Consider:
"How do I make educational games engaging (and) not boring?"
Everything that can be used to make a game engaging in general - and there is a lot to do - should apply to educational games too. The matter is that if you set to create an educational games you are placing upon you an extra set of constraints that may make it hard to desing engaging game mechanics.
Yet, not knowing the gnre, I gotta cover from cooperative to competitive, from single player to massive multiplyer... so, gotta go the abstracts and the generalizations.
To see how to make games engaging, go back and look at how individuals play. Not just video games (and not just humans). A common theme of all play time is make-believe^1, without good immersion any game will become boring. But the immersion is not in the graphics or the music (even tho, they can help) but in the mind of the player.
^1: You should also consider: challenge, chance and thrill.
Consider how a child may enjoy playing with toys. Even if the toys doesn't have rewards or goals. How is that possible? Meditate on that.
Now, if we consider that all games have some form of make-belive, then we can consider that all games are a form of simulation. You may consider them training - except that people enjoy it!
By extension, all games prepare you for something. Even if that something is just playing the game itself. So making a game that teaches something is not hard, but when you settle to make an educational game, you are setting yourself to make a game that teach certain things - in particual those of academic value - and that teach them in a way that is measurable or testable.
Note: Another thing that makes the game boring is repetition, which sadly is one of the things that is used for memorization. So if you want to use repetition for the educational content (because it requires memorization) then your should do it in a way that is not predictible.
How to motivate
Before going futher, I would like to encorage you to ponder... why do you play the games you play?
The naive approach for decision making is that people choose what give them more value. And so there would be some value function that can be applied for each option (which assumes that the set of options is discrete and finite), that gives a value and you pick the one that give you more. Addendums to this approach will consider cost, risk, simulataneos tasks, etc...
But what is "value"? What is it that motivates people? What need do games fullfill to people?
As per Self-determination theory there are three core needs that could be fullfilled by games:
- Competence: Games can give you the skill to do things. Become better at the challenges the game presents.
- Relatedness: Games can serve a means to connect with other people.
- Autonomy: Games can provide a framework in which people can own their actions.
Note: Also constrast with Maslow's hierarchy of needs: Games can provide a sense of belonging, can improve your self-esteem, and may help you become a better individual.
You can use tricks techniques such as random possitive stimulus rewards and conditioning creating habits that are known to bring poeople back to the game.
On the other hand you may inspire curiosity - and imho this is what you should try to do - so that people doesn't come back because they may get a new virtual medal^2 but because they want to know what's next, it is the unknown that calls to them^3. Of course this will be easier with certain topics than others. This is done in a game by surprising the player, and being ever changing. If the game is always the same, well it is boring.
^2: It is known that people attribute real value to virtual things, so yeah, they come back for the virtual medal.
^3: Which is also an argument of why people don't find interesting a game where they do what they do for work. Instead people are interested in games where they are and do things that they don't in real life.
So: More replay value => more engagement.
Yet, if the game is meant to have some challenge. And as the player has more experience their skill is expected to increase, and as such the challenge should increase too. Otherwise the game will become dull over time. If you consdier a competitive game, then you will have to balance the skill level of the players.
A measure of player skill will be handy, it will allow you to provide the appropiate challenge and also will be useful to see how well the players/students are learning the game/lessons.
Difficulty curves and how to adapt them is broad topic, and while it should not be overlooked, that discussion doesn't belong here.
Consider how to use randomness and combinations to provide structure to the game. For instance the game in the OP shows a game where - I pressume - numbers appear at random. Yet, the structure of the game doens't really change, it only makes you stop and think what number to pick. Now consider a game like Mario Kart where the random roll of the power ups (even if it doens't have an uniform distribution) may change the way you play the game (from an defensive to an offensive game play, for instance). Or the random maps of Age of Empires or Civilization may alter the means you use to expand your territory. Or for Minecraft, depending of your spawn location you may have to decide if you dig, build or explore first. And the next day, the player knows that the game may have a new experience for them, and so the player comes back.
