EN24 discord
sov map

Leaks: AAA Reaction to the loss of HED-GP

January 20, 2014

The following post was originally posted on the Against all Authorities forums:

Default HED-GP After Action Report
tl;dr – We lost HED, had great form up and really put our best foot forward here. We have a long war ahead of us but losing due to CCP is not a lost at all. Lets keep going, tighten up our defenses and fix issues before they bite us again.

So, we lost HED-GP. For those that were there, thank you for enduring an 8? hour trip through CCP Server and Programming Hell.

It was in every sense of the word a nightmare, however there are from where I sit an huge amount of positive things that came out of that fight.

I wanted to break down what happened, what our game plan is going forward, the positive things we are doing well and the issues we need to improve on.

The fight

Warfare in EVE is not unlike any other analogous you can think of in war. There is strategic planning, tactical implementation, unit and formation consideration and the list goes on.

HED was not the ideal place for such a fight. While a station system, its not of any great blow to our logistical chain to empire, it has historical but not military nor supply chain importance… but we wanted to defend it. And frankly I was extremely impressed with our numbers. At peak, we had 250 -A- + SCDOT (didn’t get an exact number) logged in to various fleets from sub-caps to titans. A great turn out.

Today’s fight in HED-GP we formed in our coalition:
1000 Sub-Capitals
700 Dreads
220 + Supers
100 + Titans

N3/PL formed:
350 Carriers
116 Supers
30 Titans
100? Sup-capitals

In order to make sure they had the system and their formation locked down in place, they formed super early. Hours and hours before the timer they were set up on the IHUB in HED-GP. In our belief that server issues of years gone by had been fixed and addressed, this seemed trivial and a waste of their time. In reality it became the reason for their victory, much to my surprise. Ill touch on this soon.

We arrived 1st with our sub-capitals and immediately began engaging both their small number of sub-caps and also a mobile jammer they put in place to prevent our capitals from cynoing in close. Due to max TiDi this alone took at least 30 mins, probably closer to an hour.

Once the jammer was down, our next move was to bridge in all Dread fleets, nearly on top of them and go to work on their carriers and the limited amount of dreads they brought in as the fight began.

However, fate and CCP’s limited server capabilities (for a fight of this magnitude) were not with us. Almost immediately I began hearing reports that dread pilots couldn’t online modules, were stuck in “warp tunnels” or just plain black screened trying to enter the system. Seeing the system at max TiDi before the dreads entered I guess I wasnt entirely surprised, I was fully expecting the server to go down.

However that didn’t happen. Server stability improvements did not translate to actual service to connecting clients.

Back years and years ago before TiDi the server worked on a 1st come, 1st serve basis. In reality what this meant was he who was on grid 1st, got the servers full attention and those who came in 2nd got nothing. TiDi was, I thought at least, the solution. Coupled with slowing down server time itself (1 second turns into 10 seconds in how long the server has to respond), I also had the impression that each client connecting to the server would be given equal (and random) chance to access system resources. What actually happened though makes it apparent to me that the issues of old are still very present in the server response. If you are on grid 1st, you are getting more server attention than those who come in last.

In short, our Dreads jumped into system and were left not only defenseless but unable to apply any damage.

Seeing this unfold we decided not to send any supers or titans on field, expecting them to run into the same issue. In another blow to our senses, the only thing that appeared TO be working was Titan DoomsDay effects, probably due to them already being on grid.

Our dreads couldn’t do any damage, they couldn’t turn on their tank and our sub-capitals were having a nearly impossible time to lock and shoot. The subcap fleet I was in with my main was only able to kill approx 4 sub-caps after the dreads entered, in a span of 6 hours.

What could we have done better?

– Well, now I know that in these huge timers we are either on field 1st or the fight has a very high chance not to go in our favor. 5 years ago this was a no brainer, but we fooled into thinking that the changes to server performance and programming eliminated these short comings.

– Dreads *need* to be dual DD tanked.

– We have to make sure we don’t lose POS’s in system like what happened in HED. To our surprise our allies and us lost 9 POS’s (out of 11 moons) before I caught the issue. Some took their POS’s down, others lost them to attacks. Either way not having a jammer POS to prevent the “Slow-Cat I-Win” button from entering cost us.

Going Forward

– Replace these dreads and aim for Naglafars. Just because the server couldn’t handle the load doesnt make this fleet ineffective. Alpha has and always will be king, plus we need to step up the number of dreads we can field regardless. This was a full SRP op so make sure that you request SRP for them.

– Keep up the great numbers. We can always improve, ideally 10% CTA participation on a regular basis and seeing 20% on major CTA’s from corps.

– Shore up our defenses. We need to stay on top of POS alerts, get them passed up from corps to the alliance leadership ASAP. We are working out a better automated solution, but always manually let us know.

– Pray that CCP find a solution to these issues. I’ve been playing for a long time and frankly the server situation today is no better then it was 7 years ago, except now we can have double the amount of players who can be disappointed.

Questions inbound.

Was there any benefit to our excessive sub capital force in HED being dominixes, or perhaps more reachingly, overwhelming sub capital strength whatsoever? A fleet we know adds to lag exponentially due to drone doctrines and the server stress they bring? Was there benefit to having 10x their subcap numbers and ballooning HED to its server max of just over 4000 characters?

Why weren’t dreads cynoed into a deep safe and warped in? Even if we’d moved half the subcaps in we’d have had easy superiority to clear tackle and was there not historic precedent (PL in Y-N or CVA v. AAA 2009) to know that such a move would be disastrous? Essentially why commit to an act that has only ended in failure?

Is there any clear benefit to dominixes over maelstroms or any other non drone based long range doctrines during this type of battle? Drones assisted to bliss were rarely firing and it seems likely individual pilots may have more success in a doctrine not requiring so much from the server in an already stressed environment.

Are our allies still committed to this fight? Oftentimes such disastrous defeats see the losers go into a demoralized slump and fairweather friends bugger off. Essentially, was this our 49-U6U?

Answers
– Yes, there is a ton of benefits to having superior sub-cap numbers. You control the battlefield with sub-caps and also provide protection for capitals. No, we shouldn’t based our doctrines on what the server can handle as that is entirely outside of control and never a static number or situation.

– In hindsight we could have cynoed in our capitals at a safe 1st and warped them in. I wasnt running a dread fleet at the time, but I think I would have cynoed them in at zero before knowing now how the server reacted.

– The reasons for the Domi fleet are many. Ill be straight up.

1 – We believe Drone assist is a broken mechanic and having 1 person controlling the damage output of 249 others, directly, is not in the best interests of EVE. Its the coalition’s goal to (ab)use this mechanic so that CCP wakes up and fixes it… Maybe they will fix carriers having endless amounts of sentries on hand, while battleships are limited to at best 3 flights.

2 – A single person can control the damage output of 249 others. No misfires, misclicks, killmail whoring etc. That is pretty powerful.

3 – Domi’s put out 2.1k Alpha every 4 secs vs 6k alpha every 14 seconds for Maelstroms. I am a huge Maelstrom fan and due to the sentry drone’s weakness to bombs I am partial to the doctrine, but with current game mechanics the Domi is pretty powerful. However as N3 found out, kill the drone assignee and its game over.

4 – I can’t speak for everyone, but myself personally I am even more motivated. We had a legitimate opportunity for victory snatched away by an issue way outside of our ability to mitigate or prepare for. If the server was fine and we died in a fire there would be need for serious introspection, but we are currently unable to tell if an Alpha Dread + Domi doctrine is as effective as we hope it will be.

Feel free to discuss this in the comment section below!