top of page
  • m.decapua

Deconstruction: Grow Series by Eyezmaze


Grow, by Japanese indie developer Eyezmaze, is a series of point-and-click Flash mini-games published on in the late 2000’s. The first Grow game was released in 2006, and the most recent, Grow Park (for mobile, ported to Flash) was released in 2015. The series went rather unnoticed, occasionally garnering online reviews but was largely lost in the sea of browser games at the time. A poll on the forum shows that there were approximately 45,000 regular players at the peak of the series’ popularity in 2009, though some YouTube walkthroughs have reached over a million views, so the player base is certainly out there. I will be discussing the first six games of the series in this document, which are the six originally developed for Flash and with the same main premise.

The premise of these puzzle games is selecting icons in a specific order. Each game presents hundreds -- or hundreds of millions -- of permutations for the player to choose. With only one win state, the player uses trial and error as well as the interpretation of visual feedback to repeatedly attempt to crack the code.


As noted above, the Grow series saw fairly limited gameplay, even at the height of its popularity. Moving the games to the mobile market was a fair choice on Eyezmaze’s part, as browser games have largely fallen off the radar. The playstyle of the Grow franchise is also much more conducive to the experience one expects out of a mobile game -- quick, on the go play sessions in which there is no punishment for losing or exiting mid-game.

Audiences that continue to play the Grow series on a browser are playing for simple, reasoning-based challenges. The inputs are all mouse clicks, making it an accessible game for players who are not used to controllers, as well as those who have slow reflexes or limited motion. Rewarding feedback after each move keeps players immersed, but the majority of gameplay takes place in the player’s mind. This makes the Grow puzzles ideal for strategy gamers, especially those that prefer low-risk environments without violence or stress.

Players that enjoy the Grow series would also be fans of other low-risk puzzle games, particularly games that reward the player for reaching a well-defined win state rather than racking up points. Other games that fall into this category are The Incredible Machine or Sudoku, games where there is one right answer to the puzzle and the player is motivated to reach it.


When beginning a Grow game, players are presented with a blank environment and a series of clickable icons, as seen above. Different installments in the series have varied amounts of icons, with the average around eight. In a game with eight different icons, there are 40,320 possible permutations of player inputs -- with only one right answer. The icons may be contextualized in the game environment (Grow RPG has players “growing” a battlefield by selecting components like a castle and a treasure chest) or may have seemingly no context or connotation at all (Grow 2 presents players with a series of shapes that eventually grow into different creatures and contraptions). Grow Cube is arguably the most effective game in the series, as the icons strike a balance between in-game context and enough ambiguity as to not give away the answer.

Largely, these games work on a turn-based system. Each time the player selects a component to add to the game space, the components they have already placed may level up. Components cannot level up more than once a turn, and many of these upgrades can only be unlocked if other components have reached a certain level. Clicking the icons in the perfect order will allow for the correct amount of turn cycles and the ideal synergy needed for each component to reach its maximum level. When all components are at MAX, the player wins the game.

After all the icons have been selected and all components have been added to the game, the player is presented with the levels that each component reached. These levels are not based solely on the component’s position in the sequence, but on the overall synergy of the player input. Some icons may also display MAX even if they were not selected at the right time -- for instance, if the player clicks an icon too early in the game, but the component still reaches its “full potential,” it will display the MAX score. This data indicates to the player their overall success as well as which items still need more turns to grow; players can synthesize their scores to see how to rearrange their selections on their next run. See the image below for an example of the post-game score display in a win state.

The primary gameplay experience can be summed up as follows (in the beginner English of Eyezmaze himself):

“You click the panel and makes things in field grow.

You would complete the game when all panels' level would become Lv.MAX.

You have to think about the order since there are deep relations with each panel when they are growing. “

Each Grow game features some variation of this statement as a caption to the game, followed by apologies from the developer for taking a long time to produce and revise the games.


The player experience in Grow is characterized by qualities: logical reasoning, persistence through trial and error, and wonder. Through the many variations of the same basic concept (clicking components in the right order to unlock a win state), these principles remain the same. When looking at the larger Grow library, even dabblings in genres outside its niche, these components are present in the player experience.

