Pathway to an Event Server

[Talkos]

Preface: I am not an admin for the Arma servers and the opinions stated below do not have any sort of official status. And apologies if this is already being discussed and I step on any toes, as that is definitely not my intention on putting my thoughts out in the open.

I have been talking to various people for a while now on the viability of an Arma2 event based server, and in the past month have started hosting open house events in order to gauge viability and interest in the possibility, sort of as a open testbed just using my own computer as a temporary host. These events have proven to be popular enough that there has even been some backlash, but I feel that I should explain my overall thoughts and direction of where I think this could lead.


Premise 1: A 24/7 server would be an impossibility to keep primed and active.
I think many people would agree that the time investment in Arma2 means that a 24/7 server would sit empty most of the time. It would also be very restrictive on the types of missions that can be run. Either very bland warfare or domination missions, or continuous missions such as dayz or wasteland. I think the previous attempts to start an Arma2 server focused on the earlier idea, and our most recent attempts focused on the DayZ option, neither proved viable. On the other hand, the influx of people to DayZ created a large and diverse population of Arma2 players that want more than just hiding from zombies and sniping.

Outside of Arma, we've seen this phenomenon across our servers. I have a lot of experience priming and helping out DoD, and although it has a very vocal minority that plays it, people were simply burned out after an extended period. Thus achieving server population was a very difficult thing to do which burnt out a lot of players. On the other hand, it had considerable success with its PUGs(scheduled pick up games) which were often played on maps that were not in the regular rotation. I've noticed a similar thing with TF2, when the server was flagging, often the Highlander events proved popular despite a waning population. I think that there is a realization that the same may be happening with a resurgence of BF2 nostalgia, it's an old game and nobody expects it to be able to stay primed every night, but having BF2 nights on a weekly or monthly basis has quite a bit of support within the TPG community.

Premise 2: The most popular Arma2 communities offer event based as opposed to continuous missions
Prime example being Shacktac which regularly has 200+ individuals playing, plus a waiting list weeks long. Other examples include United Operations, Veterans Gaming(although Veterans has both a constant server and events), and 15th MEU. Many of these will go as far as to vet their applicants before letting them join in the regular mission sessions.

Premise 3: Event based servers offer benefits in administration and mission options
The benefits of an event based server are numerous: Less burnout on players and admins. Additional mission variety due to the ability to have recognized start-times, as well as forewarning to players if mods or maps are required. Tailored missions to the number of players involved in order to keep excitement high: A major issue with larger missions of few people show up is that a single individual might be expected to take on a platoon of tanks, by knowing the population the administration can choose missions based upon it easier than having a large scale mission on all the time with frequent deadzones of activity where it is simply not fun.

I am not familiar with how the old Arma2 server chose its missions, but in my experience, selection for map rotation is a sluggish and often cumbersome thing. Requiring decision by committee and often entering fad maps that people soon groaned to see coming up. In a weekly or bi-weekly event night, one individual or a small team of individuals could take responsibility for choosing the missions played and if they proved to be less than popular, we can simply note not to play them in the next event night. Also, it could create more community involvement, as clans or individuals may take responsibility for choosing the missions for a week or day.



