As The Dust Settles: Western Cape [PS: EVIDENCE OF WAR BUG]

Day 369, 13:13 • Published in South Africa South Korea • by Watcher

As you might have noticed, there was a little to-do in the Western Cape on account of a Minipax contract that just got over with. Although the client was ultimately unsuccessful, we are proud to have played our role in fulfilling the local coordination of the war. I don't have time for an extended analysis, but the short version is that the client preferred to use its own forces to provide the majority of the damage in the fight, and those forces failed to provide sufficient swing to counter what was admittedly a greater than expected Indonesian resistance.

Sorry for lack of legends on these, I'm pulling them right off the coals for you:
Damage Graph: http://tinyurl.com/6edmx4
User Count Graph: http://tinyurl.com/5cndw6

Red for Resistance, Blue for Indo - the scale for the damage is 0 to 14500ish, the scale for the users is 0 to 200ish, x-axis is time in 10-minute intervals over about a 24-hr period. Notice that initially the client's forces were rather effective, delivering 200 distinct users and combined attacks totalling over 14000 off the wall in just a 10-minute span, but shortly thereafter the resource chain could not support the continued efforts and apparently failed to allow further sustained attack, at which point the Indonesians all woke up and began bringing their own troops in.

I may add in further details as the day goes on, and I may discuss questions from the comments if they pique my interest.

POSTSCRIPT:

Turns out the ATLANTIS allegations of unfairness in the battle system have some merit. Not only is it broken, it is broken in a way that is biased towards the defending side in any given battle. Look at these by-player numbers here:

# butnaru - 19350 Damage (17632/171😎
# RupeTot - 13324 Damage (12503/821)
# Bomberclad - 12521 Damage (11774/747)
# nemeiah - 12257 Damage (11233/1024)
# Juroichi - 10223 Damage (0/10223)
# blink-az - 9771 Damage (0/9771)
# all-x - 9719 Damage (0/9719)
# Misho - 9294 Damage (9142/152)
# arcadia - 8963 Damage (8137/826)
# flik_kenni - 8477 Damage (0/8477)
# Pier - 8349 Damage (7742/607)
# Lonestar - 8101 Damage (7978/123)
# rikwandi - 7156 Damage (0/7156)
# Bima - 6515 Damage (0/6515)
# Ryuhou - 6118 Damage (0/611😎
# garpheet - 6070 Damage (0/6070)
# Prince Rizkyputra - 5328 Damage (0/532😎
# bisiacco - 5181 Damage (0/5181)
# sandygee - 3964 Damage (0/3964)
# zzZZzz - 3926 Damage (3775/151)

Those are the top 20 by-player damage rankings for the last battle, excluding the initial indonesian surge. By-side damage is split by (Resistance/Indonesia) in terms of which side the person was fighting for. Notice that while the Indonesian damage counts are all (0/##😵, the resistance have about 7-9% of their total logged for the indonesians. In just the top four listings, that reduces the effective damage done from 57,452 damage to 48,832 - almost a 10,000 point difference, just in the top four fighters. This is significant, and I hope we can see admin action on this flaw.

As a side note, I've seen this occurring to a lesser degree in other battles, always to the detriment of the offense. In the second resistance in the Cape, Durruti has done fully 5% of his damage for the Indonesian side. So, don't deride the claims of a broken system. I won't say whether it would have made the difference here, but it would have been interesting to see.

EDIT:

Here's an independent report confirming the existence of a bias:

Article (ENG)

In the interest of full disclosure, it should be noted that in talking with eFRA I've found and corrected a bug in my own data-gathering procedure. My damage numbers during the most active portions of the fight are inflated quite a bit due to the speed at which the fight was progressing, so I encourage all people doing secondary analysis to use the numbers for eFRA for their work. The bug has been fixed, and all future Minipax publications will use the new routine which corrects the error. Sorry about all that. 😃