Using Analytics to Create Viral Games
My note on the video Using Analytics to Create Viral Games by Gavin Eisenbeisz
Viral games are "made" by content creators.
Content creators play games that increase their retention rate, because that's how they get paid
Make games that naturally creates videos with high retention rates.
[[Game Retention Study.pdf]] was made studying retention on no-commentaries gameplay video.
(So something that directly matches how your game can create high retention rates for content creators)
How to increase retention rate on videos
- Work toward a goal
Author's DevLog template video:
- 1st 5 seconds: explain the problem / story -> what goal do I need to accomplish in this video
- 10-15seconds: Additional context
- 45-60 seconds: Rapide progress toward the goal
- Twist / Additional challenge
- Rest of the video: Progress interrupted by intermediate conflicts
- Close to the curve in [[Game Retention Study.pdf]]
- Constant remind the public of the goal and where we are -> like in a car ride with child "Are we there yet?!" -> Progress bar
Game features decreasing retention rate
Keep in mind that some features are bad for retention in a video, but good for player experience, so it's a balance.
- Bad menu screen π½
Applicable only on no-commentaries gameplay video, because on edited video, bad menus will be edited-out most likely - Player deaths πΌ
Every time the player dies, you loose retention -> the tension to move toward the goal is release, so you loose viewers
Why ? They are followed by long game over screens, then loading screen, then have to retrace already traveled paths
=> Cut respawn time to the smallest possible time / instant
In horror game though, players death is expected at least once with a jumpscare - Level / Match completion / Win screens π½
Probably cut in edited video but valid for livestream / no-commentaries gameplay - Notes / Wall of text β«
Notes are worse for retention the longer they are and the sooner in the game they appear
=> If note can't be avoided: Few words, big fonts, pictures, later in the game - Expositions dump β«
Similar to notes - Big cinematics / monologs / dialogs. Worst if not acted - Backtracking πΌ
When player goes to a long dead-end path / when player died and the waypoint is too far off - Credits π½
People leave as soon as the credits rolls
=> Put the credits as a button in you main menu - Walking simulator π½
Variety in environment is needed, as well as landmarks, and goal visible from a distance - Dullness πΌ
Overly dull / slow gameplay moment - Lighting / Obscure style π½
Harsh art style / difficult to read the scene - Repetition πΌ
Core gameplay repeating in under 10sec
=> Add variety in the gameplay
Features that help hold retention
- Chase scene πΌ
Slowly increase the intensity of the chase as time goes on.
Maybe inching the player closer and closer to dying, without actually dying.
Enemy must be close, and visible
Don't end the scene to abruptly => transition to a new goal - Boss fights πΌ
- Parkour π½
Not a lot of data on this one, but even in the same game, levels with a lot of parkour holds attention more than levels with high action - Clear definition of goals β«
Most important feature, basically encompassing all of the above: having a clear goal is creating anticipation of something to come and increases stakes. Creating features with clear definition of goals is key to keep audience engaged (and thus make content creator more likely to pick up you game)
Features that bump retention
- Jumpscares β«
People are re-watching the video at this point.
Subtle jumpscares / stinger are not as effective, but still work
Re-using jumpscares loose the impact => Avoir re-using the same jumpscare - Secrets π½
- Sexualization πΌ
Personal note: I don't want that in my game anyway - Random interesting moments πΌ
Random / Funny / Out of the blue moments that happen once in a game - Trippy stuff / illusions π½
Nuance features and other notes
- Action scenes
Player is in control of the situation (vs a chase where the player is not at all in control)
No real challenge / tension, the player is going to win
Actions scenes can cause bump, and hold for some seconds, but retention will drop drastically after that
Slower / more strategic multiplayer shooter (PUBG) are good are retaining views, because the tension build, the action is intense and the resolution is uncertain - Cutscenes
Can be bad, like big expo dump / long dialogs
But good if short / show the goal - Mystery and Tension
A mysterious story can hold attention, but not if it's build with big expodumps
=> brief conversations / environment cues - Subtitles
Nuance because sometimes the content creator will do the voice adapted to their audience