TPG Server vs Temporary hosts aka Lessons Learned from the Testbed

So some may now wonder why to host a TPG server when so far we've had success with simply having someone host.
1. Stability - currently our hosts are subject to the whims of their internet. Lightning storms or lag spikes can easily kill their connection and thus the mission.
2. Capacity - The limitations of computer power have limited the number of individuals in the missions. We have peaked at around 14-16 simply because to go higher than that on someone who is also attempting to play often causes crashes or high lag
3. Openness - In most of my hosted games I try to leave it unpassworded, however since there are only a few missions that seem to have caught on, and those with a relatively low population count, often they hosted games will be passworded. I would liken this to a Left 4 Dead game mentality which although fun, doesn't offer a great deal of possibility for new people to join.
4. Visibility - a TPG server would be easier to locate and offers a recruitment tool for TPG itself, although I've made a point to tell smurfs that they should visit TPG for more information, an intermittent message over the server is always a better option. Also, given my own schedule and those of the others that have been hosting it(props to the NOM and PHX crew) I simply can't tell people who wander in "Yeah, we do this every Friday, search for XYZ."
5. Accessibility for Mods - There are several launchers available for Arma2. The one familiar to DayZ players is likely DayZ commander, which makes it very easy to update to the correct version and mod on the server one wishes to join. The one used for every other Arma2 map and mod is called PlayWithSix which offers similar features, in that one can select a server and through a single click update to the correct version and download any mods needed, as well as starting up Arma2 with those mods enabled. This is currently non-functional for privately hosted games. As it is in order to play a mission with a mod or different map on a hosted machine, one needs to download the appropriate files, make sure they have the same version, check in their arma2 expansions in-game menu to make sure they have the right ones checked, and cross their fingers and hope that things work.
6. Mission design - Paradoxically, the bigger the server population is, the easier it is to create a mission for it. In my opinion, a good TvT mission requires at least 20 people, on the other hand, the mission designer has less requirement to create AI behaviors. E.g. A mission that is a recreation of a CSS hostage rescue mission would require hours of time to make in singleplayer or cooperative, but can be created for TvT simply by placing slots for Terrorists, CTs and an independent hostage.


Challenges and requirements
Admin commitment. Simply put, although the server will be down 5 or 6 days a week, at least one admin would likely need to be on during the time it is up in order to advance the missions. In addition there is likely to be an additional commitment of prechoosing the missions, or editing them to TPG standards(TPGers don't like spending hours as a bird). This is not necessarily one that need be shouldered solely by the admins however.

Soon however, we will have built up a database of missions that offer repeated enjoyment or be slightly modified to keep people on their toes.

Timeframe Limitations
Arma 3 is in Alpha right now, and I think we need to recognize that once Arma 3 comes out of alpha in a year or two and becomes a stable product with mission and mod support, much of the population will migrate. Thus, a pessimistic look may see this event server idea as only viable for a year. Personally I think that it will take about two to three years from now before Arma 3 fully supplants Arma 2 in functionality and mission variety, but either way, this is not going to have a run as long as CSS or TF2 here at TPG.

Additional Costs
While we had the DayZ server, my suggestion would be to have the events coincide with required server restarts. Turning the DayZ server into an event server for a day as well as serving a function of restarting the server. However with the closure of the DayZ server, we end up with a server that is open one or two days a week and otherwise not doing anything if it has a dedicated server. Unless it is simply run on existing TPG architecture, of which I am relatively ignorant of the current setup.

Viability tests
I was actually very surprised by how popular the hosted missions were. I expected that I would have to beg and plead for people to try it out, instead there have been groups escaping Chernarus pretty much every night since I started it, and groups of a dozen or more in the other missions that I would host on the weekends. I think that these proof of concept tests among the TPG population show that there is a significant demand for this type of directed Arma2 gameplay as opposed to the more sandbox gameplay that DayZ and Wasteland offer.


Options:
I see three main viable options of models for how a TPG hosted server would look like.

A. Closed(Shacktac) Model - Require people to register at TPG to join, and require a password on the server. Offers a buffer against teamkillers and trolls, but limits openness and opportunity to recruit people to TPG. Also in my opinion against the TPG ethos of inclusion.
B. Mixed(Veterans) Model - Have the server alternatively be open and closed. Open nights would be vanilla or near vanilla game missions which require no mods. Closed nights would be more organized and allow a greater variety of missions.
C. Open - Have no password on the server. Maximum amount of new faces, but requires closer admin attention. Also may limit mission selection, as new people would need to be briefed and are not guaranteed on going along with the plan.

Personally I would recommend either the mixed or open model.

Anyways, that is pretty much the sum of my thoughts on the idea of an Arma2 event server. I hope that this opens up the floor for some discussion on whether to start one, or whether to keep this as it is with simple temporary hosting, or if anyone has any other great ideas on how to overcome the challenges that we face.

[SmokenScion]

Looks Great.

[Merritt]

Since all my gaming time is spread between Arma 3 and the Arma 2 mods anyway I'm in