Andrew Hill
ahill@cc.gatech.edu
CS 8803D - Advanced Internet Application Design
Spring 2006
This proposal describes an application which will be used to place a bid on an eBay auction in the final seconds of the auction. The structure of the paper is as follows: the motivation and objectives section will describe in further detail the application’s design and purpose, the related work section will describe current technologies addressing the issue at hand, the proposed work section will discern the project from current technologies already implemented, the plan of action will describe a schedule of events which will take place in order to assure a successful project, and the evaluation and testing method section will outline the procedures which will be implemented in order to verify the correctness and functionality of the final product. Finally, the bibliography section will contain the resources used in researching the technologies discussed in the related work section.
One of the most frustrating aspects of the online auction community as a bidder is being outbid in the final seconds of an auction leaving one with no time to place a counter-bid. Another typical situation is when an auction a user would like to win is set to end while at an inconvenient time for a user to watch for these last second bids such as in the middle of the night or when he/she is at work. A final related problem is that of connection speed or type such as a user on a narrowband connection or who loses his/her connection at an inopportune time such as at the auction’s end.
The objective of this project is to increase an eBay buyer’s chances of winning an auction by allowing him/her to place his/her bid during the final moments of the auction automatically. This will solve the problem described above concerning the time of the auction’s end. In addition to an automatic system to perform the task, another main benefit of my application will be user security and ease of use. The proposed application will reside on the user’s personal computer which will yield the highest level of security as a tradeoff for the problem of connection speed. However, it will be possible for the application to account for available bandwidth and latency in order to place the bid at the optimal moment desired by the user. This leaves only the problem of loss of connection which will be traded off for security.
Bid timing is an exceptionally important in successful online bidding. Approximately 75 percent of final bids are submitted in the final 3 percent of an auction’s duration [1]. This technique is commonly referred to as sniping and is considered legal by eBay’s terms of service, “Any bid placed before the auction ends is "legal" on eBay” [2]. There are many services such as [3] and [4] which offer users a way in which to snipe auctions. The main problem with such a service is its lack of security. Conversely, the main advantage of such a system is that its connection is centralized and, therefore, the likelihood that it will lose connection to eBay is greatly decreased. Additionally, it is not necessary that your computer be running or connected to the internet to complete the task [5].
Alternatively, eBay offers what is known as proxy bidding. When you place a bid on an eBay auction, the minimum possible bid is placed visible to other users, however, the user’s maximum bid is stored in order to place automatic bids based on other user’s bids [2]. The obvious problem with proxy bidding is the possibility of an emotional bidding war where users bid higher as a result of excitement [5]. These bidding wars take time and are, therefore, made impossible by sniping which takes place in the last moments of an auction.
This project will implement an application in C# which will gather data from a user-defined eBay auction page and react accordingly. The program will take as input an eBay username and password, an eBay auction number, a time variable (most likely seconds) within which the bid will be placed, and the user’s maximum bid. The program will then retrieve the data specified by the auction number and continually update itself until the auction reaches the specified threshold at which point it will place the user’s maximum bid assuming that it is greater than the current bid.
C# was chosen for speed and convenience. The application needs to be very time efficient, and as a result needs to be written in a language which can execute rapidly such as C, C++, or C#. The input was chosen to specifically adhere to users’ needs. The eBay username, password, auction number, and maximum bid are needed for obvious reasons. However, the time threshold will be implemented such that a user can specify his/her own risk factor. This risk factor is such that the less time allowed, the higher the risk that the bid will not be placed in time but the higher the possibility that the auction will be won.
This project unlike centralized web services will guarantee user security in that it will not require users to place their private eBay information such as username and password in a database. Additionally, it will give the user control over all aspects of the sniping process, including the ability to remove an auction from being sniped in the last moments of the auction before the snipe for any reason. Moreover, the user will have full control over how close to the end of an auction the sniped bid will be placed.
The system architecture can be seen in the figure below. As can be seen from the figure, data is sent from auction pages to the servers which is then broadcast over the internet. The application then gathers the necessary information such as remaining auction time and current price. The program then continually gathers data from the page, calculating whether or not to place the bid and when the time is correct, broadcasts the bid (assuming it is greater than the current price) along with the username and password.
This project will can be considered successful if the application developed is able to connect to an eBay auction page, download real-time data, compare that data to user input and react accordingly. If the application is able to successfully place an eBay bid according to user input, the application will be successful regardless of whether the auction is won. If the application attempts to place a bid but is denied as a result of auction ending which can be attributed by the threshold being set too low by the user, the project should not be considered unsuccessful.
| February 8-15 | Project Proposal |
| February 16-21 | Revise Proposal Given Instructor's Feedback |
| February 22-March 1 | Connect to eBay Auction Page |
| March 2-16 | Retrieve Data from eBay Auction Page |
| March 17-31 | Commit eBay Bid |
| April 1-12 | Compile Findings into Research Paper |
| April 13-20 | Finalize Research Paper and Prepare Powerpoint Presentation |
| April 20-25 | Prepare Project Demonstration |
The application will be tested in phases consistent with the plan of action. Connection to the eBay auction page will be tested via pulling some simple, verifiable data such as auction number to verify connection. Retrieval of eBay data will be tested by comparison of data retrieved from the auction page with the actual auction page itself. Committing an eBay bid will be verified via the My eBay page which tracks all bids for a particular user. A successful bid will be verifiable through eBay and all intermediate steps will be verifiable against the actual auction page itself.