PDA

View Full Version : March 21 Kovrova Coast Event AAR



tyrspawn
03-21-2010, 07:41 PM
Not going to give a mission synopsis - mainly because I was unable to operate with effective Command and Control (C2) and never felt as if I was a platoon leader.

I think this mission revealed the instability and danger of using the current com system - it effectively makes TEAM/TACTICAL play with a group larger than 35-40 impossible. Coms broke down at all levels but the highest due to technical failures:

1. Individual - being unable to hear a guy 5 feet away from you who is being fired at and asking for help, unable to report that you are hit and need a medic
2. Fireteam - unable to direct team members or assets effectively to act as a unit rather than as scattered individuals
3. Inter-squad - each squad was unsupported ,as they were unable to talk to other elements, making play as a platoon impossible
4. Platoon - I was unable to give orders and once elements were set into motion corrections could not be made. This resulted in dozens of unnecessary casualties as situational awareness was extremely dampened.

Effectively each player was by himself, unable to rely on others, and even when there was intra-squad and fireteam cohesion, elements were unable to talk to eachother, resulting in individual unsupported elements going up against a seemingly coherent enemy resistance. This resulted in long delays, heavy casualties, low morale, frustration and consternation.

I also had very low FPS the entire time.

Other than that:
1. Higher's plan was logical
2. Radio SOP and discipline was implemented effectively where possible
3. Individual leadership skills were effective and exceptional
4. Admirable dedication by all players, and exceptional patience in not quitting when com stresses were becoming unbearable
5. The mission was excellent and I would love to play it again
5a. For those who say it was too difficult, I disagree, it's like a typical wac: it requires good C2 and attention to tactics, and without the ability to command AT ALL, it was a failure

Gill
03-21-2010, 08:38 PM
Keep in mind, I wasn't there. But... why were you not able to give orders? You set a command level short- and long-range radio frequency, right? There's also different slots you can program different frequencies for into your radio box that comes on-screen. Example:

Company
Long range: 1234
Short range: 123456

Platoon Leader
Plugs in Company Commander's short and long range
Long Range: 1123
Short Range: 112345

Squad Leader
Plugs in Platoon's short range and long-range
Short Range: 111234

Fireteam
Plugs in Squad Leader's short range
Makes their own little short-range.

Fireteam talks amongst themselves, can communicate via short-range to squad leader. Squad leader can talk to his fireteams and pass stuff up to platoon leader. Platoon leader talks to his squad leaders and can pass stuff up to company commander. Vice versa back down.

Does this create a lot of chatter for the company commander? Yes. That's what radio protocol is for. If you're talking on a higher command net, identify who you are and who your traffic is intended for. Protocol at the fireteam/squad level isn't really needed. Just say who it is and blabber away.

beita
03-21-2010, 08:40 PM
Plugin got broke'd :(

tyrspawn
03-21-2010, 08:41 PM
Plugin got broke'd :(.

Gill
03-21-2010, 08:56 PM
Haha... well, that answers that.

Falcon_262
03-21-2010, 08:57 PM
I'd like to play this mission again with working comms. In my opinion it is the most tactical, and most dynamic mission we've made for ARMA II yet. Heck I'd play this with one platoon. I think the ONLY failure in this mission was the complete breakdown of comms from the platoon level down which made the mission effectivly unplayable.

Toyguy
03-21-2010, 10:07 PM
I'd also love to give this a go again when the revamped plug-in is available. As it was, we fell back on group chat which worked OK, even if it was a bit slower and clunky. I guess you could look at it as a simulation of comms jamming by OPFOR :) One thing I would suggest is that if the plug-in borks, let's not discount the use of VON and/or appropriate text channels just because we don't normally use them. They still make a reasonable fall-back plan.

I do think the scenario needs just a little tuning. For the apparent size and skill of the opposing forces, it seems unrealistic to commit a force of Blue's size with no artillery support. I don't think the weapons section alone was adequate. We also had a number of Javelin gunners and assistants but as far as I ever heard, we never saw any armor. Maybe we just never got far enough for them to appear.

Finally, it looked like the poor chopper pilots might have been able to use some help. In addition to landing all the initial forces, then returning respawns and ferrying JIPs, they were supposed to handle resupply. Another chopper just for supply runs might be a good addition. 1-1-A spent a fair amount of time combat-ineffective due to running low on ammo and morphine and waiting for supplies to arrive. At that point, I think it was down to just Waldo flying.

