Forums » Suggestions
Initial Android control tutorial: Show background
I think the initial Android controls tutorial should display a sector background, similar to the later steps of the tutorials. As it stands, the player is presented with a black background and some stars or debris. I don't feel this is very inspiring and is certainly not indicative of what they will experience when in the game.
Even though the later tutorials do have a proper background, how many people may be dissuaded from trying more by the first screen?
Even though the later tutorials do have a proper background, how many people may be dissuaded from trying more by the first screen?
I think the initial Android controls tutorial should display a sector background, similar to the later steps of the tutorials.
The reason why it does this, is the "stars" in the initial tutorial actually demonstrate 3D movement, helping the player to understand the ramifications of their motions. Having a traditional background does nothing to indicate side-to-side strafing motion, for instance.
Furthermore, adding the background behind the starfield would essentially make the starfield invisible, removing the value to the beginning player.
how many people may be dissuaded from trying more by the first screen?
Well, that's why we measure exactly how many players fall-out of their initial runtime experience, and exactly where they fall out, all the way from the first moment they run the game, up to their docking with the Capitol Station.
I think we currently have around 50 different (anonymized) analytic progress points in the first 5 minutes of the game, also cross-referenced by input type, platform, language/locale, and other parameters.
Then we A/B test changes that we make, against similar cohorts of players from the same platform (for instance, all the iteration you've seen on Android over this past year), correcting for "ambient" new-player cohorts and not previous users coming back from Newsletters, for instance.
I don't feel this is very inspiring and is certainly not indicative of what they will experience when in the game.
We used to start the game with a "battle scene" with the intention of being very inspiring, and truly "showing the game"; but we lost far more players using that.
I do see value in providing the most graphically exciting scene to the new-player as possible, as quickly as possible. But there has been a lot of iteration and tradeoff between being "inspiring" and the fact that new players really struggle with the controls.
One of the things that we discovered back around 2004, when we launched on the Mac, was that unlike PC.. many of our Mac users had never played a game before, of any kind. Which was a completely different level of user than what we were expecting, and a much bigger challenge to introduce to the game.
There seems to be a substantial breadth on mobile as well, some players may be experienced action gamers from PUBG and Fortnite, while others have never played more than "Angry Birds".
So, anyway, the best thing that I (or my industry as a whole) seem to know how to do, to improve this.. is to scientifically measure the results of changes as robustly as possible, and experiment with different designs, while watching for statistically significant variations.
I hope that's helpful.
The reason why it does this, is the "stars" in the initial tutorial actually demonstrate 3D movement, helping the player to understand the ramifications of their motions. Having a traditional background does nothing to indicate side-to-side strafing motion, for instance.
Furthermore, adding the background behind the starfield would essentially make the starfield invisible, removing the value to the beginning player.
how many people may be dissuaded from trying more by the first screen?
Well, that's why we measure exactly how many players fall-out of their initial runtime experience, and exactly where they fall out, all the way from the first moment they run the game, up to their docking with the Capitol Station.
I think we currently have around 50 different (anonymized) analytic progress points in the first 5 minutes of the game, also cross-referenced by input type, platform, language/locale, and other parameters.
Then we A/B test changes that we make, against similar cohorts of players from the same platform (for instance, all the iteration you've seen on Android over this past year), correcting for "ambient" new-player cohorts and not previous users coming back from Newsletters, for instance.
I don't feel this is very inspiring and is certainly not indicative of what they will experience when in the game.
We used to start the game with a "battle scene" with the intention of being very inspiring, and truly "showing the game"; but we lost far more players using that.
I do see value in providing the most graphically exciting scene to the new-player as possible, as quickly as possible. But there has been a lot of iteration and tradeoff between being "inspiring" and the fact that new players really struggle with the controls.
One of the things that we discovered back around 2004, when we launched on the Mac, was that unlike PC.. many of our Mac users had never played a game before, of any kind. Which was a completely different level of user than what we were expecting, and a much bigger challenge to introduce to the game.
There seems to be a substantial breadth on mobile as well, some players may be experienced action gamers from PUBG and Fortnite, while others have never played more than "Angry Birds".
So, anyway, the best thing that I (or my industry as a whole) seem to know how to do, to improve this.. is to scientifically measure the results of changes as robustly as possible, and experiment with different designs, while watching for statistically significant variations.
I hope that's helpful.