Mancha! by Lakso

Hello, my name is Lakso and this is MANCHA!, the classic game of tag.
This is my first entry to a Ludum Dare.
In Mancha! you'll be playing tag against an AI, in a constantly shrinking field.

Try not to be taged until the time is over.

As this was my first game, I tried to keep it simple while fun. The game was made with Löve and edited with atom. The sound effect where made with sfxr.
for some reason the source code isn't apearing on the links, so Here it is.
| Windows | https://lakso.itch.io/mancha |
| Source code | https://github.com/Laks0/Mancha |
| Original URL | https://ldjam.com/events/ludum-dare/42/mancha |
Ratings
| Overall | 671th | 2.694⭐ | 20🧑⚖️ |
| Fun | 660th | 2.5⭐ | 20🧑⚖️ |
| Innovation | 685th | 2.289⭐ | 21🧑⚖️ |
| Theme | 452th | 3.342⭐ | 21🧑⚖️ |
| Graphics | 584th | 2.658⭐ | 21🧑⚖️ |
| Audio | 374th | 2.842⭐ | 21🧑⚖️ |
| Given | 12🗳️ | 16🗨️ |
With a gamepad plugged in, I was forced into a gamepad+mouse situation, where I could only control the game with the gamepad but could only control the menus with a mouse.
I started the game with a gamepad plugged in (as it usually is) and was surprised to see WASD and the arrow keys not working. I got to just a few seconds left on the clock when I picked up the gamepad and started moving around, but I was able to win that game regardless.
And that's my biggest problem with the design of this game. Because of the shrinking level, you will always be right next to the AI at the end of the game (they tend to stick to one edge and not move much). So all you need to do to win is wait 29 seconds, tag, and hope you're lucky enough to move away. Anything else is a waste of time, as far as the game's concerned. I would have much preferred a king-of-the-hill system, where you track how much time each player has spent being "it". Maybe make the score increase faster towards the end of the game to add tension. Then whoever got more points being "it" loses. That would lend some significance to the bulk of the game.
The other problem I had is that the AI had insanely quick reflexes. You do get a little bit of knockback which means it is possible to start running before the AI reaches you, but more often then not I'd get tagged back immediately. I wonder about starting the two players further apart after a tag, or giving a half second of invincibility.
For your first entry this is very good. You kept the scope low, and you put a lot of effort into one aspect of the game, while making sure the rest was clean and polished. And it is fun and satisfying to play. I just wish the game wouldn't invalidate a game where I spent the first 20 seconds running away only to be sniped at the last second.
It would be better if the AI will escape from me in last few seconds.
I found a trick. I just need wait 29 seconds and give the block ring to it in the last second.
I do like the AI's behaviour, it's reactive to yours, but somewhat delayed too. Feels pretty natural for something so simple.
Last thing, I think you should have given users control over the window's position. I was trying to stream your game, and it firmly lodged itself in my non streamed screen, lol. :)
I think we can change the winning condition to something like total time of holding the hat is smaller, so the player will have a motivation to give out that hat ASAP.