Siege of Centauri
Get it Now

Siege of Centauri Dev Journal: What Makes A Good Tower Defense Game?

Published on Friday, August 23, 2019 By GGTheMachine In Siege of Centauri Dev Journals

We love Tower Defense games at Stardock. We’ve played our fair share, and many have served as inspiration and reference for our work on Siege of Centauri. Before I started on Siege, I went on a Tower Defense binge to analyze and understand what makes for a good Tower Defense game. Here are the pillars that I think are critical to developing a good tower defense game.

Variety of Creeps

Tower Defense is inherently a repetitive genre, and generally without much narrative punch to keep players engaged. As a result, Tower Defense games need to prioritize ensuring the game doesn’t feel repetitive. Meaningful variety of creeps is critical, not just with different stats but also qualities to make countering them require different towers or tactics. Swarm creeps should be vulnerable to splash but resistant to Cannon Towers, for example. You can think of it as designing creeps and towers in a rock/paper/scissors counter system, but with more interesting quirks mixed in such as shields or healing. Creeps can’t just be superficially different; the player needs to interact differently with them to keep them on their toes thinking about how to respond. Defaulting back to the same strategy over and over is where the feeling of repetition comes from.


Variety of Towers

The design of towers and creeps should work hand-in-hand. A rock/paper/scissors design is a good start, but it should be a lot more interesting than just that. Towers should have different qualities that determine how well they perform in specific positions. Some can shine in choke points, others could be long-range and fire over mountains, and others being support towers. When towers have distinct qualities and roles, players can try out different strategies and will need to when facing different creep types and on different maps. Tower variety doesn’t have to be more and more separate towers, it can also be branching upgrade paths.

Variety of Missions

When towers have unique qualities and excel in varied situations, differing map design forces the player to improvise as what worked in the previous mission may not work again. Some maps can be symmetrical while others have vastly different lanes. Some maps can have multiple entrances or multiple exits, or just one. Some missions may have air units while others don’t. Missions can also vary significantly in other features and quirks such as walling, bonus objectives, and scripted events. It’s great to have additional features in a tower defense game, but they don’t need to be consistent across each mission. Ultimately, the variety in creeps, towers, and missions needs to have consequences for how the players interact with the game. The result of creep, tower and map variety elegantly intertwined together is depth, strategic diversity, and a fun challenge.


Interaction & Pacing

Tower defense games can feel passive as the towers are doing all the work, you’re just building them! If the player is only performing actions once every 30 seconds, the game will likely feel boring as it’s mostly watching and waiting. Watching and waiting isn’t inherently a bad thing, it sure is satisfying watching your towers blow stuff up, but it can’t make up too much of the game. A tower defense needs to ensure players have stuff to do to equalize the ratio of gameplay to watching. You don’t want to make Tower Defense overly twitchy, "gameplay" doesn't have to be performing actions, it could just be deliberation of each action.

  • Creep leaking lethality
    Creep leaking should be common, but not overly punishing. When creeps are leaking, the player has to think about building new turrets forward and using abilities instead of just being safe or only reinforcing the front line. To make leaks less punishing, I’d recommend having some objective to protect rather than simply a life go down if creeps run off the map. In Siege, we have the Colony to protect, which is equipped with basic weaponry. If creeps get past your defenses they'll apply some damage before being taken out. It’s not good if some creeps get through, but not overly punishing unless they're tier 2 and above.
  • Support abilities
    When players have abilities to deploy, they're always thinking in the back of their head “Hmm, should I use this ability now?” Abilities don’t all need to be high impact; you can have some cheap abilities to encourage players to be active with their use.
  • Incoming wave panel
    Allow players to study the incoming composition and direction of attackers. This lets them think about what they should do in advance instead of blind luck. If a player is thinking strategy they’re playing the game.
  • Rush waves
    If all else fails, simply give the player the option to fast forward or rush waves so they can play at a faster pace. Rushing Waves should come with an incentive of bonus resources or score.
  • Focus fire
    Manual focus fire of towers can give players something to do. Manually focus firing isn’t particularly fun and can feel overly twitchy, but it’s an option to get the player more hands on. Sol Survivor had a cool approach of making it an orbital ability.




