Game Design #2: Listening to Feedback (And when to ignore it!)

I’ve learnt that one of the biggest challenges with game design is learning to process feedback. By putting your designs through rigorous testing, you’ll often get a lot of high quality feedback from helpful testers. And sometimes, added to the feedback are suggestions on how to fix the problems that were identified.

This is a great thing, but getting a ton of feedback comes with its own challenges. The most difficult of which is deciding when to listen to said feedback, and more importantly, when to ignore them. And sometimes, while a point of criticism has merit, it might be a symptom of a larger problem that testers might not notice (and rightly so, given that they approach the game from a very different perspective!)

Last year, I passed out a prototype of Endogenesis to a friend for him to perform playtests. One feedback I received was that in certain cases, the game would last for far longer than expected, and end up feeling like a drag. This came as a surprise, as reducing game length was one of the earliest issues I had successfully tackled in early versions… or so I thought!

Winning in Endogenesis

To go on further, I’ll need to explain the general flow of the game.

  • Over the course of the game, cards are drawn and revealed from a 36 card deck called the Realm of Chaos.
  • Among other things, these cards include Monsters and Legendary Monsters, the latter of which is important as slaying them will give the player a Prism (Those hexagonal things above).
  • The goal of the game is to be the first player to collect a certain number of Prisms. (3 for 3-4 players, 2 for 5 player games.)

The Realm of Chaos deck contains monsters, legendary monsters, event cards and distortion cards.


The specific issue being raised was that for the extraordinarily long games, the Realm of Chaos would sometimes be depleted and therefore would need to be replenished with the discards. This would lead to an exceptionally long and draggy experience, as players potentially have to go through many normal Monsters before encountering a Legendary one, when all it takes to end the game is just one more Legendary to be slain.

What was peculiar was that in previous versions of the game, this case had never turned up before. Was this a rare problem that had always been there, but was only uncovered recently, or perhaps there was something new with this version was causing it?

Suggestions on how to fix the issue

Along with the feedback came suggestions on how to fix the issue. Here were a few:

  • Introduce a new alternate victory condition: If the Realm of Chaos deck is depleted, the player with the most Prisms is considered the winner. If multiple players are tied with number of Prisms, then use another form of player resource as a way of tabulating score instead.
  • After the Realm of Chaos is reshuffled once, all monsters within are considered to be Legendaries and will now award Prisms, so that the game speed accelerates.
  • Make the Realm of Chaos a “killswitch” — if the deck is depleted, the game ends and no one wins. (yikes)

As you can see, these ideas generally revolved around adding new rules, or exceptions to existing ones. This is a solution I was generally adverse to, because in most cases, making a rule change just to patch a corner case feels like a clunky band-aid than an actual fix. Here’s an excellent line from an article on elegant design by Chris Anderson:

“If you are writing your rules and see yourself using the word “except”, you may just be wasting people’s time.”

Having made many changes to the game over the past two years, it was difficult to remember the rationale behind every decision I made. Thankfully, I had kept careful notes on the more significant changes. Upon referring to them, I’m reminded that in very early stages of the game’s development, I decided that there would be a total of 9 Legendary monsters in the Realm of Chaos for a very specific reason. This was because 9 Legendaries was the minimum number of monsters required to remove all cases where the game could extend past the exhaustion of the Realm of Chaos deck:

  • For games with 3 players, the longest possible game would see two players get 2 prisms and 1 victorious player getting 3 Prisms. (2/2/3) This meant that 3 player games would see 7 Legendaries slain at a maximum.
  • For games with 4 players, the distribution of Prisms would be 2/2/2/3 at most, so that’s a total of 9 Legendaries.
  • And since 5 player games have a modified goal of 2 Prisms, the maximum distribution would be 1/1/1/1/2 (7 total)

In other words, having 9 Legendary Monsters should’ve been enough to prevent the re-shuffling of the Realm of Chaos!

I then turned my attention to the changes made in the latest version, and was finally able to isolate the problem: introduction of these two new cards:

Smite: Instantly destroy a monster and receive its rewards, but not its Prism (if any).
Hallucination: Used when a monster perishes: Its carcass transforms to that of a harmless sheep’s, offering its killer no rewards.

These are Wonder cards, and they come from another deck called the Realm of Wonder. They contain powerful abilities that can be activated instantly from hand. Because of their potentially game-changing nature, players are awarded Wonder cards very sparingly, with 1-2 cards per player on average per game.

Smite and Hallucination were new additions to the Realm of Wonder. When either are used on a Legendary monster, the monster would be destroyed and discarded without a Prism being awarded to any player, therefore causing the potential number of Prisms awarded in one cycle of the Realm of Chaos deck to be reduced from 9 to 7. In other words, while it would occur rarely, under the right conditions these cards could cause the Realm of Chaos to be depleted without a declared winner!

With the problem isolated, I could easily fix the problem by removing these cards. However, as I did like the cards’ effects and felt that they added to the gameplay, I decided instead on a small tweak. Here’s what they look like with the new changes:

Smite: Instantly destroy a monster and receive its rewards, but not its Prism (if any). Shuffle the monster back into the Realm of Chaos.
Hallucination: Used when a monster perishes: Its carcass transforms to that of a harmless sheep’s, offering its killer no rewards. Shuffle the monster back into the Realm of Chaos.

An additional effect was added, causing any monster that falls victim to either cards to be shuffled back into the Realm of Chaos. With the new change, the balance was now restored! Legendary Monsters will never be discarded unless slayed the proper way, with a Prism awarded to a player. No new rules or exception had to be introduced as well.

So this was one instance where I had to be careful when dealing with feedback. As game designers, we have unique insights and perspectives into what makes your game work. But we also have unique blind spots and biases, that can only be revealed through the feedback gleaned from playtesters. And that’s why processing feedback can be such a tricky affair —  being able to make the right call is what will require a lot of experience.

If you’ve gotten this far, I’d like to thank you for reading! Here’s a great article on handling feedback by Eric Jome that has helped me tremendously: How To Listen – A Designers Guide to Feedback During Testing


  1. An interesting read! I wanted to write a more in depth response, but I can’t see my text as I’m writing it. Your text color is set to black and your comment box is set to black, so I’d have to highlight what iw rote to see it. I see a red line, so apologies if something looks weird.

    Anyway I think it’s really cool how you came to the conclusion that you did. I usually break my game, then rebuild it with less parts to see what works. Then slowly add until I either fixed the initial issue or figured out why that issue occurred in the first place.

    • davidgoh

      Oh darn, thanks for pointing that out! Problem with customizing themes extensively is that one tends to miss out the smaller details. It should be fixed by now!

      I’ve been keeping up to date on your process, and I’ve noticed that you have no reservations making huge changes to your game when you feel that it’s warranted, which is a really good thing. A lot of designers get hung up over keeping certain things, and end up being resistant to change… It ends up being a mental block that keeps a designer stuck in mediocrity. Thank you for sharing your thoughts!

Leave a Reply to Fertessa Scott Cancel reply

Your email address will not be published. Required fields are marked *