Stage 3 of the Iron Bank Cycle

With Stage 3 of the Iron Bank cycle done, we sat down and crunched some numbers on just how many players and how much damage was involved in the Iron Bank cycle.

One of our key metrics for success in Stage 3 was getting newer and younger players involved in community events. AvA is a very intensive and deep play experience, which is difficult for low-level and younger players to engage in. Stage 3 of World Events was meant to be intentionally easy to be involved in, with a low barrier to entry, while also providing incentives and benefits for our mid- and high-level players.

Overall, we had six times the number of Alliances participate in Stage 3 than participated in Stage 2, the AvA portion of the World Event. This increase was driven by a huge number of smaller Alliances participating in the event. In total, more than a quarter of everyone who logged in was participating in Stage 3. While higher level characters did much of the damage, lower-level characters certainly pulled their weight; On average, characters that were level 10 or above completed three full waves of orders. This was a big improvement from AvA, where most players below level 30 don’t participate much, if at all.

We met on Wednesday to go over improvements for Stage 3 in the Braavosi cycle. Aside from the obvious - looking at total World Event health, rebalancing attacks, making tributes more exciting and expire when they should - we identified a couple major improvements that could be made. Better feedback to our players throughout the event, showing that they’re engaged in something epic in scale and scope. Individual rewards or encouragement to log in every day and participate. How we improve Stage 3 so that our players find it more exciting and interesting is our goal before Stage 3 of the Braavosi cycle.

Now on to the analysis!

 

Who contributed to unlock the World Event?

After 19 hours and 49 minutes, the World Event unlocked and players began to send out orders. A Thousand Eyes and One, Kong’s Landing, and House of Baratheon were the top three contributing Alliances who opened the World Event, with Kings of the North Worldwide, Guardians of the Fourteen Flames, Strongholds, Elyssi Dracarys, Winter is Coming Alliance, and others coming close behind. However, those Alliances only contributed small portions to the overall supply - it took the effort of thousands to unlock the World Event.

 

How much damage was done in the World Event?

(click to enlarge)

 

Looking at the above chart, damage peaked each day during US prime hours (18:00 - 02:00 UTC) each day. Some players speculated that Labor Day weekend, a major holiday in the U.S., impacted the damage being done. Damage per minute and damage per hour stayed relatively stable each day, except for Sunday, where it increased by 10% over the previous day averages. Monday morning saw the highest Damage per Minute, with the highest peak hitting 1,320,000 damage per minute.

The first world event thresholds were set to:

  • Contributions: 6,000,000 supply.
  • Part 1: 500,000,000 damage.
  • Part 2: 1,000,000,000 damage.
  • Part 3: 1,500,000,000 damage.
  • Part 4: 3,000,000,000 damage.

Total damage to complete Stage 3: 6,000,000,000

Total damage done by the end of Stage 3: 5,770,928,346 (2,770,928,346 in Part 4)

Difficulty increased for each part, but damage per minute stayed relatively high. This is because the difficulty of the challenge affects the amount of additional damage done in the World Event. A failure does “regular” damage, which is in a range of possible amounts (for example, 500 to 600 damage), while success does double the regular damage, and a critical success does five times the regular damage. Unlike questing or adventuring, losing a World Event order doesn’t mean you get nothing - it means your Sworn Sword will do the normal amount of damage and receive a wound.

This helps explain why damage per player per hour was steady during the event:

 

(click to enlarge)

 

With little difference in effort between sending out one order versus sending out five orders, most players sent out a full wave of five orders every time they logged in. We consciously made orders not take Command Points, and give Sworn Sword and Character XP. By the end of the event, our players had earned 91,531,500 XP for their characters and 320,360,250 Sworn Sword XP. That’s enough to level 300 Sworn Swords from level 1 to level 127 - the highest they can go!

 

How did we decide on the damage and contribution amounts?

It’s a good time to discuss how we balance massive systems like the World Event. The damage formula for Stage 3 is based on the same damage that you see in Boss Challenges and Alliance Challenges. By taking representative challenges in a meaningful sample size (n=10,000), we were able to control for level, age, and number of Sworn Swords. We then found the maximum possible damage the entire community could do, and the minimum we expected them to do. The balance part of the question was answered by the design team looking at participation rates versus desirability of rewards, and extending that to the World Event.

