Site-based server selection

One or more non-transitory computer-readable media storing one or more sequences of instructions which, when executed by one or more computing devices, cause:

• selecting, at a client computer, a first bitrate file;
• receiving first data from a first copy of the first bitrate file that is streamed from a first server computer at a first site having a plurality of server computers;
• determining a first throughput estimate for the first site based, at least in part, on a first throughput of the first data from the first copy of the first bitrate file that is streamed from the first server computer;
• associating the first throughput estimate with the first site;
• receiving second data from a second copy of the first bitrate file that is streamed from a second server computer at a second site;
• determining a second throughput estimate for the second site based, at least in part, on a second throughput of the second data from the second copy of the first bitrate file that is streamed from the second server computer;
• associating the second throughput estimate with the second site;
• selecting a second bitrate file, wherein the second bitrate file is different than the first bitrate file;
• determining that a first copy of the second bitrate file is stored at the second server computer at the second site and a second copy of the second bitrate file is stored at a third server computer at the first site, but not the first server computer;
• requesting third data from the first copy of the second bitrate file from the second server computer, without attempting to download data from the second copy of the second bitrate file from the third server computer, based, at least in part, on a comparison between the first throughput estimate associated with the first site and the second throughput estimate associated with the second site.

A client computer for improving experience quality of digitally distributed content and configured to:

• receive a first list that enumerates a first plurality of server computers that each host a copy of a first bitrate file;
• receive data indicating that a first computer of the first plurality of server computers is associated with a first site;
• receive data indicating that a second computer of the first plurality of server computers is associated with a second site;
• receive, in response to a first request, a first portion of a first copy of the first bitrate file from the first computer and determine a first throughput;
• receive, in response to a second request, a second portion of a second copy of the first bitrate file from the second computer and determine a second throughput;
• associate the first throughput with the first site and the second throughput with the second site;
• receive a second list that enumerates a second plurality of server computers that each host a copy of a second bitrate file, wherein the second bitrate file is different than the first bitrate file, and the second plurality of server computers includes the second computer and a third computer, but not the first computer;
• receive data indicating that the third computer is associated with the first site;
• determine, without requesting additional data from the first computer or the third computer, that the second throughput is greater than the first throughput, and in response, send a request to the second computer for data from a first copy of the second bitrate file, without sending a request to the third computer for a portion of a second copy of the second bitrate file.