I think i just found something very very interesting.
First of all i started writing an nba 2k15 explorer, which is working (just exploring the archives for now) but only for a limited number of archives like 0W which is the file that i worked on. I did not like the way that the archives are stored right from the start but i continued working on a really "workaround" way of reading the file.
Yesterday trying to test my tool with various archives to see with what its working or not, i found out that it could not find anything at all in archive 0A. I immediately opened this with my hex editor, and i found out that this very archive is the beggining of everything. Right from the start, the lack of extensions in the archives seems to be like a continous file just split in various archives for obvious reasons.
Now what i discovered is that this 0A, contains in its start, ALL the archive names together with their sizes, and right after this section, there is a section of nearly 8 mb's in size (0x860E80 in bytes, thats a HUGE amount if you are thinking with bytes) containing ALL OFFSETS FOR ALL THE MAIN ARCHIVE SECTIONS OF EACH ARCHIVE. But thats still not enough. These offsets need to start from somewhere, and thats the 0 offset in 0A file. In other words: The offset that leads to the first main archive in 0B file is equal to the length of the 0A file , the offset that leads to the first main archive of the 0C file will be equal to the length of 0A+0B and so on

That changes all plans for an explorer application, because something like this means that you simply CANNOT read any of those files separately. YOU NEED ALL THE REFERENCES in order to read them without errors. I need to find a proper way of loading all of them somehow and just load contents of the selected archive on demand. But i guess thats a good re-start
