Blog

Hello everyone!

Once again, this article is directly written by Inateno, for the sake of simplicity.

At the beginning of June, we launched a major call to our beta testers to help us with a first playtest session.

And the first word that comes to mind is “wow,” followed by “Thank you!”

Indeed, “wow” because I was impressed by your participation and willingness to help us.

And truly, thank you because this session has taught us a lot!

It helped us find many bugs, of course, but above all, it helped us understand the strengths and weaknesses of the game. To be honest with you, we were afraid to launch these playtests! We didn’t know how you would receive the game, whether you would like it or not.

And overall, among the 49 responses we received on the form, you all loved the experience and what the game offered. 

Of course, nothing is perfect, and thanks to your feedback and our analysis system, we managed to identify the little hiccups that bother us.

This article is dedicated to the playtest debrief. We will explain “how it works” and show you some graphs!

PS:

• If you don’t want to spoil anything from the game, I invite you not to read this article as we will go into details, content in the biomes, and map names.
• If you couldn’t participate in this first playtest session but would like to participate in the second, don’t worry, you can read this article without fear.

As a reminder, the purpose of a playtest is not just to find bugs (although it comes with it, it is not the central objective).

In our case, the playtest had several objectives:

• To know what you think of the game as it is (graphics/gameplay).
• To check if you were getting lost too much in the world or not.
• To have an overview of the balancing and difficulty.
• To see what things are not necessarily understood or not obvious, in order to correct/improve all of that.

Summary of responses to the playtest session form.

At the end of the playtest, we sent a form to gather answers to certain questions. It was, of course, anonymous and not mandatory.

Here is a summary of what the players thought, based on a selection of questions.

•  94% considered the game “neither too difficult nor too easy, “4% found it “too difficult” and 2% found it “too easy.”
•  67% didn’t feel lost but weren’t sure where to go, 20% knew exactly where they were and where to go, 12% felt “a little lost,” and 0% felt completely lost!
•  96% of people think they had “the required amount of lives”, 2% found it too many, and 2% found it insufficient.
• Players mostly loved the platforming sequences.
• Regarding the first mini-boss, the overall feedback is good. As expected, players had “difficulty reading the patterns” and “felt destabilized,” which are relevant insights, and a few people found it too difficult.

Among the best memories of the playtest session, two standout points are:

• The giant worm
• The demon that chases us and can’t be killed but activates things for us.
• Some players particularly enjoyed the trampoline mechanic.
• Lastly, other players specifically liked the level design/platforming and the game’s fluidity and movement.

Among the “worst memories”:

• the villainous creature that takes the medal is, of course, the terrible sausage on legs, or “Dark Frogs” as nicknamed by one of the players (our technical name is “SpiderSwarm”)
• Next, the trap monster (that burrows into the ground) was quite frustrating because players didn’t know how to approach it. However, we have made improvements to it and hope it will be better now!

• On a scale of 0 to 5, we have 63% with a rating of 4, 30% with a rating of 5, and 6% with a rating of 3.
• As for whether they would like to participate in another playtest, it’s a resounding yes at 92%, while the rest prefer to wait for the beta.

There’s no need to go into detail about most of these responses, but some points need to be addressed in more detail.

Let’s start with the first mini-boss of the game.

Its technical name is “Elite Spider,” and it grants you the “WallJump” ability!

It’s not too difficult; in fact, it has relatively low health.

The monsters in the “spider” family have an unusual behavior in video games, and almost all players who participated in the playtest were disturbed by these creatures’ patterns.

We are divided on the changes to be made to these creatures because they are original and break the conventions of AIs we know in metroidvanias. However, they gave you quite a challenge.

For now, we have decided to reduce their numbers on certain problematic maps.

The “dark frog” wins the award for the annoying mob, as it is a common monster found throughout biome 2. However, this creature is very fragile, and 1 to 2 sword strikes are enough to defeat it (depending on your attack boost).

But it’s the same issue I explained earlier; we wanted unusual monsters that make you think before charging in.

Analysis of the session data

Let’s move on to a more in-depth analysis.

It’s necessary to explain our tool a bit before proceeding, as not everyone may be familiar with how our playtest was conducted.

We created a fairly simple tool that generates a unique one-time code.

Through a form on a “pre-launch” page, participants had to enter the code, and then the game would load.

Once the game was launched, it was connected to a data server, and the game would send information about what each player was doing.

We didn’t collect a ton of information because it wasn’t necessarily immediately useful, but we had the following data:

• Player changes map (destination map with “time,” allowing us to measure time spent per map)
• Player takes damage (with the ID of the entity that inflicted the damage)
• Player dies (also with the ID)

Based on this data, we were able to generate the following graphs:

PS: The names you see on the X-axis are “technical” and temporary names. We didn’t necessarily have names for the creatures when we were working on the game, and several of them need to be renamed!

