Prototype Phases For A Hit Casual Game – Purple Diver

Here at Voodoo Games we recently launched Purple Diver, which we developed with Viatcheslav Tarasov (the developer behind Ragdoll Archer, Truck Traffic Control, and Snake Loves numbers). And we’re pleased to say it quickly jumped to the top of the charts, now having around 15M downloads worldwide.

So in this post, we’re going to talk you through how the game came into being and what we learned from making it. (If you’re new to hyper-casual games, make sure to check out our 3 snackability tips for making hit games).

Why we love Purple Diver

For us, one of the best things about Purple Diver is its monetization potential – in a game like this we can show a lot of ads. We know this can be irritating for some players, but we think the nature of the game and the depth of the gameplay keeps people engaged and their annoyance levels to a minimum. The ads don’t cost us too much to get either.

With that said, let’s look at the different iterations Purple Diver went through before we settled on a final version.

Version 1 – inspiration and initial numbers

The game first came to us around December 2018. Viatcheslav had always wanted to make a jumping game, and was inspired to put this into action by the success of Flip Trickster, published by Lion Studios. Obviously they’re very different visually, especially once you go into the challenges, but it definitely played a part in the creation of Purple Diver.

The good

In its first form, Purple Diver wasn’t that different visually to the current version. In fact, it was actually more complex, and quite a bit harder to play. Having said that, we were actually pretty pleased with the difficulty level in version 1.

If you were to pick up the first prototype and start playing, chances are you’re going to be hitting the edge of the pool quite a lot. So it was pretty tough to move on to the next level. But our players got more skilled pretty quickly, and the level of difficulty made them keen to master it. We compared it to games like Ball Blast – yes, they’re frustrating, but they’re skill-based games which make you feel great when you finally do beat them.

Version 1 also had some promising stats. Day 1 we got around 45% retention, and day 7 was around 15%, so we knew that this game was definitely worth pursuing.

The bad

The cost per install (CPI) for version 1 was a bit challenging for us. It was between $0.30 and 0.40. That’s not bad, but there was definitely room for improvement.

We also had a fairly complicated user interface (UI). For example, the end-of-level animation had a big menu with a bunch of information. It was hard to absorb and not particularly useful. So we knew we had to simplify the rules and rewards in the game to keep players interested.

The onboarding (i.e. teaching players the game’s mechanics) in version 1 also wasn’t quite what it is today. There was no effective tutorial. And while in the main we were happy with the difficulty of the game, the first few levels were harder than they needed to be. For example, if you hit the red rings you’d die straight away, which is no longer the case.

In short, here’s what we learned and needed to improve:

• day 1 and 7 retention was strong, so we knew we had a good game on our hands, 
• the difficulty of the game was pretty spot on (if not just a tad too difficult),
• our CPI wasn’t great, and needed some improvement,
• and we needed better UI – the game wasn’t intuitive enough and overloaded the users too quickly.

Version 2 – simplifying the gameplay

The main thing we wanted to do in the second iteration of Purple Diver was to make things more straightforward. We gave each world a single goal – so whether that was hitting the duck, or diving really deep, you only had one thing to do and you knew what it was.

We also simplified the UI by removing the complicated ratings from the end screen. This made it easier for players to focus on a single goal. And finally, we added a tutorial. The key performance indicators (KPIs – metrics that measure the performance of games) showed us that these were definitely improvements. We were now at around 50% on day 1 and around 18% on day 7.

Unfortunately, the CPI was still too high on this version of Purple Diver. So even though we had a great game with great retention, it was still just too expensive to publish. That’s when we decided to get more teams involved from across Voodoo. We worked with our marketing team to simultaneously carry out more iterations. That meant we could work on the game more actively and with bigger budgets, and using creatives that looked for lower CPIs by varying the visuals and the gameplay.

Version 3 – using what we’d learned   

On version 3 we applied a few basic lessons we’d learned from other games. So while we were working on getting a better CPI with the marketing team, we also looked at things like making the game more forgiving and rewarding. A few things we tested out were:

• A version where the rings didn’t kill you,
• where you didn’t need to reach a level’s goal to make it to the next one,
• where collectibles were available and where there were easier touchdowns at the bottom of the pool, and more.

And we didn’t like it at all. In fact, it actually hurt the performance of the game.

Version 3 taught us that making difficult games like this easier i.e. games where the player’s skill determines the outcome (so chance isn’t a big part) and where you’re able to upscale through time, can actually end up hurting retention. Which reminded us that there’s no simple, one-size-fits-all answer to game design.

Version 4 – getting it right

By the time we got to version 4, our marketing team had done a great job at reducing the CPI – from $0.30 to 0.40 all the way down to our launch KPIs.

So we then decided to go back to our best version, which was version 2. Obviously, we made some tweaks to improve the balancing and on-boarding and added some haptics. And it was that, along with the CPI improvements, that meant we could launch what today is one of our top-performing games.

So, what did we learn?

There are a few things we took from developing and launching Purple Diver that we’ll be able to apply elsewhere.

• We can get solid CPIs for new games even if they use ideas that are already on the market. We’re not talking about copies here, even though those are prevalent in the top charts (we’d never ask our studios to copy someone else). But modeling and improving something that’s already on the market can be a great source of new ideas – as long as you pivot the game plays.
• Sometimes you have to ignore the basic rule of making gameplay too forgiving or rewarding – it doesn’t apply to all games. It can be potentially more important to get onboarding and upskilling right, especially if you’re happy you have an effective, if difficult, game.
• Launching a game before reaching the KPIs can work. This and involving more teams is part of an initiative we’re taking to publish more games, and get them out earlier. So while we might not see a huge payout on day 1, we can at least cover our and the developer’s costs. And we can start paying our developers before we reach the impressive metrics we usually get on published games. That doesn’t mean that iterations and coaching come at a later stage though. We still want to work with developers and studios on games that are getting 40% on day 1 and 10% on day 7, and with a CPI of below $0.50. But it does mean that when we do get to those metrics, we want to use our capabilities to carry out A/B testing and publish faster, then get to the KPIs.

Want to talk to us?

If seeing what we did with Purple Diver has inspired you to complete your own game, feel free to show us what you’re working on. We’re always keen to help game developers get across the finish line. And if you fancy some more reading, have a read through Voodoo diversified and lowered game product KPIs.