Anyway, it was a good time regardless and the 1-1-A leaders did an outstanding job given the conditions. I think I got lost only once, for about 5 minutes.

1longtime
03-21-2010, 10:09 PM
This was an extremely disappointing event.

1) Slotting in must be faster.

We can't have 20-60 players in a holding pattern for 90 minutes. It's simply too much time. I have no answer for this, except to simply make it happen faster... I'm confident it can be done with our admins keeping things moving. Having extremely complicated mods/plugins makes it difficult, so quite simply if a player doesn't have it set up at start-time, they should be told they can't play. Set a hard cut-off for roll call, and set a goal for mission start to occur within 20 minutes after roll call.

This event was worse because the slotting was done twice, the second time in staggered fashion, which seemed necessary after the first crash but still took an unbearably long time.

Having a warmup JIP mission will help tremendously... don't ask me why, but the server handles it better when a large group gradually JIPs into a mission, and then switching to a new mission isn't an issue. Set up a 100+ holding mission and allow players to JIP and wait while ingame... then switch to the real mission. The gradual JIPs for the "warmup" mission seem to actually warmup the server. I have no idea why this works, but it does. This is a trick that is used every night in other clans, and it's effective... missions with 60+ players are started fairly regularly without crashing. Please consider trying this idea for the next event.

2) Expert mode

We aren't capable of it. With 60 players, many of which are often sharing areas of the map, and all looking essentially identical... it doesn't work. We need green diamonds, or perhaps the "TAGS" mod so they can be turned on/off as needed (so you can just "sneak a peek" at who is around you and you won't need to have diamonds or tags for the entire mission).

Playing a video game robs us of spatial awareness and ability to see each other's face properly. With large groups playing, and particularly after a respawn, keeping near each other is near impossible. The comms failure made it worse of course, but it's still too much and kills the fun to constantly trying not to lose your team. That's only my opinion of course. I sense that many people will disagree.

3) Laggy mission

Maybe it was just me, but I almost never lag and the FPS during the event was very low for me. The mission had many, many ungrouped AI with lotssss of waypoints... there are tricks for this, and I felt the mission's memory "footprint" needs to be adjusted for large missions like an event. In particular, perhaps not spawning enemy units until they are triggered, so only a small group of enemy AI are active at any given moment. There are also eventHandlers for kill cleanups that are important for long missions, as the bodycount stacks up.

4) Comms - what next?

The comms issues were off the chart. Perhaps an overall executive decision to completely disconnect from TS3 and try to rely on VON earlier in the mission would have been better... I'm particularly frustrated because I had hoped we would keep TS2 until 3 was at least out of beta and with proper Channel Commander. I believe the TG higher admin's decision to cutoff TS2 and move to TS3 was rushed, and I think some of the ArmA admins agree with that. TS3 AND the radio plugin are beta releases and it showed.

I don't fault the admins for trying new technologies, but I think the overall goal should now be stability. We are now straddling between a new technology (TS3 and plugin) that has shown it isn't ready for deployment and an old technology (TS2) that has been removed completely and is no longer even an option. It's a terrible place to be.

*Perhaps we should halt use of the plugin until a later date?
*Or temporarily setup a completely different comms server? Back to TS2? Ventrillo? I somehow doubt that the top TG site admins would like that though... having comms servers pop up with the name "TG" attached is a complicated business, and I'm sure they would prefer all comms activity to be consolidated on one server, but this may be the only option.
*VON-only isn't a good option, so I won't even mention it for serious consideration.

I will be looking at the admins for guidance on the comms technical issues... we need your leadership badly.

I don't like to sound like a negative jerk, but I think a discussion of what happened and how to make things better is important.

Merula
03-21-2010, 10:14 PM
If I have to be honest, I'm not looking forward to play this mission ever again.
I think one concept this mission captured well was that of "battle fatigue": after 2 hours of getting killed, waiting at spawn, flying to the LZ, walking to the designated position, getting slaughtered (often without being even able to see where the shots came from) and repeating the process over and over and over and over and over and over and over again removed all fun I had at the start.
They way I see it this mission could do with a couple of secobj less and we absolutely need some sort of support.
While I do not want to have a fleet of helicopters levelling the area while a mechanized company rolls in blasting everything standing, we could have really used some indirect fire support: when dealing with missions of this size (2 full infantry platoon) you would expect to have at least 2 heavy mortar squads and maybe even a light mortar squad to "soften up" the area before the infantry moves in, because otherwise it will still end up like it did this evening.

