Figure 2.1 The genre-defining platform game Super Mario Bros. (Nintendo, 1985).
Shall we play a game? You choose: Chess, Super Mario Bros., or Angry Birds. I’m giving you a choice because I don’t know whether you are familiar with all three of them. I talked about Chess in the previous chapter: the Western’s world’s arguably most famous board game, played by physically moving pieces such as pawns, kings, and queens on a board with alternating black and white squares. By moving these pieces so that they threaten and capture your opponent’s pieces, you can ultimately win over your opponent by surrounding her king. It has changed little since it was invented millennia ago.
Super Mario Bros. is the platform game that accompanied the European/American launch of Nintendo’s 8-bit Nintendo Entertainment System (NES) console back in 1985 (figure 2.1). By pressing buttons on a little plastic box, you commandeer the jovial plumber, Mario, as he avoids evil turtles, stomps menacing mushroom men, jumps over gaps, collects coins, and saves the princess who has been kidnapped by a giant lizard. Sequels of the game keep being developed for all of Nintendo’s hardware and in addition to hundreds of millions of copies of the Super Mario Bros. games that have been sold legitimately, there are dozens of unauthorized versions of the game available for any conceivable hardware platform.
Figure 2.1 The genre-defining platform game Super Mario Bros. (Nintendo, 1985).
Angry Birds is the mobile gaming phenomenon from 2009 by Finnish company Rovio (figure 2.2). You point at and swipe your fingers over your phone’s touch screen to fling an assortment of birds at various structures, and your goal is to make the structures collapse on top of evil green pigs that have stolen your eggs. The original game, as well as a myriad of sequels are available for iPhone, iPad, and Android devices and have topped best-seller lists on all those platforms.
Figure 2.2 Angry Birds (Rovio, 2009), the physics puzzler that was on seemingly everyone’s iPhone after it debuted.
My guess is that you have played all three of these games, or at least seen someone play them. If not, you have probably played two of them, or at very least one of them. In the extremely unlikely event that you don’t know either Chess, Super Mario Bros., or Angry Birds, I’m somewhat confused as to who you are and what world you live in. Are you reading this book in the far future? I’m just going to assume you play games of some sort.
Having ascertained that you play games, let me now ask: Why do you play games? To relax, have a good time, lose yourself a bit? Perhaps as a way of socializing with friends? Almost certainly not as some sort of brain exercise. But let’s look at what you are really doing:
You plan. In Chess, you are planning for your victory by imagining a sequence of several moves that you will take to reach checkmate, or at least capture one of your opponent’s pieces. If you are any good, you are also taking your opponent’s countermoves into account and making contingency plans if they do not fall into your elaborately laid traps. In Super Mario Bros., you are planning whether to take the higher path, which brings more reward but is riskier, or the safer lower path (figure 2.3). You are also planning to venture down that pipe that might bring you to a hidden treasure chamber, or to continue past it, depending on how much time you have left and how eager you are to finish the level. You may be planning to eat the power-up that lets you get through that wall so you can flick a switch that releases a bean from which you can grow a beanstalk that lets you climb up to that cloud you want to get to. In Angry Birds, you are planning where to throw each bird so as to achieve maximum destruction with the fewest birds. If you crush the ice wall with the blue bird, you can then hit that cavity with the black bomb bird, collapsing the main structure, and finish off that cowardly hiding pig with your red bird.
Figure 2.3 A planning algorithm (a version of the A* algorithm, discussed in chapter 4) playing a clone of Super Mario Bros. The black lines show the various future paths the algorithm is considering.
You think spatially. Chess takes place on a two-dimensional grid, where cells that are not occupied by white or black pieces are “empty.” Those who have played the game a number of times and internalized its rules start seeing some of the opportunities and threats directly as they look at the board. The fact that the queen is threatened stands out like an X in a row of Os, and the possible positions a knight can go to are immediately visible on the board. In Super Mario Bros., you need to estimate the trajectory of jumps to see whether you can pass gaps and bounce off enemies, which means seeing the jump in your mind’s eye before you execute it. You also need to estimate whether you can get through that small aperture with your current size (Mario can change size) and whether that path over there leads anywhere. In Angry Birds, you also need to estimate trajectories, sometimes very complicated ones that involve bouncing and weird gravity, and you may also need to determine whether you can fit that bird in the narrow passage between that pixelated rock and that virtual hard place.
You predict the game and your opponent(s). In Chess, predicting what your opponent will do is essential to successful play. If you knew how your opponent would react to your moves, you could plan your strategies with perfect certainty that they would succeed. Super Mario Bros. and Angry Birds are usually not adversarial games (you are not playing against a human opponent), but instead the challenge is to predict the actions and reactions of the environment. When will the cannon fire? Which way will that turtle face if I land to the left of it? Will the monster lizard advance all the way if I don’t jump up on the platform? And how exactly will that complex building collapse if I knock out the bottom support, where will all the pieces land, and will one of them set off that box of TNT to create a nice chain reaction? While randomness may play some role in Angry Birds (Super Mario Bros. is completely deterministic), the difficulty stems mainly from the very complex interactions among the various objects in the game.
