tar wrote:Top 3 wrote:
spacewar : 15 : tar : Space Wars
: 15 : WRX2 : recordable since
: 13 : German Krol : m35b6
our korean friends are peculiar
The phenomenon you are observing is well documented, but until recent events was unexplainable.
A WRX2 defector has, in return for his own island near Cabo San Lucas, revealed to Western investigators their first details of the organization and structure of the WRX2 network.
The phenomenon at hand is a result of the method in which the inp retrieval spider traverses the massive WRX2 network of computers. According to the defector, the network is set up in a hierarchical structure according the the strength of the individual player at that location, based on previous MARP submissions, and results of certain Asian scoreboards and competitions. The entire structure can be envisioned as a virtual pyramid, with the average players at the lowest level (there are no below-average players in the WRX2 league - players performing at a lower level are removed through methods not suitable for public discourse). These average players make up the largest skill level, those slightly better are slightly fewer in number and occupy the next "higher" level in the "pyramid". The structure thusly terminates with the legendary Kang-Kyu, alone at the "top".
The simplest traversal method for the spider would of course be to start at the top and work down, but this would place too great a band-width burden on the sole top computer. Therefore, a slightly more sophisticated algorithm is used, whereby the spider starts at the lowest level, searching it completely for a score beating the MARP submission which automatically triggered the spider's operation, before moving on to the next "higher" level. Only if none of the other several thousand computers in the network contains a recording which would regain the MARP high score does the spider ever reach the "top" computer.
Since the spider stops as soon as it has found such a recording, and because the computers searched first are the lesser caliber players, often it will be seen that the MARP submission is only slightly bested, or even tied. Of course if
that score is bested by another MARP player, then the entire process is restarted, searching for an inp that would beat
it.