0
As you know, we're continually striving to improve the accuracy of our items database. An issue that has come up recently is some random spikes in the price of an item. Good examples of items that jumped around enough to cause problems were:
Boiled Egg
Visor
and other Tiki Tack items...
Their price would basically multiply many fold due to a couple bad outlier prices being submitted to our database.
With our past outlier detection system, if more than one outlier entered our system, we could easily identify and remove it. However, during additional tests using two outliers in a data set, our outlier detection failed to notice that outliers even appeared in the data. We have now created a secondary system to deal with this.
On another note, the other day, we hit a record of 94 users concurrently submitting data to the database in a one hour period. Good job, automation rocks!
We'll continue to monitor the integrity of our prices and improve the system further if we need to. Thanks everyone for supporting the database thus far.
That can be fixed one of two different ways though so we can look at that when you're back from your larking about at uni. 