You assess yourself. “Know yourself,” said Socrates. He was probably not talking about Chess and certainly not about Angry Birds, but really, knowing yourself is an invaluable asset when playing games. Overestimating your skill will make you play recklessly and most likely lose; underestimating your skill means that you will not attempt that risky strategy that could have won the game for you. Also, you need to take your affect into account and correct for it. Are you currently off-balance because your plan did not work out, unhealthily buoyed by your recent success, or perhaps driven by lust for revenge for that bastard move your opponent just made to capture your queen? Well, then you need to take that into account. Don’t try that ten-moves-deep strategy if you know it’s based on wishful thinking rather than careful assessment of the situation. The same is true for Super Mario Bros. and Angry Birds: if you did not know your own skill level, you would not be able to progress in the game because you would try strategies that were too hard for you. You might also be better at executing some tactics, such as long jumps or setting traps with your knights, than others, such as precision shooting or moving in quickly to surround the king.
You move. It is true that Chess does not involve much in the way of motor skills, at least unless the game degenerates into a brawl, but the other two games certainly do. Super Mario Bros. has you pressing two keys and a D-pad, which is itself eight direction keys, very frequently (often multiple presses per second). Angry Birds demands very fine control of your finger movements on the screen in order to shoot the bird in the right direction with the right force and activate its special ability at the right time. In both games, these movements must be coordinated with what happens on screen and perfectly timed. It is the sensorimotor aspects of these games that tend to picked up very quickly by five-year-old kids but not always by their frustrated parents.
Of course, other games offer other challenges. First-person shooters such as Halo or Call of Duty challenge your spatial navigation in three dimensions, and in multiplayer modes, they throw you straight into the complexities of team strategy. Role-playing games such as Skyrim and Mass Effect require you to understand the motives behind the actions of complex characters, resolve ethical dilemmas, and navigate perilous politics (at least if you play them the way they are meant to be played—although you can get pretty far in some of them by just shooting everything that moves). Economic simulation games like SimCity and Transport Tycoon require you to understand and influence complex economical systems.
One way of trying to outline what types of cognitive challenges games offer is to look to psychology or, more precisely, psychometrics, to see if there is some handy list of cognitive abilities. We could then try to figure out how each of these abilities is required (or not) for playing games of different types. It turns out that there are indeed such lists. In particular, the Cattell-Horn-Carroll (CHC) theory divides general intelligence into eleven different “broad cognitive abilities,” which are further subdivided into many more specialized cognitive abilities.1 This taxonomy is based on statistical analysis of hundreds of different cognitive tests and is widely accepted in the psychometrics community (though as new empirical evidence comes in, categories are modified and added).
Table 2.1 lists the eleven broad cognitive abilities from CHC theory and gives some examples of situations in games where they are used. Note that this is very far from a complete list; I’ve more or less listed some of the first examples that came to mind, trying to get some diversity in terms of game genres. My guess is that almost any game would make use of at least five different cognitive abilities (Super Mario Bros., Angry Birds, and Chess certainly do), but this is just a guess and I’m not aware of anyone having done research on it. Someone really should do that research.
| Broad cognitive ability | Example use in games |
| Comprehension-knowledge | Communicating with other players in all manner of multiplayer games, from Bridge to Gears of War and World of Warcraft |
| Fluid reasoning | Combining evidence to isolate suspects in Phoenix Wright; solving puzzles in Drop7 |
| Quantitative knowledge | Controlling complex systems involving lots of quantitative data, such as in SimCity, or character management in Dungeons and Dragons |
| Reading and writing ability | Reading instructions in games, following conversations, and selecting dialogue options in role-playing games such as Mass Effect; writing commands in text adventures such as Zork |
| Short-term memory | Everywhere! For example, remembering recently played cards in Texas hold’em poker or Hearthstone |
| Long-term storage and retrieval | Recalling previous games of Chess or StarCraft that resemble the current game to gain insights into strategy |
| Visual processing | Spotting the possible tile matches in Candy Crush Saga or the enemy snipers in Call of Duty |
| Auditory processing | Becoming aware of approaching zombies (and from which direction) in Left 4 Dead; overhearing secret negotiations in Diplomacy |
| Processing speed | Rotating pieces correctly in Tetris; micromanaging battles in StarCraft; playing speed Chess |
| Decision or reaction time/speed | Everywhere! For example, countering moves in Street Fighter or deciding what fruits to slice in Fruit Ninja |
In sum, we use many different forms of intelligence when we play games, more or less all the time. This sounds like a lot of hard work. It’s amazing that playing games actually relaxes you, but it does. (I took several breaks to play games while writing this chapter.)
