Atlas Update - February 2020

Greetings Dragon Lords!

I know that we have been promising an update about what we are planning to do with Atlas in the first chunk of this year. We have been working with Early Birds as well as looking at many of the comments that people have been posting about some of the changes that were released alongside the launch of the Atlas Frostwreath Season.

Below are our plan for what we are committing to working on at this time:

Edit: removed draft #1 (outdated; but you can read it for historical context by checking the edit history if you’d like!)


Please discuss and ask questions in the Official Discussion Thread!

For any questions directed at employees, please make sure to tag myself & @PGMichael.

EDIT: Earlier, outdated drafts have been edited out so that’s it’s easier for everyone to find just the latest proposed changes. You can still find the historical, older drafts by reviewing the posts’ edit histories.

UPDATE - 2/5: (Draft #2)

We appreciate all of the feedback that we have received so far regarding the changes that we posted on Monday. Based on the feedback, we have the following updates:

Edit: removed draft #2 (outdated; but you can read it for historical context by checking the edit history if you’d like!)


Update - 2/10/19 (Draft #3)

We appreciate all of the continued feedback that we have received since posting our last round of updates. Based on the feedback, we have the following updates, most of which will be under a header:

Edit: removed draft #3 (outdated; but you can read it for historical context by checking the edit history if you’d like!)


Please note that we will likely be holding off on the changes to castle ownership until a later date.

1 Like

UPDATE - 2/14/20 (Draft #4)

We appreciate everyone giving us further feedback around the proposed Atlas changes. Below is our updated proposed list of changes!

Edit: removed draft #4 (outdated; but you can read it for historical context by checking the edit history if you’d like!)


Update - 2/19/20 (Draft #5)

We appreciate all of the continued feedback that we have received since posting more updates on Friday, the 14th. Based on the feedback, we have the following updates, most of which will be under a header:

  • Wave 1 Tweaks
    • Tweaked the max proposed defender revive from 110% to 100%
    • Added in increased max troops you can build/revive in one slot from 8K to 12k - LIVE NOW
    • Team Glory Scaling will remain based on your team (not your 5TA, as previously proposed)
    • To conquer, your primarch needs to have more troops than each other 5TAs troops present
  • Wave 1 changes should now be considered a “Final Draft”
  • Wave 2 is still very much up in the air and open to discussion and major revisions
    • Cut the seasonal land reset, perimeter travel and new ownership limit.
    • We’ll consider evolutions of those ideas in the future, and further discussion and revision (and after seeing where Shrines takes Atlas).
    • Other ideas, like shuffle as proposed by Kate and others, are currently being discussed but are not ready to share yet

Wave 1: Current Season (ASAP)


Remove conquer penalties - LIVE NOW

Problem: After teams spend months building up infrastructure, having a castle conquered is currently a very costly experience when 10% of their levels, which are taken off the top meaning the most expensive levels, are lost in the conquest. Castle turnover should be a less crushing experience

Solution: We’ll be reducing infrastructure destruction from 10% to 0% on conquer via a param change.

Fix Conquest Mechanics

Problem: Currently, to block a conquest, all the defender needs to do is to keep sending 500 troop count primarchs to a castle. This:

  • Does not cost anything meaningful for the defender since the troops involved are such small numbers
  • Is enough that it will prevent the attackers from hit-and-quit. 10:1 losses on 500 troops is actually quite expensive

Solution: Teams may conquer a castle if the following conditions are met:

  • There are 0 guards present, AND
  • Defending owner (not 5TA) has less than 50k troops present
  • Your primarch has more troops than each other 5TAs troops present, AND
    • Example: There are three other 5TAs present. They have 50k, 100k and 200k troops, respectively. Therefore, to conquer you would need a Primarch with >200k troops (the largest enemy 5TA present).

Glory - Player (Level) Component

Glory Bands are Too Generous

Problem: Glory bands based on numeric level difference are too generous (to the detriment of lower level players). Glory bands are too small for high level players (where a few levels difference means less than at lower levels).