Time spent by map

In red, we have the maximum; in green, the minimum; in orange, the average; and in purple, the median.

As we can see from this first graph, a significant amount of time was spent on maps b2_01 and b2_02.

This is normal since these maps are quite large and introduce new mechanics and monsters.

The number of deaths/time spent was particularly high on these maps, considering they are relatively early in the game.

Therefore, we have decided to make modifications to these maps to make them less punishing and facilitate learning.

To measure the impact of our modifications, we calculated the number of deaths per map per day.

This provides a more meaningful perspective.

Deaths by map, day 1

The first day of playtesting was busy in terms of session numbers, as if you all wanted to be the “first” ones, haha.

In any case, we can observe a sharp spike in deaths on maps b2_01 and b2_02, followed by a second spike on maps b2_06 and b2_07.

The second spike is not as problematic; b2_06 is a rather technical map where players are being chased, and b2_07 is the map with the first mini-boss.

Therefore, we didn’t make major modifications to these maps, just some fixes.

However, as mentioned earlier, we made modifications to b2_01 and b2_02.

Deaths by map, day 2

On Day 2, we can observe a slight decrease in deaths on b2_01, but it’s not sufficient for b2_02.

There was a sequence of monsters/spikes that was quite challenging, and it was easy to fall into an “infernal spiral” and lose all of your health very quickly.

Therefore, we continued to make modifications to that specific map.

Deaths by map, day 3

On this graph, we can observe a positive trend with a significant decrease in the average and median number of deaths on the problematic maps.

We continued to make some adjustments here and there, as you can see from the overall graph progression.

Day 3 had the lowest number of playtests, so we had fewer data points, and the analysis continues on Days 4 and 5 to confirm that the modifications are effective.

Deaths by map, day 4

The downward trend is confirmed on Day 4.

Deaths by map, day 5

Once again, the trend is confirmed on the 5th day, so we have validated the modifications made to those maps.

Naturally, if we look at the “maximum” number, we might be inclined to be influenced, which is why it is important to look at the average and median curves as well! We also analyze the “min” and “max” curves because in some cases, they can be revealing.

Why is it so important / Which curve is considered “good”?

You might be wondering why we are putting so much effort into reducing the number of deaths on these particular maps.

As you know, I like to explain what happens behind the scenes in video game development.

Having a high number of deaths on a map can potentially lead to player frustration and even rage quitting, especially when the map is at the beginning of the game and the player hasn’t fully engaged with it yet. It’s easy to think, “This game sucks,” and quit. However, as players become more immersed in the game, invested in the story or other elements, their tolerance for persistence increases. Similarly, when the stakes seem higher, death becomes more acceptable.

I can summarize this with the following statement: “Dying to an ant is frustrating, but dying to the first boss is expected.”

And that’s what this curve represents.

Our mistake as game designers was trying to introduce too many new elements at the beginning of Biome 2. As players, you weren’t familiar with these monsters, and having too many of them at once, with tight placements and new behaviors, led to a higher-than-desired number of deaths.

Working on these maps was challenging because they were well-designed from the start, but not impossible to improve.

Additionally, I believe it will be necessary to add one or two maps that precede these challenging ones, allowing more time for fun and gradual discovery. Some players enjoy the hardcore aspect, but personally, I prefer a softer introduction, especially since a few maps later, the difficulty spike returns. However, the game is worth it, and the challenge is impressive enough not to frustrate players with repetitive deaths.

Total damage per enemy

This graph isn’t particularly meaningful because the total number of deaths or damage changes depending on the frequency at which an object/monster appears.

However, I find the numbers and graphs amusing, so we’re sharing these graphs with you.

Total deaths per enemy

What are the dates for the upcoming sessions/beta?

We know that we want to have more sessions, but we’re not sure about the exact dates yet! I think we’ll have another session at the end of July, and maybe one last session in August!

The closed beta is expected to arrive in September/October! (we’ll do our best 💪).

Until then, we’ll be working hard to continue the game production, and we’ll keep you updated on the dates.

The final name reveal is scheduled for Wednesday, July 12th at 8 PM (French time) in a live stream on our Twitch channel!

We’ve been waiting for this moment for a while. Join us on our Twitch channel on July 11th. We’ll discuss the game, have a little Q&A session, and finally unveil the new game trailer along with the name reveal!

We hope to see you there 😊

Our Twitch channel is located at: https://www.twitch.tv/dreamirl

And if you haven’t already, we invite you to join the official Discord server, a place for discussion and chat about the game.

PS: please check your private message on Kickstarter / your email, as we sent a message to each “beta” backer.