Conquer Club

Large attack

Found something broken on the website or in the game, that isn't a plugin enhancement issue? Let's check it out.

Moderator: Tech Team

Forum rules
Please read the Community Guidelines before posting.

And don't forget to search for previously reported bugs first!

Large attack

Postby mc05025 on Sun Mar 29, 2020 7:59 pm

Game 19752412

This is an example of a friend's game where the problem is obvious but I have observed in my previous games just much less pronounced.
In trench escalating games large attacks can easily happened so it is kind of important to follow the generic rules. In these big numbers the % declination from the normal result should be tiny while it isn't. Sometimes it is as high as small attacks' results

In this example orange attacked green in last turn doing an attack 521,816 vs 524,065 and resulted winning with 190,550 troops leftovers

The first and most important bug is that this is statisticaly impossible. Obviously some kind of shorcuts are adopted to handle large attacks but these should at least give meaningful reasults

The second bug is that the dice stats aren't being updated with these numbers. More specificaly for both players the attack was recorded as 65535 killed troops for both attacker and defender which is obviously the upper limit of the datatype you are storing the result (small positive integer for MySQL) and has nothing to do with reality. The overall battle outcomes are also not updated with the entire attack as it is obvious from happyfeet's stats. Actually it seems like it is updated by adding 65535 to both attackers loses and defenders loses which gives a fake extreme luck for defending troops (you can see the +2% at the last image)

Here is the moment before the attack. The moment after the attack is the last position you can see as the game finished immidiatly with resignation.

http://prntscr.com/rp3zdh

Here are the obviously wrong dice stats after the attack

http://prntscr.com/rp40mn


PS. In this game the error might be meaningless. But I have witness at my previous games same problems. For example at an attack with 8000 troops we killed 7950 which seems again statistically imposible but it is hard to prove it.
Also I have seen players with huge declination at the overall dice stats like -2% in thousends of attacks. The above problem seems to be the explanation

I will be adding all attacks with more than 2000 troops attackers I witness for more material

2800 troops killed 3500
User avatar
Field Marshal mc05025
 
Posts: 388
Joined: Sun Mar 14, 2010 3:09 pm
2

Return to Bug Reports

Who is online

Users browsing this forum: No registered users

cron