Interesting. But if that it was you were doing until now, then how do you explain the bracket movement of Hoguns and Unrepentant for the past 3 weeks? (How are brackets decided (bug?) - #75 by Maxx)
Because that could never have happened if a GW points score modifier was the deciding factor in GW bracket movement!
Oops - my bad - I mistyped that… I’ll go back check my notes and correct it shortly
Already made the video for it before noticing this, rip. lol
It’s a matter of getting the modifiers correct really. Too little, and it skews one way and guilds will incorrectly drop down, too much and guilds will unfairly shoot upwards (to be fair, it’s a horribly complex little formula with more going on that what I’ve simplified it down to - apologies for that)
Anyway, we tinkered with the modifiers (and a couple of other weightings) for 3 weeks now, until we figured there’s just a better way to do it than the current method.
There’s nothing wrong with the modifier approach per se… figuring out the “value” of points and standardizing them is really where the core of the problem lies in our current solution.
So Hoguns and Unrepentant being 5th and 6th in bracket 1 but being demoted was wrong? we should have stayed in bracket 1?
Of course you should have stayed in Bracket 1. We (Marthos Guardians) had fewer points than you did (finished #8) and we did stay in bracket 1.
There is no reasonable score modifier that can explain that discrepancy. More factors were involved in that than a simple GW points score modifier.
probably in that formula there is the “quality” of the guild included.
just curious, were UR and Hoguns lower level than all the others in bracket one?
Was it wrong?
In a mechanical sense, no… The system was doing what we told it and functioning as intended. In hundreds of other cases it performed some pretty clever gymnastics to get things to exactly where they should be. Of course, we won’t hear about those because they magically worked… but we could see it working well.
In a did-it-meet-player-expectations sense, yes… You guys weren’t happy about the result, you let us know, and we’re modifying our mechanics accordingly.
Indeed… there’s more there than points. As I explained. We were trying to evaluate the value of “points” based on guilds, their members, their opponent guilds, their opponent guild members, etc… then standardizing those values.
Nothing nefarious going on, I promise… 
So number of likes given to your posts wasn’t taken into account? Are you sure? 
Can we know the corrected rules for Bracket 1 & 2? It seems to me that for B1, rank 8-9 should go to B2 and rank 10 to B3… can you confirm?
It’s…beautiful 
“Less is more” is my favorite motto and it certainly applies here.
However, I do have to sympathize with new guilds starting in very low brackets (I read there are around 500 currently). Can we have a general estimation how long it will take these guilds to climb the brackets in the new system? Assuming they are active and strong.
Currently, if they’re established with some decent members, they might land in bracket 400. From there it would approximately 4 years to get to bracket #1.
I think 1 year is more reasonable for that… so we’ve got some work to do with assisting overachievers in lower brackets I think.
FIxed! I’m not in the office at the moment, so I don’t have access to my notes, but I think I’ve got that right now.
Lower brackets could be decided purely by points, GW prizes are too small at the level for people to try to game the system.
It all sounds very logical, I do have one question however.
Let’s say 30 people with good fully traited cards and 5+ star lvl 10 kingdoms decide to get together and start their own guild, fully oriented on GW. They might even take over a “dead” guild, that already had all statues at max level.
Do they then start at the bottom of the pool (bracket 500), and does it then take them 400 weeks to move up to the top 100 bracket?
Or is there some code in place, for special situations like this, that can sky rocket a guild up a lot of spots?
For example, let’s say they do start at the bottom, but win their first week with a score of 700k points, where all the other guilds in the bracket (and probably in the 200 brackets above it as well) only hit a week score of not even 100k.
Seeing as how the rewards are the same below a certain rank, how about using the system you mentioned only for the top 10 ranks, and then either rank the remaining guilds by last weeks score and divide them over all brackets?
Would it be an option to reduce the overall number of brackets by increasing the bracket size for lower brackets? Something like +50% size for each bracket down, just to toss in an arbitrary number. Rewards are essentially the same within lower brackets, so this won’t chance much. It would allow dedicated guilds to rise to the top faster and make brackets feel more relevant. The demotion/promotion span would obviously also need to adjust accordingly.
It’s definitely fun stuff to think about… I would love to see your statistics on points scored per bracket, people playing in lower brackets etc etc…
Meaning I’de love to see them, I don’t expect you to show them to us
And I should be working on the new version of HiLo Video Poker, but now I’m thinking about how I would tackle your problem instead… lol
Looking at the rewards, this is what I would probably do…
For the record, I’m just thinking out loud here, because it’s fun stuff.
The top 100 (top 10 brackets) use the system as proposed above, as it sounds good.
Below the top 100, you have already defined bracket groups: 100-199, 200-399, 400-499. 1000+. You could use those, or make groups of 100. And then rank all the guilds in those group by score, and the top x% (for example 10%) move up one group, the bottom x% move down one group.
The tricky point of course is the flow from “low” groups to bracket 10.
PS, if you wanted to make it “more effort” to get into the top 10, you could have the top 15 or 20 use the proposed system, and then below that use the groups. That would prolong the time of being able to hit bracket 1 by 5 to 10 weeks…
But I honestly don’t think you should. A clean new guild now has the disadvantage of low bonus multiplier, so it would have to work hard. Let’s say you use groups of 100, and the new guild starts in group 500-600.
It would take them probably 4 weeks to move up to group 100-200.
Since the top 10% can’t move up anymore, but more like the top 5%, it might already be harder to get to the top 100, but lets say they do, 5 weeks total.
Now they have to win in each bracket, to move up 2 spots, so 4 more weeks to hit bracket 3 (9 weeks total).
Now finish in the top for 2 weeks, and we have 11 weeks, it would take you to hit bracket 1.
If that’s even possible, as those high brackets are filled with guilds that have (close to) 50% bonus score for troops, where as in 11 weeks, I doubt you have more then 30% bonus.
So best case scenario it would take a new guild 3 months time to get to the top.
Seems fine to me… 3 months is a long time 
Well, first off, thanks for all your hard work there trying to make it better.
For me this bracket system is just not working at all tho. Would be much more simple and easier and prolly less coding too if it was simple good old ranking system where there is no jumping from bracket to bracket, where you would move up and down ranks depending how did your guild do last week score based. And to prevent exploiting, abuse and other unexpected movements between ranks you set up some modifiers like someone already pointed out and you guys are very familiar too ofc. And even if there is obsession with brackets and those will stay then it must be changed. It’s just silly if getting up to top brackets take years, even months is too much. System should not block great GW guilds moving up. Some new guild might have all members old GoW players or mby even some old guild that just have not participated yet. For example if at this moment my guild would end up in bracket 453 and you can move up 2 brackets per week at max I would not even bother to do GW to be honest. Just setting up some formula that prevent lets say rank 6 guild to swap places with rank 378 guild as obviously lower rank guild had easier fights and got more point cos of that would do the trick. And that way low rank guilds that just started could actually get where they belong soon enough. And anyways why you should drop 2 brackets if you are in bracket one? If there was decent point system then mby you would be next week like rank 11 or 14 or whatever but not like rank 20-30 cos of mby your guild wasnt as good as top 9 this week. Its kind a like in real life, if you doing sports in premier league and lose a game you dont get put to some amateur league … you have to work your way down to amateur league not gonna get there straight away.