While I don't doubt that the C2 failure impacted heavily on our chances of success, I still believe that even if it didn't this mission would still be, if not impossible, insanely difficult. Case in point, even when comms were working and we had proper intra-squads coordination, we still were slaughtered.
If we tried flanking, the enemy would engage us from hidden area, if we tried using a base of fire element to cover the maneuver element, the bof would get supressed and the assault element killed. Whether we tried using cover to get close, or degrade the enemy positions from long range we still were slaughtered.
Every time I was near a LR radio I could hear how this squad was retreating, this other squad was pinned down, and that other squad was annihilated with only a guy left.

While this mission as it is can probably be won, it will STILL be through attrition, and I don't think it's a good design choice, also because AIs have no morale, while we human players do. The only reason I stuck through was because I'm too thick-headed to quit and did not want to make the other feel like I was quitting them because I could not stand getting shot again and again for almost no gain.

To conclude, we need some believable amount of extra support (because not having any in such a big and important (according to the story) engagement doesn't makes sense), and while having respawning enabled (nicely, I might add), it should not be a requirement to have even a slim chance of succeding. It should be winnable (no matter how hard) without having to depend on respawning.

And yes, I actually had fun while the squad was still present, even when we were all killed again. And I am grateful to the various leaders and the pilots for trying to hold together this trainwreck. It's just that after a while all the fun drained away, and for a while there was only frustration. And then the frustration drained away as well, leaving me tired. Simply tired.

Zedic
03-21-2010, 10:19 PM
Disregard.

ChiefBoatsRet
03-21-2010, 10:28 PM
http://img20.imageshack.us/img20/3394/zedic.jpg (http://img20.imageshack.us/i/zedic.jpg/)

ACOGRecon
03-21-2010, 10:35 PM
"You must spread some Reputation around before giving it to ChiefBoatsRet again."

darrrrrrnit!!!!!

Boondocksaint
03-22-2010, 12:19 AM
great post chief

hifibry
03-22-2010, 07:27 AM
Could the breakdown of comms have possibly been attributed to many players not having their configs set up correctly? For some reason, it defaults controls like short/long range broadcasting to really stupid keys like W, S, A and D.

Gill
03-22-2010, 12:45 PM
Had lots of problems with that Saturday night. Players not having long/short range bound to something else than their movement keys. "Who's hot-miking?!!!" Got tiresome.

Grunt 70
03-22-2010, 01:06 PM
I was an automatic rifleman. Impulse was my FT leader. Donegal was my SL. Boon was the other FT leader. I think we were 1-2-A. (But on the slot in screen we were 1-1-F)

Initially the mission was a blast. We were on the second wave of reinforcements and for a long time I felt fully immersed and we were operating pretty well as a squad. Fire teams were organized and Fire Team leaders were leading.

Enough people have piled on the comments on the comms status. As a grunt with no radio (which I love) it took a while for the confusion to filter down but that is about the time we lost effectiveness in our ops. From that point forward I felt like a fish in a bowl. The two can't be unrelated.

Shortly after my Squad Lead departed. Boon stepped up admirably to lead the squad but after another hour or two he had to go and my own FT lead had to leave at the same time. In addition I had to step out for 20 minutes for a honeydo and when I came back there were only 4 players left in my squad, Two JIP's and myself and Merlo. I tried to establish comms and reform the squad but by the time we were assembled and in the field the mission was called. This was after a solid 5 hours in game.

The mission was too long for a Sunday event. 2-3 hours is optimal. 5-6 hours is untenable and is the primary cause we lost effective cohesion with so many players having to drop for RL issues.

That doesn't mean it wasn't great though. The concept of a campaign is great and I applaud the admin team for managing and running the show.

Toyguy
03-22-2010, 01:56 PM
I believe there is some discussion going on about adding the Fireteam HUD to the basic set of mods TG uses. That would help a lot I think with the Expert mode issues mentioned. I have to confess though, I do like being able to see where I am, either by GPS or a map icon. Just seeing myself is sufficient. I found this map hard to manage because there were not a lot of usable landmarks, at least for a guy of my limited experience.

For these "signature events" I also agree with the comment that if a player isn't setup with plug-ins, config files, etc. at the roll call, then cut them loose until next time. Since we had to pre-register, we all knew what was needed and could have easily checked things out sometime in the days prior. Arriving unprepared on Sun afternoon just doesn't cut it.