So far, we have discussed only the individual skills you exercise when you play a game. But you do not exercise them the same way all the time; you are building your skills as you play. It certainly does not feel as if you are taking a class while you are playing a game (if it does, it’s not a very good game). Yet you learn. Here is proof: you are much better at the game after playing it for some time than you were when you started. Try playing one of the early levels in Super Mario Bros. or Angry Birds again. Or try playing a Chess computer at novice difficulty again, the one that beat you roundly the first time you tried. Piece of cake.
Raph Koster, a famous game designer, has made the argument that learning is the main reason games are fun.2 Good games are designed to teach you how to play them; the better they teach you, the better designed they are. You have fun because you are learning to play the game, and when you stop learning, you stop having fun. If there is nothing more to learn, you grow tired of the game. Therefore, a trivial game that you can beat on your first attempt is not interesting, and neither is a near-impossible game that you cannot make any progress on. A well-designed game instead offers you a long, smooth difficulty progression where you can keep learning as you play. We can say that the game is accessible and deep.
For example, when you started playing Super Mario Bros., you first had to learn what the buttons did—button A makes Mario jump and pressing the D-pad in different directions makes him walk left or right—and you then had to learn how to tackle the various challenges that the game presented. “So, there’s a walking mushroom approaching. What can I do? Aha! I can jump on it!” As you progressed through the levels of Super Mario Bros., you would have noticed that the challenges presented became trickier and trickier, but also that you were better and better prepared to handle them.
The oft-imitated design of Super Mario Bros.’s levels typically introduces a basic version of some challenge (say, a jump over a gap or an enemy caught in a valley between two pipes) and later presents more advanced versions of the same challenge (longer gaps, different kinds of enemies in the valley) or combinations of several earlier challenges (a long jump over a gap, after which you immediately land in a valley full of enemies). Every time, the completion of some previous challenges has prepared you for tackling the new, more advanced challenge. And after a while, when you thought that there were no ways left to produce new, interesting challenges by varying the existing challenges, the game throws in some new ingredient that offers further variation and deeper challenges. One such new ingredient, introduced rather late in the game, is the spiky enemy, which cannot be defeated by jumping on top of it. Adding spiky enemies to existing challenges forces you to develop new strategies to cope with the familiar-looking but fresh challenges. Finally, even when you’ve managed to finish the whole game (beating the boss at the last level and rescuing the princess), there is much left to discover, including hidden areas and treasures, and how to beat the whole game in under ten minutes (if you’re of that persuasion). Super Mario Bros. is widely regarded as a masterpiece of game design, partly by virtue of being a masterpiece of pedagogics: a deep and rewarding course where the next improvement is always within reach.
The story is much the same for Angry Birds. First, you learn the basic motor skills of swiping your fingers to fling birds, before proceeding to understand how the various birds interact with the materials the towers are built from and which parts of the towers are most crucial to hit in order to raze the whole tower. Every once in a while, the game throws in new types of material, new birds, and other devices to expand the range of challenges. Even in Chess, the progression is similar, with the obvious exceptions that very little in the way of motor skills is necessary and that learning takes place over many games of Chess rather than on many levels of the same. First, you learn the basic rules of Chess, including how the pieces move and capture. Then you learn more advanced rules, which presuppose mastery of the simpler rules, including castling and when the game is a draw. You can then move on to learning heuristics,3 first simple and then more advanced; then you learn opening books (lists of good opening moves), the quirks of particular players and playing styles, and so on.
The idea that playing (games or otherwise) goes hand in hand with learning is not unique to game design. The developmental psychologist Lev Vygotsky talks about “proximal zones of development” in children’s play, where kids typically choose to play with objects and tasks that are just outside their capacities because these are the most rewarding.4 Relatedly, the creativity theorist Mihaly Czikszentmihalyi’s concept of flow states that flow can be experienced when performing a task that is so hard as to challenge you but not easy enough to bore you, and where the difficulty of the task increases as your performance improves. Czikszentmihalyi developed this concept in reference to artistic and scientific creativity, but it applies just as well to game playing.5 From a seemingly completely different perspective, the machine learning researcher Jürgen Schmidhuber introduced a mathematical formalization of curiosity. In his model, a curious agent (human or artificial) goes looking for tasks that allow it to improve its model of the task, and therefore its capacity to perform the task.6 In other words, according to Schmidhuber’s theory, a mathematically optimally curious agent does the same thing as a young kid learning about the world by playing with it, or as a discerning player choosing to play games she likes or choosing challenges that seem interesting within that game.
To sum all this up, it seems that games challenge your brain in more than one way—way more than one way—and, furthermore, that good games are designed to keep you challenged by ramping up the challenge (and providing additional challenges) in a pedagogical manner. Schools should take note (some do). It is very likely that the good games, those that we choose to play and keep coming back to, are so good at least partly because they succeed in persistently challenging our brains in multiple ways.
So you definitely use your intelligence when you play games. At the same time, we saw in the previous chapter that it is possible to build software that can play Chess or Go better than any human while seemingly not being intelligent. So how come intelligence is needed for humans to play games, but not for machines to play them? What’s going on here? It is time to try to nail down what we mean by artificial intelligence and, in the process, what we mean by intelligence.