Games aren't just for fun. They are highly practical for a busy parent. A game is a kind of an automated babysitter - distracts the kid long enough to change diapers, makes him eat his meal, keeps the kid entertained during a long drive, and most importantly works as a personal teacher (okay, that one is pure sci-fi at the moment).
After testing tons of Android games with my almost 2yo son, I am starting to think that most game developers don't test their games on children. Perhaps they don't have kids of the right age in the first place. The games are then littered with disastrous usability flaws to the point of making the game useless. If you are a game developer, read this checklist to get your next game right.
Multi-touch: Children have small hands. They cannot carry the phone in one hand and tap with the other hand like most adults do. They grab one corner of the phone with their thumb touching the screen. Then they tap with their other hand, possibly using multiple fingers. Alternatively they grap the phone with two hands and then move one of their thumbs over the screen.
Most games just lock up the moment the first finger touches the screen. From toddler's point of view, the game just doesn't do anything. All modern phones have accurate and reliable multi-touch, so go use it.
Ads: I keep wondering why people put ads in toddler games. Do you expect the toddler to buy something? Do you get paid per view by clueless advertisers? Or do you think the busy parent trying to evaluate your game before giving it to their kid has any time left to click the ads?
There's no business in ads for toddlers, so stop doing it. Childlock apps will intercept clicks on the ads and the kid will end up back in the childlock menu, which is frustrating for the child who quickly loses interest in your game.
If you want to earn money, just make two versions of the game. One version that is good for maybe 15-60 minutes of play. Parents can use it to evaluate the game and kids get enough time to start liking the game. The second version is paid and it should be good enough for a few hours of playing.
Phone buttons: Don't use them. Kids don't understand what they do. They press these buttons all the time by accident. The best you can do is to ignore all soft and hard buttons. Your game should react only to taps on the visible screen. That's all the toddler is interested in.
Memory: Since it is so easy to exit games by accident by pressing Home button that cannot be intercepted, the only solution I can see to this problem is to make it equally easy to return back to the game. Your game needs to remember its last state and it should fully restore the state if it is restarted quickly after being closed.
Childlock: By no means, do not create your own childlock. Toddlers have short attention spans. They will get bored with your game quickly and the best thing to do in this case is to let them switch games.
There are specialized childlock apps that provide high-quality childlock while allowing kids to switch between approved games. Your own childlock will just interfere everywhere and it will require you to implement menus that end up confusing the toddler when he accidentally triggers them.
Menus: Toddlers cannot read. For a toddler, all menus look alike. Every screen in your game is conceptualized as a place, an enclosed space, a kind of room. Touching stuff on the screen can get the kid to another place in the game.
Since menus all look alike, they are perceived as a single place. When the kid navigates from menu to another menu, it feels like going through maze - all rooms look the same.
Picture menus are a little more varied, but they still look very similar to each other. Ideally, your menus should be unique graphical screens that make it clear to the kid where he is and and where he can go. Or even better, avoid menus and just keep everything on one screen.
Icons: This is related to menus. Many icons just don't make any sense to toddlers. The kid will then avoid clicking on your app in a childlock environment or on home screen. Put something recognizable in all icons. Icon is like a door through which you can peek into the room behind it.
The same applies to menu entries. If your menu is a unique screen, then all menu entries should be icons on that screen. Note that icons aren't necessarily square nor do they need to be placed in a grid.
Features for parents: This includes upgrade offers, review requests, and game settings. Toddlers often play without supervision. That means your whole game should be child-safe and child-friendly with no exceptions. That means no parent-only menus.
Even if there is a riddle or something else that prevents kids from actually using the menu, it's terribly annoying when the menu pops up in the first place, because kids don't know how to get rid of it. If you really need to include settings, then make sure this functionality is triggered by slider/riddle on the visible screen (no soft buttons) or in some other way that prevents kids from accidentally activating the functionality.
Pointing accuracy: Touch screens alone have quite low pointing accuracy. This gets even worse with kids who don't understand that the software perceives touches as singular points. Toddlers just point and drag with their fingers in the approximate area of interest.
Too many times I have seen icons with white space between them that doesn't react to touch. Kids will miss the little icons by a few pixels and then wonder why nothing happened. Extend your tap areas as wide as possible. Similarly, when dragging, make sure the drop area is huge.
Stuttering: This happens surprisingly often. When kids tap the screen, they often perform sequences of taps or even keep tapping one place continuously. Many games have some event handler somewhere that activates animation or sound upon tap. When such game receives quick sequence of taps, the animation/sound gets restarted upon every tap and the game appears to be stuttering. Toddlers find it interesting, but in the end it distracts from the game.
Responsiveness: Toddlers have next to no patience. They play a couple of minutes at a time. If your game takes one minute to load, it could consume one third of their total attention span. Worse yet, the boring waiting will likely turn the kid off and the main content of the game will never reach the kid.
Also nobody hands 600€ phones to toddlers, especially when not fully supervised, because kids do destroy stuff rather quickly. A 2yo kid is most likely holding an older cheap phone or some cheap tablet. Make sure your game runs smoothly on such hardware.
Price: There's no problem paying for the game. I pay for physical toys at the toy store too. The problem is how much your game costs. Toddlers don't spend that much time with individual games and they often completely ignore some games. They need variety, which means the cost of games adds up quickly.
Unless your game is a real goldmine of fun, it's unwise to price it above the minimum price allowed by Google Play store. It's actually a good idea to combine games in one purchase to drop the per-game cost even lower, but make sure you read the point about menus before you mix multiple games in one app.
I can name Puzzingo as one of the games that do many things right and a few things wrong. That's the only game I have bought so far. Most games I have tested however repeat every single mistake I am listing here. Make your game right and I will be happy to pay for it.