Knight Squad AI Update

There has been a lot of work done on the AI since the last update because it was a major feature needed before showing off the game to the public. We knew we would not always have 8 players playtesting at the same time so we needed the game to be fun even with only 1 person testing it. The first big test session was at the GDC and it went incredibly well. Most of the times there was around 3 to 4 persons playing at the same time and most of them couldn’t tell which player was human and which one was an bot. Playing a capture the flag game in a team with 3 other persons against 4 bots was one of my favorite moments because we really had to work as a team to win on harder difficulties.

Adjusting the different difficulties was one of my main concerns because we wanted the game to be as much enjoyable as possible for all kind of gamers and even for non-gamers. To my surprise I had to dumb down a lot the AI on the easiest difficulties but it confirms that playtesting the game often with people outside of the development team is really important because we are so used to play the game that it always seems too easy. I knew I had done a good balancing job when I was watching my family having fun on the easiest difficulty and then watching my gaming buddies being challenged on the hardest one.

Here are the main parameters modified depending on the difficulties:

-Reaction time
Longer reaction time makes the game easier because the bots are slow to react to the player’s actions.

-Pathfinding and turning speed
Adding some randomness in the pathfinding result and slowing down the turning speed will help on the easiest difficulties because the bots will have a harder time moving through the levels.

-Strategy, goals and decisions
The bots on harder difficulty have more goal choices and they take more complex decisions. Ex: when being attacked from behind, an easy bot will ignore you but on hard he might turn around and attack you.

Here is a video of the updated AI playing a capture the flag game:


Pathfinding added to First Time Knight!

Hey guys!

I have a first working version of the AI for First Time Knight which means having some basic pathfinding and a simple goal (killing the closest enemy in our case). The pathfinding comes from a library I’ve found on the asset store ( With this library I had the choice to enable the diagonals or not. In my case I wanted the diagonals but only when there were no walls corner to corner so I had to make some modifications to get the desired result.


Another optimization I made to the paths is to add long diagonals. It feels more natural when the bots are in a big open area.


The result is pretty neat as you can see in this video:



Happy New Year!

Happy new year to you all!

The development of BeatBlasters III is going very well and its release should be approaching quickly! Here’s the first trailer of the game :

Let us know what you think about it!

What’s coming up next for Chainsawesome Games in 2014? We are currently working really hard on polishing BeatBlasters to make it the greatest experience possible. We will then slowly start development on the full version of our First Time Knight prototype which is an 8 players competitive top-down game.

This is pretty much what is coming next in the short term! Stay tuned for more news!


First Time Knight video

Finally! A video of the game we created during the 48h game jam named Pixel Challenge in Québec City. We won first place in the professional category with this game.
Let us know what you think about it and if you think we should continue building it!


Chainsawesome Games wins the 2013 Pixel Challenge

WOO! We just won the 2013 Québec Pixel Challenge with our 8 players fast paced action game. The challenge was quite an experience for the team! Congratulations to all participants for creating such awesome games, shorts and concept art during these intense 48 hours. We will put online a video of the gameplay shortly so stay tuned!