Contribution was balanced by looking at the total quantity of non-permanent items across the community, and assigning contribution values appropriately. The threshold was adjusted using a mixture of desirability of each type of item, and the future utility of the AvA spoils, versus the desire to open the Wold Event as quickly as possible.

Like most online game design, there is a lot of extrapolation and hypothesizing before a system goes live. Now that we have data on Stage 3, we can tweak the balance for the next stage. Our goal is to make Stage 3 fully completable, but not without significant effort. Why you should have significant effort before receiving rewards is a great discussion we should have in another blog post, or over drinks at King Richard’s Faire this month or PAX East.

 

How many players participated in Stage 3?

(click to enlarge)

For the lenth of the event, we had around 30% of our total active players sending out orders. The chart above graphs hourly participation rate for players. The peaks and valleys between days are clearly visible, as is the increase in participation on the last day. It is clear that Labor Day weekend had little to no effect on participation rates. GoTA has a very international playerbase, with about half of our players being international, but most of those players are either in Europe or the Americas. That means our lowest population times are between 0100 UTC and 0900 UTC.

We split participants into four groups:

  • Those who lightly participated (41% of participants, sending out up to 15 orders).
  • Those who participated (17% of participants, sending out 15 to 45 orders).
  • Those who heavily participated (34% of participants, sending out 45 to 255 orders).
  • Those who were deeply committed to the event (8% of participants, sending out 255+ orders).

For the next world event, we want to make Stage 3 more interesting and engaging than this first attempt. A core question for the next time is, how do we encourage our players who only sent out a few waves of orders to send them out each day, or multiple times a day? How do we make the World Event even more engaging, so they want to?

 

Who did the most and least damage?

The top 10 damage dealing players were, in no particular order:

  • Stunning Steve
  • Cahling Buhlsheet
  • Andryn Penrose
  • Dhyan Mahyhaer
  • Gregor Dane
  • Elyana Payne
  • Tregor Pendragon
  • Maximus Un Vicesimus
  • Romules Wolf
  • Knicknack Paddy Whack

We also had 450 players who did, on average, 25 damage or less per order. The clear winner for least damage done in the World Event was Macha Of The Morrigan, a new player who did an average of 1.1 damage per order, for 10 total orders.

From our Alliances, Kong’s landing did an incredible 3.5%, or 200,000,000 damage, of the total World Event damage. Their high-level, highly optimized players were able to do consistently high amounts of damage, helping push the World Event bar further than any other single Alliance.

For those of you not in Alliances, don’t think that your contributions were unneeded; Non-Alliance players were the fourth-greatest source of damage in Stage 3 of this World Event.

Kyra Rambton’s Sworn Sword Amerei Greenfield holds the dubious honor of being wounded the most of out any Sworn Sword, suffering an amazing 79 wounds from Stage 3 of the World Event.

 

And what about those Braavosi ships?

As of this blog post, we have 1,710 peerless Braavosi ships floating in the Narrow Sea, and 2,900 legendary Braavosi Longships coasting beside them. So far, the Peerless Braavosi Warship is the clear favorite for our players.

 

Moving Forward

Some of our most ardent players always ask us whether game systems work the way we intended. Certainly we want them to work well, but what’s most important is that our players have fun, or at least find the changes interesting and intriguing. We can and will make Stage 3 more fun and interesting for the next cycle, and the cycle after that. Online games let us constantly improve. It’s because of players like you that we have the opportunity to reflect on honest, thoughtful feedback, and continue to give you new experiences in Game of Thrones Ascent. If there’s one emotion that sums up developing for an online game, it’s surprise - surprise that our community comes up with new and novel ways to play our game, and take it in directions we would never have thought of.

Don’t forget that we’re continuing to push out quality of life fixes for GoTA! Next week sees the Sworn Sword selector not closing when adding Sworn Swords to garrison, a Remove All button for garrisons, and the fix to the phantom wound bug, in addition to new Alliance challenges and the next cycle of Alliance vs. Alliance combat.

Thanks for reading!