How to start an RSPS in 2026

We ran RSPS toplists from 2011 to 2019 and watched several thousand servers launch. The overwhelming majority were dead within a month, and they died for the same handful of reasons every time — almost none of them technical.

If you are thinking about starting one in 2026, here is what the last fifteen years of the scene suggest you should know before you download a base.

Pick a base, do not write one

Nobody starts from nothing any more, and you should not either. The open-source bases cover the revisions people actually play, and the difference between a server that launches and one that does not is usually whether the owner spent three months building a networking layer that already existed.

The practical choices:

• 317 / 377 — the classic 2007-era feel, the largest pool of existing content and plugins, and the most competition. If you want a familiar economy and PvM loop, this is where the players are.
• 474 / 508 — mid-era, thinner community, distinctive if you want to be different.
• OSRS-style (revision 200+) — the closest to modern Old School RuneScape, the most demanding to keep current, and where a lot of the current audience actually is.

Whichever you take, read its licence, and read the code before you run it. Bases circulate with backdoors in them. This is not a hypothetical — it is a recurring event in this scene, and the usual payload gives someone remote admin in your game or a shell on your box. If you cannot audit it yourself, pick a base with a public repository and visible commit history rather than a zip from a forum post.

The legal position, stated plainly

RuneScape is Jagex's intellectual property. Private servers exist because enforcement is selective, not because they are permitted. Jagex issues DMCA takedowns and has pursued the larger and more commercial servers, and the pattern over the years is consistent: what draws attention is scale and money, particularly servers that advertise aggressively, sell heavily, or use Jagex's assets in marketing.

You should decide up front how much risk you are taking on, and understand that hosting providers respond to takedown notices quickly and without much interest in your side of it. Keep backups off the server that hosts the game.

We are not lawyers and this is not advice. It is just what has visibly happened to other people.

Hosting: what you actually need

RSPS are much lighter than people assume. The server is single-process, mostly memory-bound, and the CPU load per player is small.

For a launch:

• A VPS with 4GB of RAM and 2 dedicated cores handles a few hundred concurrent players comfortably. That is €15–25 a month.
• DDoS protection is not optional. This scene attacks each other, routinely, and the first time you rank well on a toplist is a common trigger. Use a provider that includes filtering — OVH, Hetzner with their protection, or a game-hosting specialist — rather than the cheapest box you can find.
• Put the website on separate hosting. When the game server gets attacked, you want your site, your forums and your Discord invite still reachable so you can tell people what is happening.

You do not need a dedicated machine on day one. You will need one at around a thousand concurrent players, which is a problem you would love to have.

Build the community before the server

This is the part nearly everyone gets backwards, and it is the single strongest predictor of whether a server is alive in three months.

A server launching to nobody is a dead server, because the first thing a new player sees is an empty world, and an empty world is the fastest way to lose them. The servers that survive launch with an audience already assembled.

That means, starting six to eight weeks before you open:

• A Discord, and actual activity in it. Not an empty server with fourteen channels. Two channels and forty people talking is worth vastly more.
• Development updates with screenshots. People invest in things they watched get built. Post progress weekly even when it is small.
• A beta with real players. Twenty testers will find things you never would, and half of them become your first regulars and moderators.

If you cannot get a hundred people into a Discord before launch, launching will not fix that. The game is not the hard part — attention is.

The economy decision you cannot reverse

Decide your rates and your donation model before launch, because both are effectively permanent. Changing rates after launch devalues everything your existing players earned, and every server that has done it has taken a visible hit.

The one that matters most: decide what donations buy. Cosmetics and convenience keep a server alive for years. Selling power kills servers, slowly and then suddenly — first the competitive players leave, then the economy inflates, then the people who stayed for the community find there is no community left.

You will be tempted, because power sells about five times better than cosmetics in the short term. The graveyard is full of servers that took that trade.

Launch week

Get listed on toplists before you open, not after. Listings need time to accumulate votes and appear in search, and the traffic arrives when players are looking, not when you are ready.

For the first week specifically:

• Have vote rewards working on day one. Voting is how new players find you, and the first week of a month is when the ranking is most winnable.
• Watch for the exploit. There is always one. Have a plan for rolling back and communicating rather than pretending it did not happen.
• Be visible in your own Discord constantly. Early players are choosing whether this is a place they want to spend time, and the owner being present is a large part of that answer.

Why most servers die

From watching a lot of them:

1. The owner ran out of energy. This is the top cause by a distance. Running a server is a second job with an audience that notices when you take a night off. Decide honestly whether you have six months of that in you.
2. Launching to an empty world. Covered above.
3. No updates after month one. Players tolerate bugs; they do not tolerate the feeling that nothing is being worked on.
4. The pay-to-win spiral. Revenue dips, so power gets sold, so competitive players leave, so revenue dips.
5. One person holding everything. No second admin, no backups anyone else can restore, no plan if the owner disappears. A great many servers ended not because they failed but because one person got busy.

None of that is about code.

If you are still in

The scene is smaller than it was in 2014 and there is still real room in it, because the median server is poorly run rather than technically deficient. A server that launches with an audience, delivers vote rewards reliably, updates every fortnight and does not sell power will be in the top ten of most RSPS lists within a couple of months. That is not a high bar; it is just one almost nobody clears.

When you are ready, list your server — it is free, and being ranked from day one is worth more than being ranked after you have already lost your launch traffic.