To end on a positive note again, I loved the planning ahead of time, and the posting of the orders and the plan on the forum with full diagrams, etc. That was very well done and made the tasks easier as well as more immersive. All of us in 1-1-A knew exactly where we were going and what we were to do when we got there. Of course, once we hit feet dry, the enemy started shooting at us at the LZ and the plan underwent some emergency modifications :D At that point it was very immersive!

ThePieSpy
03-22-2010, 02:14 PM
Apologies to my squad for leaving mid-game, apparently Comcast decided to cut out randomly (again) and it just came back a few hours ago.

As other's have said, it's clear that we need to find a different communication solution for larger events atleast until we can get a version of the radio system that works better with large playercounts. As anyone in leadership can attest it was near impossible to coordinate outisde of ingame text between individual players let alone other squads and platoons. I had to be away from my squad for long periods of time just to try and coordinate a simple attack that shouldve taken place shortly after we hit the LZs.

From what i could tell though the mission looked excellent, and my gratitude goes out to all who contributed towards it.

Lastly a thanks and apology to ACOG and Cheesemenbashi, as ontop of trying to coordinate with me and their fireteams without an effective means, they had the stress of dealing with a missing squad leader after my internet went kaput.

EDIT: On a happier note i would like to say the method of slotting people squad by squad seemed to work pretty well, and I would hope that the admins continue this practice in future events.

waldo170
03-22-2010, 04:00 PM
I was Romeo 1 1 with Dredge as Romeo 1 2
I ended up flying for a good 5-6 hours straight which was intense. Me and dredge had to deal with some ZU-23's which were placed so that they were over effective and shooting far beyond the prospected AA areas (IN Red) Everything was going smoothly (except comms) air wise, once supplies became sparse is when things started becoming hard. Me and dredge manually loaded the supplies we could, but with the single box of medical supplies we could not resupply the corpsmen because most of the supplies had already been taken by respawning units. Once dredge due to RL issues, I was the only pilot left. I then made the resupply run a priority. ON my way to LZ Yankee, I was shot out of the pilot's seat by a PKM gunner. The bird landed itself but was still spooled up. Upon respawning I grabbed another bird and tried approaching the LZ from a different angle and much lower. Nonetheless I was still shot out of the helicopter from the PKM gunner, with a little help from the ZU-23.

This being stated, I believe AA markers should be properly marked. Overdoing AA is not necessary because the only pilots that would be allowed to fly in a large Event like this would hopefully be ones that would respect simple NO FLY ZONES. A second thought would be take away all guided or AA capabilities and leave it to the accurate and deadly PKM gunners or RPG7 guys to harrass LZs and Transport birds.

Since this mission was not completed, will the next stage in the campaign be a mission about a retake on the island with larger force? Or continue on and pretending we won?

All in all though the mission was very fun.

ChiefBoatsRet
03-22-2010, 04:11 PM
First I would like to thank all those that put in a lot of time in hard work in putting this mission together because the mission itself was difficult and very good. As we all saw , excellent comms were needed to get that many people to where they had to be to survive let alone win. The technical breakdowns on a large mission is almost expected now so most of us were not surprised . After the first hour I was in almost constant desync which made it difficult for me to keep up, so after about an hour and a half I had to give it up. It dosen't really matter because the next great mission is just around the corner and when you finish with it... all this is forgotten.

cheesemenbashi
03-22-2010, 05:58 PM
What most impressed me was the patience of the guys in my fireteam and squad to stick around so long doing not very much, good job to niko, erppa and pineapple who were 'the last men standing' before the 9 or so that were alive died, and I left to do something else. It's obvious what buggered the mission was the the failure of the plugin and from that comms breakdown came the lack of moment, the loss of cohesion and the large amounts of dying.

The 'superversion' of the plugin, if you go by estimates on dev heaven (http://dev-heaven.net/issues/9550), is due out this week and has "500% performance improvement" among its many listed features. If we plan to have any more events of this scale and we think the plugin'll have issues we should revert back to teamspeak comms and VON for that event but here's hoping the new version will sort the problems.

waldo170
03-22-2010, 11:54 PM
its not with the plugin. It the TS server not being able to handle 60 speakers and listeners in one channel alone. the TS handles each channel as a pipe persay, too many players in one pipe and well you get what we had on sunday.

Merula
03-23-2010, 01:06 AM
its not with the plugin. It the TS server not being able to handle 60 speakers and listeners in one channel alone. the TS handles each channel as a pipe persay, too many players in one pipe and well you get what we had on sunday.

