If we have two brackets, for instance bracket 3 and 4:
In theory the winner of the lower bracket could get the same reward as the second last in the better bracket.
And the runner-up in the lower bracket could get the same reward as the loser in the better bracket.
Then it would be no advantage in losing your way downward.
But, I guess that litle extra reward for winning a bracket is too small to care about.
Great job @Sirrian . Appreciate some of the changes, and looking forward to trying this out soon! 
[quote=“Mekkalyn, post:30, topic:21806, full:true”]
I don’t understand what the point of these brackets are. It doesn’t change anything. You’re still basically competing for the top prize which will never be achievable for the average guild.[/quote]
In the previous thread, somebody mentioned that a higher bracket’s lowest reward shouldn’t be smaller than the highest reward from the lower bracket, since that would encourage intentional losing to get into a winnable bracket. Based on that philosophy, a global ranking/rewards would end up being equivalent to specific bracketed rewards which take that into consideration, but it would be much easier to implement. I think @Annaerith describes it best:
Easy to understand description, and sums up what I was hoping for, “narrower” match-making (rewards were just incidental).
I feel the opposite, and am really happy with this match-making change. I was hoping for a slightly wider bracket size (somewhere between 20-50), but my main hope was to play guilds I’m familiar with and around the same level as mine. The familiarity is because it could create some fun rivalries and discussions here on the forum, having a war against guys you know. Competitively, I’d much rather play top teams that have decent and challenging defenses, rather than cake-walk over weak guilds. Some of the “cheese” you mentioned is really a balancing issue and global problem with the whole game regardless of mode. So fixing those and “broadening the meta” will give top teams more options and let us play against different troops and strategies, far preferable to fake defense diversity by lop-sided pairings. And I think “random match-making” would also be harder to get rid of in the future and we’d be locked into it for a long time (even if some of if the meta issues are fixed).
I think people are assuming everyone needs to fully level their sentinels. I see this as the top 20 or so may try and fully level their sentinels, the next 80 may do to level 4, and then everyone else will be anywhere from 0 to 3. Being down in the 1000 bracket, I don’t expect to need huge sentinel buffs. But as my guild starts to move up, I would think about investing a couple gems each week.
Sentinels wont be leveled equally. Everyone will max out magic sentinel and other sentinels probably as high as level 2. Maybe some will max out attack too, but not many.
That is not an entirely correct assumption though.
Just choosing some numbers to illustrate how Lower brackets winner can have higher reward than the Higher brackets loser. This would in no way encourage intentional losing, as there would be nothing to gain.
Higher: Lower:
500 420
490 410
. .
. .
420 340
410 330
Regarding the brackets… there’s more to consider here than just the math though… the solution has to tick 3 boxes:
So, in the example above, we see a range of 100 Gems for a bracket.
But what happens if we have 100 brackets? Or 200 brackets? What about 1000 brackets?
Obviously in the final case there, the top bracket would have to be earning 100,000 Gems, but even in the smallest case, the top bracket would be earning 10,000 Gems.
And also, as you move down from bracket to bracket, there are going to be thousands if not tens of thousands of players earning WAY too many gems each week, which is just not feasible for the economy.
So, obviously the solution is to base each bracket on smaller (MUCH SMALLER) ranges of Gems, with a sharp dropoff at the top so as not to flood the economy with too many Gems… but it STILL doesn’t scale unfortunately, and as more players are added, and more brackets are added, either we have to dynamically alter the ranges, or we have to create new ranges at the top with bigger rewards, OR add more ranges to the lower end with little-to-no rewards.
In fact, when you do the math on that, you’ll very quickly find that the amounts of Gems given per bracket vary SO LITTLE (from top to bottom), that it’s simply not worth even having a difference for the most part… you may as well make the reward simply moving upwards into higher brackets.
We actually had 3 maths guys spending hours on this last week, and it’s simply not as trivial as a contrived example makes it seem… certainly not something to commit to in 4 days, and then be shoe-horned into an existing framework.
@Sirrian what about release date? Seems that we can expect it in the next few days? But i may be wrong.
#theydidthemath
This makes perfect sense, especially given the recent influx of new players!
Personally I like all the changes and I’m really looking forward to it.
But I would echo @DonBoba and ask for an idea of the release date. Pretty please with a pony on top!
I totally agree. That’s why I said in my post a few minutes earlier:
And by the way:
I had hoped it was understandable that I chose easy numbers just to show the principle (that in fact, higher reward for winner in a lower bracket than the loser in a higher bracket, does NOT automatically imply that someone want to lose intentionally.)
Whiskeyjack? Who are you oh handsome stranger?
(Sorry couldnt resist)
I am praying that the conclusion ultimately drawn will be that the underlying problem is a half-baked implementation skull matching and trait proc’ing
But if that is the conclusion drawn (which it might not be), I think to address it it would take a lot more work to than the areas being discussed openly.
You are such a brilliant communicator man and the comment I’m replying to is pure gold, thank you.
In future the above quote is going to be my go-to reply as I “peace-out” of ugly, heated arguments about game mechanics 
Why does it have to be all gems why can’t it be a few gems, glory, some keys and maybe a random epic or legendary troop?
The main issue I could see with the bracket system is it will discourage higher bracket guilds from recruiting lower leveled players. Right now higher guilds can allow people who are level 100ish with untraited legends into their guild as long as that player is active. Any players in the top 5 brackets or so, particularly the top 3, will need to have guild members with fully traited level 20s, most troops obtained, and many kingdoms at 5+ gold stars.
I think this is a huge assumption that should never be made. If it were my guild, I would absolutely want to be in the lower bracket if it means less effort for greater reward.
I suspect most guilds will have a relatively narrow range where they can be competitive, so a system like that would encourage all guilds to play at the bottom of their range in order to min-max effort vs reward. Then you end up with a system where it’s not actually fun or rewarding b/c there’s little incentive to really try.
That was a comment to me. I tried to only comment on that some logic that was used, was not as black-white as it seemed in other comments.
I’m sorry and sad that my post is looked upon as “heated and ugly”. That was never my meaning. I only wanted to comment on some effect I had seen, that I hadn’t seen mentioned before.
It does not feel great to be misunderstood so badly by a developer (taking my principle example as a real suggestions for actual numbers, when I only wanted to show a principle.)
It feels even worse to be the example of “heated and ugly”.
I obviously have no clue how my words are understood by others in a forum.
So based on this, I just want to assure you all that I will not bother you in here any more. The door already hit me.
Hey @Lun, yes, I understand perfectly… my comment about contrived examples wasn’t meant to be judgey at all - sorry if it seemed that way! I was just kind of carrying on from where you left off.
Please know that I’m not talking about you @Lun!
To be honest I didn’t see your comment, I only scanned through this thread and watched for @Sirrian 's posts.
I will be more careful in future, so that when I comment I don’t inadvertently hurt feelings or make it look like I am attacking somebody. Please accept my apologies.
One thing that bothered me and I didn’t see mentioned:
What if someone who did some or all battles of the week leaves an hour before weekly reset? Do we lose the week, or do his points stay with the guild?
Similarly, what happens when someone leaves the guild mid week, and another joins in his place? Is the new player eligible to participate in that week’s war? and if so, does he get to fight all daily wars, or just the ones since he joins?
Please dont skip my question, and respond to the next one 