Thank you , you have found my mistake. 
1 Like
Just try to scroll in output windowā¦ the blocks of debug messages looks almost similar, thatās why it looks like you see only the last one.
how can i get to Boss3 after i got the 1st versus Boss2
how do I get it to go to the first checkpoint?
Quick question: is the number of checkpoints constant?
Nope, you have an input variable for the count:
Iām Using python on bronze, and I donāt see that variable. Does that only appear later on?
Hm, Iām not sure, I remember itās there from the beginning.
No, itās not. It comes with the promotion to Gold League.
Does the number of checkpoints still vary? I mean before Gold.
The number of checkpoints in one round remains the same, but the number of checkpoints can be different in two different rounds.
Thatās what I meant thanks.
hi!
I want to do simulations to predict trajectories and to infer podās decisions based on game physics.
I find it rather hard to infer formula from game data, even after some simple experiments
(accelerating from zero speed to the maximum speed along the X-axis).
Am I missing something? How to do such things in a right way?
For example:
-
max speed is 277 for throttle=50 and 561 for throttle=100, but 277*2=554, not 561
-
a pod needs 27 moves to reach the max speed with throttle=50, but it needs 32 moves for throttle=100ā¦ I wonder is it possible to predict speed at each next step when we start to accelerate having initial speed V with some throttle=T? (Except that experiment with every possible throttle 1,2,3,4,ā¦,99,100 and saving the results in tables.)
In other words, is there a sane way to āconvertā known current speed and known (chosen) throttle into the corresponding value of speed on my next move?
Any ideas which variant of formula I could try? I tried different ones and no one fits so farā¦ Iām stuck at this moment.
Hey !
Depend of which league you are in. The definitive rules and inputs are unlocked in the Gold one.
Thereās some physic involved in the calculus, more about the collisions than about the speed, but a collision can infer in speedā¦
I donāt know how you made your tests, but be careful about the viewer. It is nice, but not really preciseā¦ A gap can also exist in the display of the stderr logsā¦
If you already are in Gold league here is an article about CSB simulation: http://files.magusgeek.com/csb/csb_en.html
1 Like
Iām not yet in Gold league, Iām in Silver. Maybe I should first focus on moving to Gold, and then to start with simulations.
Thank you for the link and I will be careful with viewer, that makes sense.
1 Like
When you go on to do a simulation you might want to use robostac brutal tester to verify its working fine https://github.com/robostac/coders-strike-back-referee 
2 Likes
My bot like, circles the checkpoint then touches it until it moves on which is really weird. Any idea what I did wrong?
You can use it to force a collision (for push them away from the checkpoint for example) or prevent a collision (by adjusting your trajectory, possibly producing a curved trajectory to your goal). I have seen a few good examples of doing this.
Unfortunately, a LOT changes in the GOLD league. This makes for a huge learning curve once you make gold: just an FYI.
Yah the same thing
If I try to implement a GA in python, am I going to have trouble getting it to run enough moves ahead for it to be useful (because itās python)?