[The Accountant] validator 27

What is so special with this validator ?

I have submitted several solutions which pass all IDE validators and fail only validator 27.

This is very frustrating since I see no way to reproduce the issue in the IDE … and my basic solution scores about 7000 points (100%) while my new solutions peak at 15685 (96%) …

Note that I don’t use any random number generator in my implementation, it is only based on the inputs.

1 Like

As a matter of fact, I also have a problem with this one … but I’ve never passed it … is there any way to know what is wrong with the method or the algorithm or the … idk ?

1 Like

I’ve just checked about this expert rule I had forgotten, but implementing it did not change the result.

If several data points tie as the closest to an enemy, that enemy will aim for the data point with the smallest id.

I’ve also double checked the rounding rules (floor for positions, round for damage).

I am also having the same problem with validator 27. No way to know what went wrong (timeout, death or no points). The difference is that I cannot even build a static basic solution that passes the 27th test.

Only times I passed it was when I was using random values.

1 Like

[quote=“mdaviot, post:3, topic:2079, full:true”]…I’ve also double checked the rounding rules (floor for positions, round for damage).

So the damage is rounded, not floored? Does that mean that the damage is rounded after everything in the formula is calculated (meaning that the distance part of the damage equation is not rounded or floored)?


This validation test is very annoying. I pass every single test with a very high score (16000+), and the physic engine is 100% accurate (no difference between the game I predict and the one given for each turn). I’ve doubled checked every rule, so there’s no way I forgot something, even the border collisions.
In case of inevitable end of the game, I always avoid being killed if I have the choice.

I have absolutely no idea what is going wrong with that test 27, it ruins every submission I do. We would be very pleased if you could give us any hint :slight_smile:

(I think other big players, as rOut, are also having the same issue)


"It is only with the heart that one can see rightly; what is essential is invisible to the eye.”

Please do not give hints on this.:yum:

Is it possible 27 has 100 agents and 100 boxes, and your “good” 16K solver takes too long, while your simple 7K solver doesn’t over think it…

1 Like

I thought so for a while, but now I’ve added a check to stop computing after 90ms and it’s not better. And in the IDE tests I’m using <10ms with my most complex logic.

It could be possible that the player starts on a data point that is 500 units away from an enemy. First thing that happens is the enemy moves - moving to 0 distance from you. If you are pre-calculating the move to determine damage then your calculation for damage could be dividing by 0 which throws an error and you automatically lose. That is just a thought. Just starting to get into this competition late and have yet to code anything. I have been just mulling things over in my head for a day before I start coding.

No, I just checked it. It is not about division by zero

I think I solved mine, validator 27 has dozens of data/enemies. A pretty big number, my problem was that I was computing how many enemies were after each data point. Which of course with these big numbers (and an inefficient function) made me timeout all the time.
I built a basic survival solution without that computation and it passed.

yes, that worked for me too. Too many enemies

Thanks for the insight :wink:

Thanks ! The rules say that there can be up to 100 enemies or data nodes, even if the IDE tests contain no more than 15-20 of each.
I probably had some O(n²) algorithm somewhere which behaves badly in this case even when I consider only the first move.

Something I also double checked before submitting again is the bounds of the area … you cannot flee the armored guys outside of the box !

Now I’m back with a good score (>16000) :smiley:

My simplest solution passed validator 27, but it does not involve any movement of Wolff at all (that’s why it is so simple).

The rules say 100 enemies, with health up to 150

right …

1 ≤ dataCount < 100

1 ≤ enemyCount < 100

1 ≤ enemyLife < 150

I’m having nightmares about 9th validator. Any hints whats going on there?