Rick, I doubt that there has been any research on this topic, other than it being easy to test almost any other service and watch as 3 second snipes go through time after time.
My question -- and I have yet to receive an answer that makes sense to me --- is:
If the auctionsniper system is constantly testing ebay response times and compensating for those response times, that compensation should result in the bid being placed at the correct time, not at 6 seconds, correct? if a user has set a snipe time of 3 seconds, and if the auctionsniper system tests the ebay response time in real time and identifies a lag of 3 seconds, the auctionsniper system should place the bid at 6 seconds, which results in ebay entering the bid at 3 seconds, which is exactly what the user requested. Instead, auctionsniper appears to be routinely entering bids early no matter what time of day or night it is, no matter what the ebay response time is, which results in an overcompensation of 3 seconds, which is eons in ebay sniper time. Before I started using auctionsniper several years ago I used other services and applications that would test ebay response time just second prior to the bid and then would place the bid with appropriate compensation to result in a bid at the time set by the user. I switched to auctionsniper because i liked the browser plugin. If auctiosniper's answer is that auctionsniper cannot accurately determine ebay response time for each bid and thus will add 3 seconds to the user's bid, then many users will be looking at other services. I could understand an occasional early bid, or a +/- 1 second average performance, but consistent +3 second performance is nuts and is not what your users expect. Apologies for the typos in my earlier message. not proofreading these.