Poor user experience and gamefeel ruin a game - GMTK2022

GMTK 2022's theme was roll of the dice. I had gone into this gamejam wanting to make a puzzle game, so I came up with the idea of a tile-based game where depending on which side of the die was up, the level would change in some way. I originally had thought about playing with height differences and such, but ultimately I chose three major mechanics during the jam to implement. Only one really matched my original vision, being lazers, but I was still happy with the end state of the puzzles and completeness of the game.

 Each puzzle object in the level could have a  a value or a logic rule to whether it was active or not. Practically this led to lazers being on or off when certain numbers of your die were face up, such as all faces but 1, or only 1, for example. You may have to land on the goal with a specific face, or go through a series of lazers that only deactivate on specific numbers requiring you to precisely orient your die.

The other two mechanics I implemented were other dice that moved when your main die moved, and pressure plates that had to be activated (sometimes with a logic rule) to spawn the goal in for the stage. These extra dice could also activate pressure plates, could block your movement, and could intercept lazers creating a buffer for you to move your die past them.

Overall the 12 levels I had implemented explored the few mechanics I had decently as well as their interactions.  I had put a lot of time into this gamejam, and was confident I could score well for creativity as well as enjoyment, even if I wasn't confident in the visuals. I had an undo button that worked even with all of the complex mechanics, and could rewind the gamestate to any previous state of the level perfectly and play the animations. There were debug controls to go between all the levels, and also reset the current level.

I rated various similar die rolling games after the jam ended to scout the competition, and this only boosted my confidence. Many games were ambiguously isometric with confusing controls, no undo feature, let you fall off the level and forced you to reset even after long puzzles, and in general created a bad gameplay experience. Very rarely I did find a game that I found to truly be creative and enjoyable. Two of my favorites from this jam were Machine of the die by Tyrant, and RollyPolly by foxtrot-roger (a couple sokoban style games I liked a lot are also on this list https://itch.io/c/2645689/gtmk2022-favorites).  Machine of the die allowed you to dash forward over gaps with the number indicated by your die's top face, and created some decent levels with disappearing platforms, and lazers. RollyPolly introduced a variety of mechanics, including the core mechanic of having two characters to control a die, not only could you roll the die, but you could also push it without changing the face, getting rid of many of the frustrating situations in other game where you are one or two tiles off with the correct face up.

However, when I received the rating of my game it was a good amount lower than I expected, for what I thought to be my best gamejam game to date. Its rating isn't necessarily bad, but compared to similar games with less inspired puzzles or fewer mechanics (or even aggravating gameplay), it often scored lower.

What went wrong - from reviews and personal opinion:

1. Camera shake
What I thought to be a simple addition to add some visual flare, and maybe my presentation score, ended up causing players a lot of frustration. Each time the player moved, the camera would adjust by one degree to make the level look like it tilts allowing the die to roll. I've never been someone to feel sick from playing games so the thought never crossed my mind that it may have this sort of effect on players.  (By the way I did this with lerp and a [SerializeField]-ed AnimationCurve in case anyone else wants to make their players feel unwell) 

2. Poor choice of colors
I had my monitor set to boost digital vibrance (75%) through nvidia control panel since it helped when playing Apex Legends to visually distinguish enemies, so the resulting game ended up looking drab

3. Levels may have been too difficult for gamejam rating
I suspect another contributing aspect to poor ratings was the difficulty curve. I'm not really someone who will want to make my games easier to my detriment (I suspect my mobile game https://zanthous.itch.io/trail-light also suffered from this problem to some degree). In particular, the level called "particular" requires the player to get an exact orientation correct for the die to roll through a series of lazers. To compensate for this, I give the player a very wide area to work with. If you can get the face of the die facing the camera to be a 4 on the line before the lazers, you can roll sideways until your orientation is exactly correct. In combination with the bad feel of the game, the extra frustration could end up with players just quitting the game, giving a rating and moving on to rate the next. (The faster you rate games, the more karma you get, and the more ratings your game will get). The issue with this system is that players quit out before seeing all the mechanics the game has to offer, so you have to hook them all quickly in order for them to see all your game has to offer.

4. Borderline too slow movement
I should have toned up the tickduration slightly. I believe it was at 0.25f or so for the gamejam version, meaning for moves a second, I have cut that down to 0.15f now.

5. Debug controls not working for some keyboard layouts
Some players could not reset the level, or skip to the previous or next level because they had non QWERTY keyboards. Mostly just unfortunate, but could be bypassed with UI (which I decided to not waste time on), or adding some backup keys only bound to letters.

6. Visuals/presentation and gamefeel/enjoyment are very closely related
An unfortunate reality of the situation is that raters are not perfect, and will not be able to objectively separate all of the categories. A games with better visuals will get more points for other categories than an exact clone that has worse visuals. I heard an example of this lately, that was in Super Meat Boy, you unlock an alternate character that is functionally the exact same as the main character and is only a visual override, but players would often give feedback that the character felt much better, or was much easier to play with.

Looking back, creativity was my worst category, almost dipping into 2.X.  To me this says many players dropped off very early. Presentation somehow was higher rated, which doesn't make any sense to me. In the end a bittersweet gamejam for me, I was personally very happy with my entry, but it ended up doing much worse that I was aiming for.  I tried to fix various aspects that were wrong with the original version in the post-jam version. You can find a post about the changes I made here, and a couple notes on their implementation: https://zanthous.itch.io/minddie/devlog/410028/post-jam-graphical-overhaul-and-q...