After the recent relaunch of the now cleared-out GalCivIII-Metaverse one of my more recent games (still) yields a negative score when winning it by conquest, resulting in a corrupt record in the Snathi-table (result of zero, but a 4billion-score in the details) - clearly an (signed -> unsigned?) FP-issue with respect to the concrete calculation ... I wouldn't mind one bad entry but it is not ironed out by other flawless endings of the game (clearly 4e9 > anything else) and even worse, no other Snathi-win of mine (with the same difficulty-level) is registered by the Metaverse and my overall stats seem to be deformed as well by this corrupt entries in my Snathi-table.

Actually I really like to play with the Snathis - so is there any chance to delete the corrupt data record?  
That'd be great - over the years I only had a few such anomalies, most of my games ended fairly reasonable.
(could be a great extension for the Metaverse to allow authenticated users to delete entries by themselves or just select the ending score/entry to want to dominate/overrule the other endings btw)

Topic-Addendum:
As a computer scientist and programmer myself I have always been fascinated by these scoring issues and the mystery behind it - surely not easy to balance that (not impossible though, algorithmically - not numerically - speaking, but I've always tried to keep out of this debate, as intriguing as it is - in the end it's the programmers' decision how to achieve comparability and rank the players).
However, from where I sit the relaunch of the Metaverse could have been a great opportunity to overhaul the calculation of the GalCivIII-scores (esp. those hilariously high integers close to 2^32 - incredibly out of proportion) or at least insert some database-constrains/triggers that could avoid such issues as mine at hand. There were a decade of concrete data from thousands of games available to be analyzed in order to find the perfect scoring-procedure - or did it indeed all get lost like my many games?