The first tenet of the Grow experience is logical reasoning. Players must use visual feedback and repeated failures to extrapolate their interpretations of the systems. Players are not provided with any tutorial, or any information on how the game works. Therefore, the first handful of game cycles are usually an exercise in deciphering the goal and types of visual feedback the player should be looking for. Interactions between different components are key to understanding the complexity of each component’s leveling path. These interactions are indicated in each turn cycle when the components level up, triggering a new animation or evolution, or remain stagnant. Combining a growing understanding of gameplay with the data provided on the end-game score screen helps the player grow and reason further between each run.

To win the game -- and to discover the systems -- the player has to be willing to repeat the game. Not five or ten times, but dozens, maybe hundreds. Recall that some iterations of the game feature one winning combination out of millions of permutations. A committed Grow player will persist through hours of play time to decipher the correct input order. One game of Grow generally lasts about five minutes if the player does not hit restart, which they may do at any time. These bite-sized chunks of gameplay allow the player to hit the restart button over and over again without feeling bored or discouraged. If the player knows they made an incorrect move early in the game, they can skip the end game scorecard and start right over again.

The willingness to persist through dozens of failures may be attributed to the games’ overwhelming pleasant and rewarding atmospheres. With the right perspective, there are no actual fail states, just opportunities to gain a greater understanding of the system. The game doesn’t punish the player in any way for reaching the end without winning -- there is no score, no “game over” banner, no death and destruction. The player is simply informed of which components were used effectively and which were not, then allowed to click the same reset button that had been on their screen through the entire game. In a few iterations of the game, like Grow Tower, the player is given visual indication of their “failures” through animations, but this is never presented as a loss, just a missed opportunity to consider next time. The player is not encouraged to restart the game or to give up even when this happens. As seen in the image below, from Grow Tower, one component has broken and fallen off the tower. Still, the player may select from the other three components and finish their attempt at winning.

A critical attribute of the replayability discussed above is the unique end result the player produces with each complete input series. Their creation, whether it is a creature or a tiny ecosystem, is very alive even when it is not the ideal win state. The animations in Grow Cube, for instance, do not only engage the player by providing visual feedback. They also reach the player on a more emotional level. The art style and behaviors of the components, especially as they evolve, keep the player intrigued. The animations evoke a sense of wonder and feel almost dream-like. The player feels rewarded each turn and each game cycle due to the unique animations and variable results.

Systems & Mechanics

Input System -- This is the only truly outward-facing system in the game. The clicking or dragging mechanics that place each component onto the game space are explained to the player through the background provided by the developer. When the player selects a component using the mouse, its icon disappears and a corresponding object is created within the game space. There is no indicator of which order the components have been placed in; the player must remember the order of their inputs within and between game cycles.

The input system rules the other systems in the Grow games. The objective, at its core, is to execute the correct series of selections. If the player fails to intuit this, they will have a fundamental misunderstanding of the gameplay. The growth of each component and their synergistic interactions, to be discussed further, rely entirely on inputting them in the right order.

Turn-based Growth System -- After the player selects a component, they are barred from making another selection until the existing components have evolved or moved. These animations may show components growing, relocating, or becoming something else entirely. Some components level up every turn, while others rely on more specific conditions. Understanding how many turns it takes for each component to reach MAX is integral to decoding the sequence. The leveling-up or evolution of each component helps to create a narrative for the player to follow as they try to interpret the context of each icon. The player is rewarded through this system after every action they take, which helps to sustain motivation and reinforce the casual atmosphere. The animations are an incredibly important piece of this system, as they provide all of the feedback and reward that is communicated between turns.

Component Synergy System -- In the proper order, each component has a direct effect on the environment around it. In Grow Tower, for example, a hand that grows out of one component presses a button on another component if they are selected in the right order. This synergy is often communicated through animations. This system provides the player with a reward for solving a part of the puzzle; the scene becomes more rich with every point of synergy. The player is also rewarded with information -- they can make a visual connection between components to help them understand what belongs where.

