Slide 1

Slide 1 text

Charugin Vadim, Lead Game Designer Pixonic, 16.10.2018 How We Developed and Tested our Hypotheses War Robots Events: 01 T X E N / 24

Slide 2

Slide 2 text

02 / 24 Events in our understanding In-game events Time limited Unique experience T X E N

Slide 3

Slide 3 text

03 T X E N / 24 Pros Paying Users DAU Increase player base Increase player engagement Increase purchase count Opportunity for experiments

Slide 4

Slide 4 text

04 T X E N / 24 Cons Recovery Addiction and fatigue Strict release schedules Unexpected problems and bugs

Slide 5

Slide 5 text

05 T X E N / 24 Why we created our first event Understanding the need Enough content in game Wanted to try new things Store featuring 1. 2. 3. 4. First event - Halloween 2016

Slide 6

Slide 6 text

Number of fights per day per person 06 T X E N / 24 Start time and duration Same time on all platforms Not very long Start before weekend and finish after weekend 03 Dec 10 Dec 17 Dec 24 Dec 31 Dec 07 Jan 14 Jan 21 Jan 28 Jan 1. 2. 3. Optimal time for WR event: 17-18 days starting on Thursday

Slide 7

Slide 7 text

07 T X E N / 24 New store icon New visuals Additional tasks New paint job system Increase maximum clan size New game mode A/B tests What we did in our events

Slide 8

Slide 8 text

08 T X E N / 24 New store icon Shows that something unique is going on Use only for large events

Slide 9

Slide 9 text

09 T X E N / 24 New visuals for maps Causes negative reaction if left for a long time More maps = longer to develop and test Shows that an event is on Provokes discussion in the community

Slide 10

Slide 10 text

10 T X E N / 24 New visuals for main screen and UI Minimal map changes, focus on hangar and pop-ups Main difficulty is maintaining good performance

Slide 11

Slide 11 text

11 T X E N / 24 New paint job system ~6% of all currency spent in-game (more on iOS, less on Android) Main objective: testing players’ interest In demand by a small part of audience

Slide 12

Slide 12 text

12 T X E N / 24 Increase maximum clan size Planned as a feature for competitive clans. Makes it easier for a clan to get higher position in ranking. More players = more activity Used by a small number of players

Slide 13

Slide 13 text

Number of fights per day per person 13 T X E N / 24 Additional tasks: 1st iteration Questline for the whole event Tasks are divided into parts with main reward at the end of each part Evaluated complexity using similar metrics from past events Start End

Slide 14

Slide 14 text

14 T X E N / 24 Additional tasks: conclusions % of players, who completed all tasks was as expected, but the funnel was uneven Redistribution of rewards within the chain helped High paying group complete more tasks on average Received/completed quests

Slide 15

Slide 15 text

15 T X E N / 24 Additional tasks: 2nd iteration Players who started too late had no chance to complete everything Tasks for a limited time instead of questline Smaller rewards, easier tasks Increased the number of people who received rewards up to 20 times

Slide 16

Slide 16 text

16 T X E N / 24 New game modes Adds gameplay variety Allows players to find “their own” mode Excellent additional activity for the event

Slide 17

Slide 17 text

Unranked mode with pre-installed robots 17 T X E N / 24 New game mode: Skirmish Most players are level 30 Increased DAU / WAU by 5% Slightly increases interest in new content Most interesting type - with unique gameplay

Slide 18

Slide 18 text

18 T X E N / 24 A/B test: popups

Slide 19

Slide 19 text

23,2% 40,0% 46,9% 19 T X E N / 24 A/B test: popups, Android results

Slide 20

Slide 20 text

20 T X E N / 24 A/B test: popups, results Best result: stylized icons Colored pictures are probably not suitable for our game style Players could make hasty conclusions and not follow the link

Slide 21

Slide 21 text

21 T X E N / 24 A/B test: push notifications Standard notifications (Event starts; going on; near end) Gifts New quests New “WR Royale” event New map Special offers Everything in one notification New topic each time

Slide 22

Slide 22 text

22 T X E N / 24 A/B test: push notifications, results Best result: new topic each time All notifications with Emoji were better than the control group The more diverse — the better Text should be as short as possible

Slide 23

Slide 23 text

23 T X E N / 24 Current Experiments: Regional Events Will it be interesting for players? Will it be interesting for players outside target audience? Will players play more on the weekdays? 25 Sep 27 Sep 29 Sep 01 Oct 03 Oct 05 Oct 07 Oct Number of fights per day per person

Slide 24

Slide 24 text

[email protected] Ready for questions! 24 / 24