  1. Change the penalty for attacking one “group” down from 30% to X%. Groups will be relatively small and granular (exact group level thresholds TBD).
  2. Group players based on the highest dragon power they could access at their current player level (this will be carefully hand-tuned into granular groups).
  3. The highest level player who flies a dragon in the battle will be used to determine the attacker’s group (same as now).
  4. The 500+ level for full glory rule goes away

Example: If the best dragons you can hatch are platinum dragons, then you can attack players who can also hatch platinum (or better) dragons for full glory. If you attack down one tier (gold), then you’ll get X% less glory. If you attack two tiers down then you’ll get even less glory (1-X)^2. And so on.

Taunter Issue

Problem: Low-level taunters can be used to dissuade attacks.

Solution: Player Level is no longer considered when attacking Taunters (unless they’re at their team’s own castle [team, not 5TA] or in an NML [bold = edit]).

Glory - Team Component

Attack Power isn’t the Best Metric + Buckets Too Narrow for Lower Ranked Teams

Problems: Team Glory Scaling currently depends on attack power, which isn’t representative of real power.

Solution: Change team glory scaling to be based on Den level.

  • Top den on the team is weighted twice as heavily as the lowest den on the team
  • 30% glory reduction for every 20 ranks higher your team is is versus the opponent (on team den score), starting with teams 20 ranks below you. (Teams within 20 ranks of your team will give full glory.)
  • Updated daily
  • Note: Level 500+ no longer bypass this multiplier (they used to give full glory no matter what)

Sticky Team Affiliation

Problem: There are teams in Atlas that are simply used for players to jump in / out of as they snipe or farm glory.

Solution: The team glory scaling factor is based on the HIGHEST ranked team you were on in the last 30 days. (So dropping to a lower ranked team will not affect your glory earnings.)


Better Glory for Successful Defenses

Problem: Defenders don’t get enough glory when they win.

Solution: If the defender wins, they’ll take at least 10% of the glory the attacker would have won (on top of any glory the defender would have ordinarily earned). This scales up to 50% of the attacker’s glory if the attacker destroys 0% of the base.

Bonus Revives for Weaker Defenders - LIVE NOW

Problem: Players permanently losing troops to much higher level players doesn’t feel great.

Solution: Increase the defender revive rate if the attacker’s glory is penalized due to a level differential (only applies if the defender is at a castle owned by their 5TA). If the attacker gets 0% glory, then the defender gets 100% revives. If the attacker gets 100% glory, then the defender gets the current revive rate. Linearly interpolate between these two cases.

Example: Let’s say I’m hitting a lower level player and my glory scaling due to player level is 70%. Let’s say the defender would normally get 60% revives. Instead, the defender will get (60% + (100% - 60%) * (100% - 70%)) = 72% revives. If my glory scaling due to player level was 10% (i.e., I’m way higher level than my target), then the defender in this example would get (60% + (100% - 60%) * (100% - 10%)) = 96% revives.

Note: Keep in mind that the revenge flag may affect the attacker’s player’s glory scaling due to player level. If a weak player attacks a strong player, and then the strong player retaliates, then the weak player will not get bonus revives!

Priority Bug Fixes & Tweaks

  • Surrender while shielded issue
    • make sure pvp bubbles aren’t included so friendly teams can still surrender castles during pvp
  • Ghost Primarch issue
  • Invader need higher levels
  • Increase the max number of troops you can build/revive in one slot to 12k (from 8k)

Wave 2: Next Season


Problem: Atlas has become stagnant. Players in mega alliances are raking in rewards and have gotten themselves in a position where they can oppress all other Atlas players.

Solution: When minor events are running in the core game, a cloud is lifted off a new island on the Atlas map. The island is visible to all Atlas players but the population of the island is sharded. Each shard is the battleground for teams that have been matched together for the duration of that week. The island is populated with a collection of shrines, that are claimed and defended similarly to Castles. While the island is active, teams on each shard compete to control the shrines. Holding a shrine garners victory points for the team. At the end of the event (called Shrine Event for now), the teams receive rewards depending on where they ranked among the teams on their shard.

More information will follow in the coming weeks about this particular feature.