it to ebay badly, the results for "win&win" are same as for "win".
Can you give a little details on how it works? Is it scraped data? etc.
I know this is targeted at a person who wants to sell items on ebay, but I am a buyer and I find this still quite useful. Not sure how up to date the data is but the graphs are an amazing feature! Makes so much sense.
I actually originally intended this for the buyer! I wanted an iphone for myself, so I scraped the 'completed listings' page on ebay, and converted it to csv to build a histogram in excel.
The data is up to the minute. But I cache the queries in order the save computational resources. Each cache expires within 15 minutes.
Initially the data was scraped, but then I moved to the API which improved performance by 200 - 300%.
This is a really killer piece of kit though. I'm very impressed. It's something I think many of us have idly imagined when struggling to determine a good starting price for a new eBay auction.
I have a suggestion: Perhaps you could build some kind of e-mail alert system that lets me know when there's an e-bay "Buy Now" item available within a user specified price range. You could also let me know if there an item I want is still within a certain price range hours before the bid closes.
I'm guess a stats nerd so I like to see that stuff haha.
I like your feature suggestion, just put it in my product backlog.
They are simple and obvious, the only feedback I would give is that I would think (not being a huge ebay person) the current charts might actually be changed up a bit.
I would be more interested in seeing the clustering of prices, or something like how long the listing has ran without a bid (indicating more likely desperate takers and the like).
If the current charts do influence the price in some way that can be used that is advantageous to the buyer, then I would rescind my comment.
Sent this site to a few friends who don't read HN, they are highly appreciative.
Your problem was the reason why I filtered out results below $100. There's just too much junk below that range like "Fender Bass HELLO KITTY STRAP". Eventually I'll have to think of a way to algorithmically exclude those results.
Thanks for the suggestion.
NEED MOAR ENGINEERS
My servers surprisingly seemed to handle it ok with one minor crash.
Otherwise, it looks like a great site!
In the meantime it's probably just as effective to put the condition as a query, i.e "iPhone 5 used"
To do it for all items would require collection of lots of data in my own database though. Will have a think about this.
- "dexter season 6" returns only listings where season 1-6 are sold setting an average price at ~$140
- "simpsons season 9" returns listings where multiple seasons including season 9 are sold setting the an average price at ~$150
In each case the 'raw data' had more than one season in it
Would it be possible to show the full link when hovering over a truncated link ("lin...")?