This case study follows the format established as part of my research into virtual pet design, which will culminate in the design of my own virtual pet.

Petz originated as a desktop virtual pet for the Windows PC. Designed by Night Trap designer Rob Fulop, the first title in the series, Dogz, was released in 1995. Originally published by PF. Magic and billed as ‘the original virtual pet,’ Petz was largely overshadowed by the release of the Tamagotchi the following year, and was later acquired by Ubisoft and reintroduced in a form that resembles little of the 1995 original. Even so, the game had a strong following at the time and is fondly remembered by enthusiasts today.

This case study concerns the original Dogz and Catz games, rather than the modern Ubisoft-published titles.

The illusion of life.

muttzThe creatures in Petz are designed to simulate the behaviour of real pets as much as possible with the technology of the time. Their purpose was to be believable; according to Furlop in an interview with The Mary Sue, ‘people really felt like their virtual pet was “real” — it wasn’t a “make believe” puppy — it was a real puppy who happened to live on their computer screen.’

Petz are fully animated, based on real-left animal breeds, each with distinct personalities and unpredictable behaviour. By looking at this promotional video released by PF. Magic in 1998, we can see the length the designers went to in order to deliver believable virtual agents:

The video shows how important it was for the designers that the Petz were both appealing and believable. The creatures are described as ‘highly expressive, believable synthetic agents with rich personalities,’ which develop ‘evolving social relationships’ with their users and with other on-screen Petz.

The entire system is designed around this idea of believability; from the way that the process of choosing a pet is presented as an adoption agency, to the way that the Petz grow over time, both visually and in the complexity of their interactions with the user.

Petz have ‘unpredictable’ and ‘autonomous’ natures, able display a ‘wide range of emotions and behaviours’ and ‘to act independently of their users, fostering the idea that they have their own will and personality. All of this ‘allows users to come up with their own interpretation of their pet’s feelings and thoughts.’

Fostering a relationship between the user and their virtual pet is a core part of the design philosophy. The idea is that users should feel rewarded for developing strong relationships with their Petz; these rewards take the form of improved interactions from the virtual creatures. Pets can be trained to do tricks and perform other actions using a reward and discipline system, creating a lasting interactive representation of the user’s investment in the virtual relationship.

kittygifThis level of believability is achieved by combining ‘animation, improvised drama, artificial intelligence and games technology.’ This multidisciplinary approach echoes the idea of ‘believability’ proposed by Joseph Bates, and lead to characters with a remarkable degree of expression and interactivity considering home computer technology in 1995. Certainly, the fidelity of interactivity and expression is far beyond that of Tamagotchi, which launched a year later.

Interaction design.

Petz is designed with a ‘natural and direct’ interface, intended to be as instinctive as possible for a piece of software tied to PC mouse and keyboard input. Input is all handled by the mouse; players can interact directly with their Petz by clicking and dragging to pet them or pick them up, or they can pick up and use toys and other environmental items, each with different mouse-bound functions.

The cursor takes the shape of a hand, capable of visual gestures to show available actions, and is implied to have a physical presence in the game world; the Petz will follow the cursor with their eyes, and interact with it if it gets close. Users can double-click in empty space to signal their pet to ‘come here’; whether or not the pet responds depends on their relationship with the player.

Performing actions with the Petz involves interacting with toys and other environmental object. These are moved and interacted with in the same way as the Petz themselves, and their functions are usually a matter of simple gestures or proximity to the Petz. For example, holding a brush and ‘rubbing’ it over a pet with the cursor triggers a ‘brushing’ interaction, while holding a chew toy in front of their face will cause them to grab it.

Petz are capable of performing a large range of actions this way; importantly for their implied personalities, they are able to interact with these object of their own accord, implying ‘favourite’ toys and actions. These self-defined behaviours can be encouraged or discouraged by the user through a reward and discipline system.

‘Rewarding’ a pet by giving it a treat or petting it tells it an action it has performed is desirable, and make it more likely to repeat that action. ‘Disciplining’ a pet with a squirt from a water bottle, on the other hand, tells it that action is bad and makes it less likely to perform it again. This system can be employed alongside the provision of special treats to teach Petz tricks like flips and rolls.

