• Sarah

Playtests 3 & 4

Updated: Jun 4, 2019

Playtest 3

This playtest tested the interaction development from the first level and creation of the tutorial and level 2. I used the music sent to us by Sam, along with pre-recorded audio of myself saying the lines for the piece. I played this to a group of students throughout different year groups to once again test whether the interaction was quick to pick up. They learnt the commands quickly which was demonstrates the interaction model working at a better performance. I then went on to playtest the game along with Adam and Venny. This identified 3 problems: confusion over who the players are talking to, how many ingredients can be used in a spell and whether it mattered what order the cards go in. With this I developed the game to be nameless, the only time Alexa's name is mentioned is to invoke the "cast spell" interaction as this can be used at any moment during the game and Alexa must awoken with that key name. I clarified that only 2 ingredients could be used in a spell, however in development we decided that this detail would be something best integrated in with the instruction booklet.

Playtest 4

Before recording the audio for the game we wanted to conduct a large final playtest to assess if there were in bugs with the speech output.

We conducted 4 run throughs of the game which helped us identify an average running time of half an hour for the game.

Overall the interaction ran smoothly, however particularly in level 2 there was some difficulty establishing what there is to investigate. This included the roots and hollows (tester said tree), the meadow and the undergrowth. The confusing wording has been changed so the area is not described as a whole, so that the individual objects can easily be identified. As well as this in the scene setter for the level 3 cave it was not obvious that the players needed to look for ingredients again. I adapted this so that this was understood, in the final interaction model.

There was a specific placement of the ingredient, finger of gargoyle in the fridge, that did not make sense and was therefore moved to the tutorial to make more sense. This brought about the problem of logic association with the ingredients for the spells as well. From this I have adapted spell combinations to be as reflective of the spell type as they can, however this could be quite subjective.

Another problem that occurred was the forgetting of what had been investigated/looked at. In development from this we had two options here, repeating the options (this would make the interaction lengthy) or encourage note taking to remember what had been interacted with. That final option worked really well, as we had already produced Enchanted Kin notebook merchandise which can be simply included in the end components of the game. On top of this I adapted the initial scene setter from the interaction model to remind players that their notebooks can help them remember what they have already done.

Interaction wise there was also no instruction of how the cards should be laid out or held. This saw a play tester put the card to hold in the cauldron. This therefore needed to be resolved as this has a different function. This was resolved with a setup instruction in the instruction manual, along with the introduction of physical inventory made by Rubi to store cards. This interaction is also now fully demonstrated in the instruction booklet.

The players also did not know what to do with the successful spell ingredient cards to discard them in the game. This led to the development of placing back in layout and turning on its back. This is now shown in the instruction booklet.

Another thing that has cropped up is the understanding of the looking system. Players have assumed that they move during the game and therefore looking right one after another will return different directions. However this is a looking system not a turning one. I have decided to establish this in the instruction booklet where a diagram can demonstrate this interaction.

A major bug that cropped up was the final level being un-completable for players who immediately cast spells and were successful on their first try. This left them with no cards left to play and no way of going back. There must be a way for these players to collect cards and take part. A work around to this was developed. The narrator confirms the players card numbers at the beginning of the final level, if they do not have enough for this (6 card minimum) the narrator will quickly find some more and instruct the player to pick them up.

It was also suggested that the forward combination of ingredients for the first level to be not-successful to promote players to explore the level further. To prevent these events occurring, I have adapted the tutorial to include a tip to find as many ingredients as you can before casting a spell, as well as adding this as a tip to the instruction booklet.

Play-testing the final model (non-build version)

After recording the narrative with the voice actress we used the voice clips in a final play test. This established two bad bugs: the tutorial not flowing right (telling the player to investigate a nonexistent object) and the use of the wrong word in the final level. This was particularly bad as the line indicated the success of the game, however the recording stated that "the dragon took out the dragon". This indicated something very different. To over come these problems we re-recorded the four lines.

Additionally we carried out a final play test with a child at the lower end of the target market alongside his younger sister. The boy had to be reminded of the options, which is where the help function in the build will come in helpful. However the younger player did not understand the interaction of looking, picking up and the casting a spell, This demonstrated that our game is appropriate for the target age range of 8-12 and not any younger.