Variances between games -- While this deconstruction covers only a few, there are a dozen games in the Grow series. Here, a few key differences in system design in the first six games will be noted.

  • Grow 1 is presented as a series of A or B decisions between components, rather than displaying every option at the same time. The growth of the creation follows more branching paths than in other games. In this version, the icons are recognizable objects that seem to have no context when related to the growing creation.

  • Grow 2 introduces the turn-based growth system to the series. This is also the first game to present all options at once and follow a more linear structure. The synergy between components is emphasized here, but there still is no context as to what the icons represent. In Grow 2, the icons are plain white shapes before they are added to the scene.

  • Grow 3 introduces a scoring system that is not present in the other games, and has a much longer array of components to choose from with twelve icons. The player’s score is increased each turn, but there is no reward for a high score and no noticeable connection between the score and player success. The score does not impact the end state in any way. It is unsurprising that this system was eliminated in other iterations -- it is unnecessary and distracting to the player.

  • Grow RPG seems to be the attempt at creating context for the components, as they are all skinned with high-fantasy symbols. The player must make their selections and then watch a battle play out in the game space, which will inevitably be lost due to not selecting the correct sequence. This might make wins more satisfying, as the win state shows a clear victory, but this introduction of a lose state weakens the series’ supportive, trial and error experience. This game includes many more systems than the originals, like a health bar, and they feel unnecessary to the player experience.

  • Grow Cube hits the sweet spot of context and interpretation. It provides the player with familiar concepts (such as tiny humans that will probably need water to live, prompting the player to select water after people) that still evoke a sense of wonder when they interact. The amount of components to select from is also pretty ideal, at ten. There are enough permutations that it is still challenging, but not so many that it feels impossible.

  • Grow Tower explores growing up rather than growing out, and gives the player visual feedback in a new way. Along the side of the screen are marks against which the player can measure their tower, allowing them to judge their success in terms of height. This presents a new incentive to beat a high score, though this is mostly superficial. If the player inputs the correct sequence, they will reach the same height every time. The measurement serves as an indicator to progress, but there is no real reward.


The Grow franchise is successful because as a whole it is greater than the sum of its parts. There are a lot of very weak games in the series, with weaknesses like extraneous systems or dilution of core experiences, as indicated above. However, when the series is examined as multiple segments of a singular unit, the player is afforded new opportunities to challenge themselves with different systems to decode. Once they have solved the puzzle in Grow 1, they can switch to the next game for a fresh puzzle that includes different systems. The continuity of experience that is maintained between installments makes the Grow series exceptional as a whole.

It can be argued that certain Grow games, such as Grow Cube, are objectively better than others. In this game, the core concepts of wonder, persistence, and reasoning are at their peak and are produced by the highly tuned systems of the game. When compared to Grow RPG, for instance, the player experience is in line with the intent to a much higher degree in this game. That provides a sense of authenticity that makes the better Grow games more effective.

Despite these fluctuations in quality, the series is very effective at producing an ideal logic-based player experience. The gameplay is mentally challenging but not physically demanding at all, and can immerse the player deeply. The variation in tone and content between games helps to break up long play sessions, while the consistency of core systems keeps the player focused.

Players who enjoy Grow want to see intellectual challenges, low-risk gameplay, and the pursuit of one right answer. Grow, as a series, satisfies all of these needs very well. The player can challenge themselves with one installment until they do not feel challenged anymore, and then pick up a new system to examine. With no punishment for restarting and leaving, and no punishment for inputting the incorrect sequence, the player does not feel pressure and instead can play casually at times or seriously at others. The pursuit of one right answer -- as well as the other experiences created by the series’ core mechanics -- drives the player through each game despite repeated trial and error. Luckily, players now have a library of twelve games to peruse through and do not need to wait multiple years for the next game to come out. In this sense, Grow is more effective as a singular unit now than any singular installment was upon its release.

71 views0 comments
bottom of page