That would explain why during the event it was the TS server that crashed regularly, instead of the plugin ceasing to function entirely. Maybe more tests should be done to see how many people can the server reliably support at once before kicking the bucket?
Hopefully the next releases of the plugin will reduce the load somewhat, but if it's a TS issue I don't know what ways there could be to make the system work: I don't think modifying the plugin to support multiple channels at once would work, since I believe the server load would actually increase, so the only solution to use it would be to limit the player count to around 40 (since everything worked almost perfectly again after the player count dropped under 50).
Still, thanks to the event we now know about this and hopefully it will be possible to find way to make everything work properly the next time. :)

Dslyecxi
03-23-2010, 02:17 AM
That would explain why during the event it was the TS server that crashed regularly, instead of the plugin ceasing to function entirely. Maybe more tests should be done to see how many people can the server reliably support at once before kicking the bucket?
Hopefully the next releases of the plugin will reduce the load somewhat, but if it's a TS issue I don't know what ways there could be to make the system work: I don't think modifying the plugin to support multiple channels at once would work, since I believe the server load would actually increase, so the only solution to use it would be to limit the player count to around 40 (since everything worked almost perfectly again after the player count dropped under 50).
Still, thanks to the event we now know about this and hopefully it will be possible to find way to make everything work properly the next time. :)
It's a shame to hear that the TS3 side of things fell down in the event. My biggest concern previously with this (and part of the reason why we've yet to commit to the A2TS3 plugin ourselves) was that it's effectively having you receive the audio info from x-many clients at once, even if you only actually "hear" a tiny fraction of that. If you take 60 people, disperse them around a map, and have them all chat at once - presumably you have 60 voice streams at the same time, and it's reasonable to expect that TS3 is not designed to have such an inhuman number of people talking at the same time, in the same channel.

As far as how this might be prevented in the future, I've wondered if the introduction of TS3 "Channel Commander" might give the A2TS3 devs an avenue to reduce the load - that, or perhaps dynamic whispering. Which is to say that whenever you are in a possible contact with someone (either by radio, local voice, etc), it could toggle CC or establish a temporary whisperlist that would be able to reduce the number of total voice streams to a manageable level.

Hopefully something can be done, whatever the solution might be. A2TS3 has some powerful potential, but if TS3 and/or a combination of the two can't cope with 55-ish players at once, that's a pretty significant limitation to have to deal with.

jaynus
03-23-2010, 03:52 AM
It's a shame to hear that the TS3 side of things fell down in the event. My biggest concern previously with this (and part of the reason why we've yet to commit to the A2TS3 plugin ourselves) was that it's effectively having you receive the audio info from x-many clients at once, even if you only actually "hear" a tiny fraction of that. If you take 60 people, disperse them around a map, and have them all chat at once - presumably you have 60 voice streams at the same time, and it's reasonable to expect that TS3 is not designed to have such an inhuman number of people talking at the same time, in the same channel.

As far as how this might be prevented in the future, I've wondered if the introduction of TS3 "Channel Commander" might give the A2TS3 devs an avenue to reduce the load - that, or perhaps dynamic whispering. Which is to say that whenever you are in a possible contact with someone (either by radio, local voice, etc), it could toggle CC or establish a temporary whisperlist that would be able to reduce the number of total voice streams to a manageable level.

Hopefully something can be done, whatever the solution might be. A2TS3 has some powerful potential, but if TS3 and/or a combination of the two can't cope with 55-ish players at once, that's a pretty significant limitation to have to deal with.

To expand on this; I believe the 3d range/distance calculations being done on 50 plus players is a major contributor to the lag as well.

Think of it this way -

1. Your client is receiving 60 audio streams at once
2. Your client now needs to calculate, via some semi-decent trig, whether you should hear them or not.

Do that 60 times every 100ms, on top of the plugin only communicating via a pretty poopy synchronous clipboard pipe, and your going to get the 5-10 second delays on comms we were getting ONTOP of the TS server crashing. I'm suprised the arma2ts.dll portion of this doesn't auto-swap channels for people. Maybe thats what the new version does. the TS3 plugin code to do that based on distance is pretty easy (but requires disabling that annoying chick voice).

Dredge
03-23-2010, 10:30 AM
Well the version the ACE devs are working on is supposed to improve performance by like 500% so, we just have to wait and see.