Jump to content


Mathematics Contributor
  • Content Count

  • Joined

  • Last visited

About seriych

  • Rank
    SerB Worshipper
  • Birthday 05/18/1986

Profile Information

  • Gender
  • Location
  • Interests
    science, PC, photo
  • Server

Recent Profile Visitors

1,534 profile views
  1. 1. Only WG can use assist. WG does not provide sufficient information on assist for third-party developers. In order to take assist into account it is necessary to know the number of battles that have been carried out since the beginning of recording this indicator. But WG does not provide this information. 2. Serious changes in the formula would entail a change in the rating name. Thus, it is the same wn8, which has always been.
  2. @Gryphon_ We also want to switch from wn8a to wn8d. As far as I understand, initially this could not be done due to the lack of necessary data in the WG API. However, the data has long and stably existed in the API in sufficient volume, which allows to switch to the wn8d method. All active (last 3 month) players with 10000+ account battles and 50+ vehicle battles.
  3. How about to increase number of battles on tanks? 50 battles is too small to correctly evaluate winrate or defense points. I understand that you had a small database, so you had to make such a restriction. But if we have full database, does it make sense to increase number of battles on tanks? ~200 battles for example
  4. Main factors are lack of time and lack of people well versed in statistics. But it is possible to make a recent rating that does not require almost any increase in a database. It is enough to apply a method similar to MoE (in Russian).
  5. No, we store data on all active players in our own database.
  6. If you are experiencing difficulties in calculating the expected values, maybe we can automate calculation on the XVM server? If you provide an algorithm. This will allow to update the values daily and quickly add new tanks. PS. Found a topic:
  7. XVM uses old v28 values for E50 and M46, and new v29 values for all other tanks.
  8. OK. I hope WN9 will solve problems of tanks such as E50 and padding damage by light tanks.
  9. Thus WN9 is not completely ready, and you do not need help from XVM now.
  10. I do not quite understand the discussion in WN9 topic. Do you need help with the collection of statistics? Or algorithm is completely ready and it remains only to realize? In any case you need to leave your inquiry here: http://www.koreanrandom.com/forum/forum/55-
  11. just for information: XVM uses v29 table with v28 values for M46 and E50.
  12. And what's the point? Why show a false ranking of all players in a few weeks? This will only show your helplessness to liars (because you do not come up with the normal method), and to other players just mess rankings.
  13. Really? It is not bad on April 1, but as usual - is a delirium. Problems with the underestimation of some tanks? Look for errors in the method. WN8 has been good because it was based on pure mathematics without subjective speculation. If you refuse this principle (especially so clearly) people will lose trust in you. Why are the values increased so much? Why does not increase values for other tanks with the same problems? Why fair players have to suffer because of a handful of liars?
  • Create New...