Lots of towers and creeps types is overwhelming to new players, so it should be gradually introduced to the player over time. For even more strategic investment and player agency, the unlock order or global upgrades could be chosen by the player. The content variety should have progression, but so should the complexity, challenge, and pacing. You want a deep challenging experience, but without front-loading it to beginners. Here are some additional ways in which missions should progress and scale over time:

  • Map size
  • Mission duration (how many waves)
  • Starting cash and difficulty of early waves.
  • Delay between creep waves
  • Number of creep entrances, forks, and exits
  • Complexity of walling/blocking
  • Frequency of features and mechanics such as air and stealth units.
  • Difficulty of bonus objectives

Progression can also be cosmetic to the gratify the player of their success. Tower Defense games generally have meta screens where the missions are selected, showing the progress throughout the meta screen and swapping to another one is a lot more satisfying than just a checklist of missions. The differences in missions across the meta screen/s can have changes to the presentation of the game, such as different biomes, UI color scheme, and the soundtrack to match it. Enemy types could be slightly reskinned to reflect the biome change. 



Quality of Life & Interface Clarity

Tower Defense is a type of strategy game, so it’s all about the player making meaningful decisions. The player should have as much information as possible available to them so they can make informed decisions and strategic choices. Ultimately, a tower defense should not be about guesswork, but calculated meaningful decisions which rely upon clarity and ease of information to the player. This requires having an excellent interface and quality of life features. Here are some examples of great quality of life features to include:

  • Tower range display and displaying upgrade bonuses
  • Incoming wave panel
  • Detailed tooltips and info about creeps and towers
  • Conveying of tower strength and qualities.
  • Target priorities, automatic or configurable.
  • Auto checkpoints/Save games


Creeps and tower effects should be as intuitive and visually conveyed as possible:

Blue Rings convey the tower in the centre is providing a bonus:


It may not always be viable to convey status effects through natural game effects, in which case 2D popups is a good alternative:



Tower Defense games typically lack a compelling narrative due to the contrived gameplay and lack of characters. Still, tower defense games should attempt to be immersive by contextualizing the game and make the player feel like what they’re doing matters and makes sense. Here are some good practices for achieving player immersion.

  • Meta screen
    A picture says a thousand words. A meta map can show territory lines of enemies attacking, being pushed back, villages that are threatened, etc. This makes it clear why they're there.
  • Expositional dialog
    Where are you, and why? Tell this to the player. This information could be dialog at the start of a mission or a mission briefing the player reads before they load in.
  • Justification for gameplay mechanics
    Have the Creeps attack a structure like a Colony that you need to defend, instead of just running to the other side of the map and despawning.
  • End game screen popups
    Display something to the player at the end that tells them why their actions mattered. How many Colonists did they save, did they get promoted to Sergeant for their heroic defense?



Replay Value

Quality over quantity is the right approach for Tower Defense games, but that means they typically have a short playthrough length. A Tower Defense game should be designed in a way that repeat playthroughs are enjoyable. Here’s some ways tower defense games achieve this:

  • Difficulty options
    What worked on Medium might not work on Hard, this forces players to mix up their strategies. Tower Defense can have many difficulties with incredibly challenging scaling, and the difficulty options don’t have to be one-dimensional. Why not have a mutator that doubles the movement speed of units?
  • Feedback
    Give feedback to the player after they finish a mission with metrics such as damage received, time taken, score, and waves rushed is a challenge to do it again and try better. Can you beat the mission this time without taking any damage to the colony? What about beating in under 10 minutes?
  • Bonus Objectives and Challenges
    Bonus Objectives and challenges can be a form of difficulty. “Win without using Lightning Towers” may force the player to mix up their strategy or use different strategies to beat each of the bonus challenges.
  • Score and Leaderboards
    Clocking a high score is addictive! Competing against other people is can be more exciting than just increasing difficulties. Score can measure factors such as speed and efficiency, which is multiplied by difficulty scaling. Scores can be global and for steam friends. Being rank #2791 kinda sucks, but beating your friend's score is fun!
  • Endless Survival
    In Endless Survival mode the quantity and health of attackers continuously increases until the player is inevitably defeated. Seeing how long you can survive is a score which can be measured against other players and can give lots of replay value. The impending doom is also fun and tense!




The crucial foundation of Tower Defense design is to have a meaningful variety of creeps, towers, and missions. The combination of these three pillars creates strategy and depth; otherwise, tower defense games can feel shallow and repetitive. Tower Defense can feel passive, so gameplay should be designed with features that give the player stuff to do or think about which makes them feel active. The content, complexity, pacing, and features should start simple and gradually progress over time to keep players engaged but not overwhelmed. The interface should be designed with clarity and quality of life features to ensure the player has all the information they need to make deliberate, calculated decisions. Tower Defense should contextualize their gameplay to immerse and justify the player’s experience, even without any compelling narrative. Tower Defense games should prioritize quality over quantity, meaning they'll likely result in a short play time. Therefore, it's important to design around maximizing replay value to keep repeat playthroughs fun and engaging.

These are the principles that we've focused on with Siege of Centauri, and I think we've done an awesome job. But we'd love to hear from you! What do you think, have we excelled or fallen short in any of these areas?