[BUG] Coders Strike Back input difference

Feedback & Bugs
#1

Hi,

I was trying to implement collision in Coders Strike Back and I found what looks like an error in the input of the game.
Both player does not seem to have the exact same input at the same turn.

Here is an exmple of one turn.

Input for player 1:
Position: [6225,3538] Speed: [-325,-63] Angle: 132
Position: [6706,4605] Speed: [-463,203] Angle: 122
Position: [5723,2451] Speed: [-68,-255] Angle: 22
Position: [7668,4336] Speed: [-281,31] Angle: 142

Output of player 1:
8000 4500 100
8000 4500 100

Input for player 2:
Position: [5723,2451] Speed: [-68,-255] Angle: 22
Position: [7668,4336] Speed: [-281,31] Angle: 142
Position: [6225,3538] Speed: [-325,-63] Angle: 114
Position: [6706,4605] Speed: [-463,203] Angle: 104

Output of player 2:
8000 4500 100
8000 4500 100

We see that the agles gtiven to player 2 for the pods of player 1 are different and according to their current state (position, target, angle) it looks like player 1’s pods have already rotated but didn’t move yet.

This situation is even more obvious on the first turn where player 1 see all pods with an angle of -1 while player 2 only see its own pods with an angle of -1.

Does anyone already encounter such error? Or is it perfectly normal (even if it is not written in the rule of the game)?

Thanks :slight_smile:

#2

Hmm… I’ve never noticed, nor have I ever heard of such an error, but I can confirm that I am seeing pretty much the same thing:

Turn 1

14313 8130 0 0 -1 1
14859 7292 0 0 -1 1
13766 8967 0 0 -1 1
15406 6455 0 0 -1 1

13766 8967 0 0 -1 1
15406 6455 0 0 -1 1
14313 8130 0 0 219 1
14859 7292 0 0 207 1

Turn 2

13809 7720 -428 -348 219 1
14281 6995 -491 -252 207 1
13353 8465 -351 -426 231 1
14781 6278 -531 -150 196 1

13353 8465 -351 -426 231 1
14781 6278 -531 -150 196 1
13809 7720 -428 -348 205 1
14281 6995 -491 -252 195 1

Turn 3

13147 7346 -591 -294 205 1
13629 6617 -544 -358 195 1
12714 8134 -653 -94 214 1
14194 5915 -369 -482 185 1

12714 8134 -653 -94 214 1
14194 5915 -369 -482 185 1
13147 7346 -591 -294 223 1
13629 6617 -544 -358 193 1

If you take this into account, then you could have a distinct advantage as player 2 (preview of enemy’s move) or a distinct disadvantage if you assume that this is not happening.

  • danBhentschel
#3

Unfortunately it seems that I didn’t registered the case as a JUnit, but I’m sure that I saw the exact same thing.
I was playing against myself and logging inputs at the very beginning of each turn, and for one turn one angle was different between the two inputs.

When I get bored with FB I may try to reproduce it :slight_smile:

Edit : and my game simulation’s result was one of the two
Edit2 : I witnessed similar issues in FB, I may post them later on.

#4

Can confirm,

Player 1 view:
10119 3516 544 -170 349 1
10645 4472 486 -191 316 1
9545 2917 360 -276 355 1
10881 5417 387 -313 306 1

Player 2 view:
9545 2917 360 -276 355 1
10881 5417 387 -313 306 1
10119 3516 544 -170 339 1
10645 4472 486 -191 312 1

#5

actually it was reported already here:

Still in our bug backlog though, sorry :confused:

#6

I’m not sure it’s the same thing. If you read through the discussion the difference in their case is 1 degree and boils down to rounding of inputs. Here the difference is 10 degrees.

#7

That’s the same bug. Player 2 recieve Player 1 angle after rotation as input. The post can be confusing (because of the -1/+1 angle) but it is the same bug.