[WK 12] Conclusion
Playtest Feedback:
[Two experienced TFT players and two players unfamiliar with the autochess genre]
- Players familiar with TFT as well as one of the unfamiliar players were able to grasp which units were more tanky vs damage dealing vs support based.
- The bow and arrow symbol and range stat on the champion inspector was a bit ambiguous so users were sometimes unsure if a unit was more of a back liner or more of a front liner.
- Players became more familiar with traits where the VFX were more visible i.e. Toy / Working / Pastoral / Hound. Terrier and Gun units seemed to be picked less often and were least likely to be mentioned when players would speak aloud.
- The fact that the combine system averaged stats out conflicted with the appearance of the dog getting bigger, which suggested the unit was getting stronger. These units would often be killed early on by Hound units especially if their traits weren't activated. This gave the impression that the unit was weaker so players would be discouraged from using the system.
- In general players felt the preperation time was too quick so it didn't allow them to properly strategise and wanted more information on the systems. Particularly players wanted to be able to see the stats of the units they were buying before purchasing them so they could make informed strategic decisions.
- Few players seemed to form proper strategies around the traits. Only one player did and that player already had context of the prototype / experience playing the game. One player did stumble across a strategy that worked well which was using the terriers, toys and hounds together. Terrier units formed a strong front line in combination with the toy's providing support whilst the hound units were effective at quickly taking out the enemies backline.
- I was initially worried that hounds would be slightly OP because the trait would become more powerful the more hound units the player had on the board. Although hound units were highly effective they were only worthwhile having in combination with a strong front line.
Further changes:
1. I made it so when players hovered over the shop frames it would display their stats on the right so players can make more informed decisions about what their buying.
2. I increased the preperation phase timer from 16 > 20 so players have longer to formulate a strategy. This might still need to be longer but I don't want to make it too long because experienced players will get bored with how long it takes.
3. I made the bonus UI frames on the left change colour to green when there's enough of the units on the board for the bonus to be activated. Hopefully this should draw better attention to the traits.
4. I also made the champion limit UI at the top change colour between green and red when there is and isn't space on the board to place new units. This should give player's beter feedback when they buy a level or they try and place a unit when the board is full.
5. On the range stat in the champion inspector I added 3 variations of text either the unit is a backliner midliner or frontliner. Hopefully this should help players to position units better instead of the number that it was before which was amiguous since it didn't relate to the grid placement at all.
6. I also changed the combining system so that instead of averaging the stats out across the two units it takes the stats of the second unit halves them then adds them to the first units stats. The fact that units can only be combined once means that units can't become infinetly stronger. It also means that now combining them feels more impactful.
Conclusion
What I learned from the project
I learnt a variety of skills whilst doing this project. For Unity specifically I learnt about the fact you can add components to animation behaviours to invoke functions after animations have finished. I learnt about how I can create input events from UI interactions, which is something I've struggled to figure out before.
I had to learn how to read code not made by me to be able to adapt my own code / systems to the processes and methods already in place for more specialised functions. Specifically this relates to when I was making the traits, I had to learn how to create different bonus' and where to call these methods.
I also had to find ways to improve the pre-existing code to allow for new functions to work. This relates to the way bonus' were read and activated previously because an active bonus would be applied to the whole team. I changed this so that bonus would only be applied to the units with that trait so I could create more specialised and identifiable traits.
I spent time creating balance sheets as I iterated on the traits system to make sure there weren't any dominant strategies in place. This was really useful because I had to analyse what was happening in the game and copare it to the model I'd created. In some cases I found that the model I'd created wasn't very effective so I changed it. In some other cases I realised the model needed to be adjusted / that I needed to interpret the results differently.
What I'd do differently or improve
I wasted a lot of time at the beginning on a error that kept popping up. It ended up being a problem with the editor, which was causing a bit of an issue but ultimatley wasn't a fatal issue. I probably should've spoken to some of the techs about it to get it resolved quicker or just moved on from the issue.
In general I had a time plan but after the first half it started to fall apart slightly since it took me longer to create the traits than I'd expected and it didn't take me very long to make the combine system. I should've accounted for contingency timing so that it would've allowed me more time to do further balancing / playtesting or to create more visuals and feedback for the player to better understand the systems.
It was difficult for me to playtest since I only knew one other person who played TFT. Towards the end I was approached by some other fans of TFT who then helped me by playtesting. It was difficult because on the one hand the focus of the project wasn't about tutorialisation however in order to playtest I needed players who understand the game. Looking back now I could've done two things though. I could've gathered a list of players, who are familiar with TFT through flyer advertisment online and in the GA so I had a group of playtesters I could keep testing with. The other thing I could've done was rather than create functioning tutorialisation I could've made a print out which explained very simply what to do.