Repeated over time, this system of reward and discipline reinforces the idea that each pet has its own personality, and that this personality can be moulded by interactions with the player. This reinforcement learning is the mechanical underpinning of the rewarding relationship the designers were seeking to build between pet and player.

Original Dogz breeds.

Original Dogz breeds.

Original Catz breedz.

Original Catz breedz.

Petz are available in various breeds, and each is programmed to behave differently. In later versions players were able to breed their pets and create cross-breeds, but in the original version they selected from one of 15 unique breeds. The idea that each breed had its own behavioural traits is a simple programming trick, but effective reinforces the idea that these are unique animals with real personalities.

Social factors.

While undoubtedly successful, Petz never reached the kinds of fevered success of Tamagotchi or Furby. The original series (Petz 1-5) enjoyed a healthy following throughout its lifespan, but it never became the ‘must have’ toy of any Christmas season. Petz sold 1.5 million copies in its original incarnation – good sales for a piece of PC software, but nothing like the alleged 27 million Furbies sold in a single 12 month period. To date the Petz series, including over 20 titles released by Ubisoft since 2004 that bear little resemblance to the original PF. Magic series, has sold around 22 million copies – a respectable figure still dwarfed by the 76 million Tamagotchis sold before 2010.

Much of this limited success, especially in the early days of the series, is likely to do with form factor. As a piece of software tied to desktop PCs, Petz faced an uncertain marketplace at launch in 1995. While home computer and internet usage was certainly growing at the time, digital saturation was a far cry from what we see today. The number of homes in America with a PC was around 54%, but the number of households whose children – a key audience for Petz and the driver of the Tamagotchi-style sales frenzy – had regular access to a PC was likely to be much lower.

Both Tamagotchi and Furby, on the other hand, were known quantities – boxed products, purchased from toy stores, which function independently of any other technology. These toys followed the marketing model established by the formidably successful Tickle Me Elmo; Petz had to follow the far more modest sales model of Microsoft’s Encarta Encyclopaedia.

It’s a situation that would play out very differently today, with Minecraft‘s 40 million monthly players20% of whom are children – dwarfing the expectations of even the most successful toy manufacturers.

Example message from a Petz fan site.

Example message from a Petz fan site.

And if we look at the history of Petz, we can see something like a precursor to Minecraft‘s success at community building and personal expression. This article by Jessica Famularm for The Mary Sue discusses the rise of a very involved Petz subculture, based around the games HTML customisation tools that shipped with Petz 3 and later. This scene, grown independently of the games themselves by fans, can be seen as a precursor to personalisation-driven platforms such as MySpace, Second Life (2003), Facebook and Minecraft (2011).

Modded breedz, like this winged cat, made for a vibrant pet sharing community.

Modded breedz, like this winged cat, made for a vibrant pet sharing community.

This long-tailed engagement through customisation and modding is an advantage unique to the PC platform, something that cannot be replicated in the physical toy market, which relies on constant new models and product redesigns to retain popularity.

The diversity of the fanbase for Petz is also worth noting – the fansites described above were moderated primarily by women. PF. Magic claimed in 1998 that Petz was ‘equally popular among girls and boys, kids and adults,’ a rare achievement at the time and notable even today in a tech market that remains skewed towards a young male demographic.

And the longevity of the fan scene surrounding Petz speaks to the designers’ success in creating a genuinely engaging virtual creatures. Dedicated fans are still playing with and conversing about the original Petz games to this day. Fulop writes:

‘About a year ago I was contacted by a woman, 23, who had actively kept her virtual dog for 15 years and was unwilling to upgrade operating systems because Windows XP did not support the original version of [Dogz] that she had since she was eight years old.’

This example, and many more like it in the dedicated Petz fan centres still active online, speak to the level of engagement that can be achieved through strong virtual pet design, and by providing tools that allow users to express themselves and build up communities around their virtual pets.