I’m wondering about one of the issues I have with the current test harness, which is:
The lead car might always win.
So I have 3 suggestions
A) Probably the best, have a special boost to use every x turns when you are behind.
B) we could see a form of slipstream, where one car speeds up more when its behind another
C) Mud can have for example a 50% chance of breaking when one car passes through it.
So that the back car potentially has an “easier” route.
I cannot really speak about the engine too much, I’m only starting my code this weekend. But I have run a lot of simulations (Mental ones) to get the perfect solution.
It is a given that the bot that does get ahead first will likely be the best bot anyways, But I do feel it could help to even the playing field slightly. I don’t know for sure though. And its based off of “the behaviour of the perfect solution”.
Looking forward to a great year.
This challenge is always a highlight for me.
I think the root of the problem is that it isn’t that difficult to follow the optimal route (i.e. avoid mud). So that doesn’t leave much space for strategy/tactics. I have some ideas:
What if there was another aspect to the race that you had to tradeoff with following the optimal route - perhaps obstacles that damage your car, but provide a boost? Then you could have some interesting decisions.
Another idea that will add some strategy - customised cars? You get 10 points to put into various categories like handling, top speed, acceleration, health etc. At the start of the race both bots provide their setup to the engine. If you don’t provide anything, your car has a default setup.
I have a feeling this will be coming in future tournaments anyway, but: weapons. Then there is racing but also attacking/defending against the opponent. This could add a whole new dimension to the game.
It might be good to put a cap on the number of powerups you can carry - this would prevent the front guy from scooping up all the powerups
Thanks again entelect team for organising a fun competition every year!
All points are good, but I especially love the sound of Point 2…
Having say upgrade points as the round progresses that we can use to adapt our car while we play.
A side note regarding point 4. Picked up power-ups does not get used / disappear. So theres no unfairness here in the sense that the front guy eats all powerups on the best routes.
With my current version I can see that there are some interesting complexities to strategy even now. And the optimum route for both players does differ considerably in some places.
Example:
If I am on x4 and my enemy is on x 9, and we are both driving at a constant 9, our routes could be different, I noticed this in my tests where some paths would have completely changed if I was “2 tiles back”.
And I know Entelect has a way of throwing a curveball at us after example the first event, which keeps things fun.
Hi Thread, we have brought this up as a team and agree with what you are saying, we are working to bring updates to the game engine and add in some new aspects to the race.
Yes thats correct. Any “bot-breaking”/significant changes or new abilities that bots can do are released ASAP after a tournament, this gives everyone a fair amount of time to tune in their bot strategies to new game rules.
Of course some of the well hidden bugs only show up thanks to playtesting, and we try to fix those early in the release timeline to keep a consistent playing field until the tournament runs.
