NOTE: This thread is not intended to suggest solutions for the unique problems presented by overloading classic servers for virtual vanilla, but I did intend to explore the wiggle room available within #nochanges.
Please respond with clarifications: knowledge, links, blue posts.
Lopsided servers would benefit from a weekly ranking that is based on bracket fill quantities of the higher populated side if those standings were filled top-down. For instance, if horde has 10k and alliance has 1k, bracket 1 would give both sides the fill quantity of the horde side: 10,000 * 0.003 = 30 that get 13,000 rank points.
Q. Is this done, or is there leeway in #nochanges to allow for this?
Vanilla players were able to check a listing of weekly results for their server.
Q. If this is available, then what is the URL? If not, then was this a purposeful denial by the classic team?
It would help to have a slash command that gives more details for each week, such as bracket, standing within a bracket, standing within the whole population counted for ranking, and one’s percentile of the maximum HKP earned by the top earner that week.
Q. Would this be possible, and would it violate #nochanges?
Are you suggesting on #1 that there would be an equal number of players at the top ranks on horde and alliance regardless of faction size? And it gets clamped upward rather than down? So a server with 50 alliance and 50,000 horde all 50 allies would hit rank 14 if they met other requirements. There’s no way this is how it is done
@Harlff - I condone extreme samples to test for weakness. You should be realistically extreme though. Realistically, that server would have shut down beforehand. If it didn’t, who would be that guy who berates devs for bestowing every credit available (with honor) to those 50 brave souls? Lastly, let us consider what would occur in your sample case to resolve this with cold hard math.
50 * 0.003 = 0.15; truncated to 0
This result would never happen. There is always a minimum of 1 at the top, and no minimum at the bottom, except that the bottom completes the whole. This does suggest that the most extreme case sets precedent for top-down fill of brackets. Now then, the question is: was there ever; in action, in code, in contingency plans; the use of the highest side population for honor count for the calculation of bracket fill quantities. If so, then that precedent was also set.
I’m almost positive that rankings are done relative to the faction population.
So in a super hypothetical situation of a server with only 50 alliance on it, you still need to outrank 49 other people for Rank 14 (although this is obviously much easier than outranking 50,000 Horde).
We used to have a Web Page that gave each servers honor ranks, ll the players that were ranked, there current position and who was a head of them and by how much!
Im pretty sure the honor system tab is borked. I got 899 first week with a rank of 2.8k or so. I got 600 honor last week and got 280 or so. I went down in rank though so Im thinking it’s tabulating properly but not displaying right.
I think whats happening is you are killing less honorable targets! I am killing lvl 60s at lvl 56 so honor is higher for me. Also consider if you have a large Horde Population you have much more people competing for ranks.
Thats what I’m saying. I was ranked close to 3000 the first week for participation at 900 honor. The second week I raked in a whopping 600 honor and saw my weekly position clock in at 290 or so. My scout rank deterriated a bit though.
Which is, I presume, a commonly held assumption. Think it through though, as the original devs must have. Population dominance would bestow more higher rank slots. If everyone played an alt to get minimum HKs, then it would double the number of people allowed in each rank. Do you see? No only does the higher population side gain a numbers advantage, it also gains the advantage of more people progressing ranks, and that dominance allows for more time to manipulate the system by adding alts to the count. Meanwhile the underdogs are fighting tooth and nail to survive on their main.
So, being literate in math solutions myself, I will pose that the right solution should have been to “normalize” both sides and fit each into congruently shaped rank stacks. That was not done. Now we are left with a predictable problem, so I made this thread to begin the discussion about what is possible within #nochanges, inviting the involvement of the community by attempting to expose facts that have been mysteriously made a mystery. An unvanilla thing to do.
Unless I am reading this wrong, it’s confirmed here.
Hmmm. Well, the ranking system was extremely imperfect and that’s no secret. I’m unsure what you mean about facts being mysterious, and that being an unvanilla thing. Which facts? There are many things we don’t and won’t ever know about the development of the original game, simply because it’s lost to time by now. And there were undocumented changes here and there even back then. Mystery is a very vanilla thing.