Consider the other side of random: grinding. Many games use griding as a way to provide virtual difficulty. Grinding is tiresome, so it translate as making a difficult task. But it is not really making the game harder or more complex, just longer. And for each step of the grind, the player get a random reward, and maybe there is a chance of a really good prize. This thrill provides engagement, but for many player it doens't really make the game fun.
It should be noted that a frustrating game maybe both a good or a bad thing from the engagement perspective. You may decide to drop the game if it is frustating, or you may be enraged with the enemies in the game (but not with the game itself) and want to get back in to play it and beat it (or kill the enemies for the sake of it). If you want to laverage rage for engagement, you have to make sure that the frustation comes from the enemies and not for a bad game (riddled with bugs or bad controls, for example) and that it doesn't break the immersion or consistency.
Your players
Ok, that's the next thing to consider: your players. Not all players are equal.
You maybe interested in learning about Bartle's Taxonomy of Player Types (which is intended for online RPGs). For abstract it classifies player in four types: Achievers (people that want prizes), Explorers (people that want to learn about the game and its mechanics), Killers (people to who like to express power and domination), Socializers (people who like to build communities).
You will find that nobody fits any category exactly, instead people are to some degree one and to some degree the other. But it may be possible to tell if they have a dominant type.
You should provide game mechanics that interest to player of the categories of player that you want to have. For a tradicional MMORPG it is argued that you want a balance of the four. But many games may do well without players of some of the categories [Or the classification may be need to be adapted for other gnres].
- Archivers will want to have prizes, titles, archivements, etc...
- Explorers will want huge worlds, secrets, shortcuts, a good lore, etc...
- Killers want taunts, competition, player vs player, anything that can act upon other players (including griefing and bullying), etc...
- Socializers will want means of communication, chat, teams, guilds, markets, etc...
And make it educational
Just like you can't put all your player on the same bucket, you cannot assume that they all learn the same way. Some may prefer to be presented with examples, some may prefer to be told what to do, and some will prefer to be explore by themselves and learn by committing mistakes.
In fact, it is a sign of a good game the one where you don't notice where the tutorial ends, in particular if it is tematically consistent. For instance consider Portal, most of the game is an elaborate turorial for the final battle, but it doens't feel like you are just playing the tutorial. Also consider Mega Man X that introduces all game mechanics on the tutorial level without telling them to you explictly. An alternative is to have the tutorial be engaging by itself, for instance VR training in Metal Gear Solid plays like its own game, while the main game will do its best to teach you as it goes.
Now, since you want to teach certain particular things in the game, you will have to build them in the game. The ideal is the case where learning to play your game is learning what you want them to learn. You want the player to have some rewards^4 if they do it correctly, in particular you need the player to understand when they have done it wrong and WHY. And you want the player to be able to get textual visual help if they request it.
^4: You may have to tailor "rewards" to the player type, it may not always be in term of archivement. Also consider if the rewards is instrinsic (for instance is a successful execution rewarding by itself) or a skinner box extrinsic (it gives you access to something else, or point to buy stuff, etc...).
Futher more you want it make sense. If you can provide an in-game reason why are you doing that task (that you are trying to teach) the better. If it can be related to the real world much better. Consdier this a test for you: when in your life do you need the skill you are trying to teach? Who will use this kind of skill on a daily basis? How can you turn that into a game? It may be required to find other problem that has isomorphism to what you want to teach (mainly because academic problems are abstract) in order to add them thematically in the game.
Remember that the game has to teach how to be played in addition to whatever is your agenda for the current educational program. So if you can build in it affordance and utilize mechanics that the player may already be familiar with (perhaps from another game, or earlier in the same game) then it will be more accesible.
If you are pitching this to teachers you want statistics. It is good to know how much the players are playing, how long are the session? how frequent are the sessions? how much progress are the players doing? It may help you to improve your game by finding what are the cases where the players are not really engaged, and what are the cases where the players are having trouble learning. This will help the teachers too - they may take it as an aid for evaluation, and it may be an extra incentive if the students can skip some tests or homeworks if they are good at the game. It may be worth noting that if you go to the path of gamifing the claass room, you want cases where the players/students cooperate.