Pixel Miner: Designing for Addiction

Pixel Miner: Designing for Addiction

In the polite society of games addiction is so loaded a term we like to use synonyms such as “sticky”. Semantics aside, I designed Pixel Miner, like any of my games, to be addictive. I wanted to have players entranced not for any mean-spirited reason, but to prove that the games I make are effective in their job of having players want to continue to play them. Because games that don’t hold players are failures, whilst games that have players returning are triumphs.

And in the case of Pixel Miner, I felt I was rather successful in my design. For the month of August 2014 Pixel Miner had the following “rolling retention” (see notes) figures of:

  • Day 1: 73%
  • Day 7: 59%
  • Day 30: 30%

What this means is that of the cohort of players who install on a day (day 0) that 73% came back the next day or later, 59% were still around a week or beyond, 30% after a month. What’s most remarkable is that at the point in time this data comes from the game only had 28 days of content which some were racing through in 10 days and others were never reaching because of a bug. So clearly Pixel Miner is doing something right. But what?

Pixel Miner is an incremental and, like many incremental games, plays upon two very deep rooted human drivers: Greed and curiosity. The game has the player mine pixels as a tiny miner, whose earned pixels can then be spent on upgrades (both equipment and speed multipliers) which allows the player to acquire pixels at a quicker rate.

As both the acquired pixels and the rate of acquisition change the player’s sense of greed is piqued, but so too is their sense of curiosity as new equipment becomes purchasable for them. In Pixel Miner’s core loop the two senses essentially feed each other creating a strong compulsion. But in having the player gated by time you are constantly delaying their gratification, having them project their satisfaction on a future time and come back repeatedly.

However, to effectively ensure the player’s constant return there were a few things that were essential to be right: Pacing and visual progress.

Pacing

In order to have the player understand why they would want to upgrade their miner we made them do it right away in a very short tutorial. They go from mining slowly with their hands to tripling their speed and using tools in their first session. This sense of both visual and mechanical progression is designed to reward the player rapidly early on and is strengthened by a few more equipment and multiplier upgrades that were minutes apart in the first few play sessions.

At this point the player is hopefully somewhat compelled by the process of improvement because each upgrade then becomes further and further paced out in time until the player is having around one upgrade per day. In essence the player is hooked on the thrill of upgrading, then upgrades become bigger and further apart.

However, Pixel Miner was designed so that the player could have a hundred five second sessions per day, so what are the remaining ninety nine sessions filled with if not upgrading? This granularity was provided by some secondary small mechanics: Boosting and treasure. Whilst I won’t go in to too much detail here, the mechanics gave the player meaningful actions that could be completed quickly and let the player dip in repeatedly.

Indeed, so effective were these secondary mechanics that something totally unexpected happened: Players began requesting that we add the time to the game. This request puzzled me as Pixel Miner was designed to have such small sessions that players surely wouldn’t miss the time for five seconds. It transpires that players were starting these session so frequently that they just left the game running. Pixel Miner became the default use for their Pebble, but they still wanted to tell the time.

Pixel Miner Balancing Spreadsheet

Some of the numbers behind Pixel Miner.

Visual Progress

The final piece of the addiction equation is curiosity through the equipment upgrades and, to a lesser extent, the unlocking of treasures. These unlockable items were designed firstly to have a range in scale and grandeur whilst also having a sense of humour. The miner starts mining with his hands and ends up on the back of a dragon, via an excavator, mutant mole and various machines from cult TV and film. As the player was teased with a small amount of information on the next and future upgrades their curiosity was triggered as to what the equipment would be and how it would look.

A great deal of time and effort was put in to making sure that the equipment and their animations felt interesting and exciting so that the players would want to continue committing to unlock them. If each equipment didn’t feel better, whilst also being mechanically better, than the last equipment then the illusion of progress was broken and the player would feel cheated out of their time and dedication. Whilst undoubtedly there were some flat spots, overall the equipment progress was a key driver in retaining players.

Conclusions

Pixel Miner didn’t do anything essentially new as its mechanics have existed in incremental games and RPGs for years, but it did package them in a new way that was right for a new platform.  Designing the game also proved out a few things for me. Firstly that having very short, high frequency sessions on a smartwatch is the right way to design for the platform, secondly that building a compelling game can be done in part in the science of a spreadsheet (i.e. balance and pacing) but that the real magic comes from the fun and playful content around it.

Notes: There are two ways to measure retention. One is counting players from a cohort who start a session on a given day and the other, as we use here, is counting players from a cohort who start a session on a given day or later. The former is a more pessimistic number and often more common, but unfortunately I only have access to the latter.