How to set up a prediction game: the six decisions that shape your season

Take one league. A Bundesliga season is 34 matchdays, late August to May — already more stamina than most leagues realise they need. Adding the cup, the Champions League and a second division doesn’t double the fun; it doubles the number of dates on which someone can forget to predict.

• One league: a steady rhythm, one weekend per matchday, everyone knows when they’re up. The normal case, and almost always right for a first league.
• League + cup: the cup brings Wednesday nights and knockout ties — and with them the question of whether a 1:1 after 90 minutes counts, or the 2:1 after extra time.
• Two leagues: works if your league watches both anyway. If not, half of you predict the second division blind — and blind predicting is the fastest route to boredom.

The date that matters is already fixed: Bundesliga matchday 1 kicks off on Friday, 28 August 2026. The Premier League starts on Friday, 21 August 2026, the 2. Bundesliga on 7 August, and the first round of the DFB-Pokal runs from 21 to 24 August. If your league is meant to be standing by matchday 1, the week before is the last one in which that can be organised calmly.

Eight to thirty is the range in which a prediction league works. Below it there is no friction — with five people you know how it ends by matchday three. Above it the table gets impersonal, the leader is a stranger, and the group chat tips from banter into noise.

• Small league (8–15), everyone follows football: you can take a system in which the best predictor really wins. Nobody drops out for falling behind — you all know each other.
• Big or mixed league (20+), lots of casual predictors: here you need a system that keeps hope alive longer. Someone 40 points down after ten matchdays with no way back doesn’t predict on the eleventh.

And one of you matters more than the rest: the person who creates it. They invite, they explain the rules, they nudge the forgetful. A league without that person doesn’t run — it dissolves. If that’s you: everything here is your decision, and nobody else is going to make it.

The scoring system is the only decision that changes who wins. Everything else decides how smoothly your league runs. This one decides whose season it is.

The classic German shape has three hit levels:

What you’re choosing is the spread between the top and the bottom level — a bigger lever than it looks. At 4/3/2 an exact result is worth two correct tendencies. At 6/2/1 it’s worth six. In the first case, whoever reads 34 matchdays best wins. In the second, whoever’s exact predictions happened to land wins — and the league stays open until May in exchange.

That is the real trade: fairness against suspense. There is no right answer, only the one that fits your league.

→ In full, with every level, the joker and the draw problem: Scoring systems in a prediction game — and what they really do to your season

Put the deadline at each match’s kickoff. Not Friday night, not “before the first game of the matchday”, not a time somebody has to keep in their head. Kickoff — per match.

• One shared deadline before the first match is fairer in theory (nobody knows more than anyone else) and, in practice, the most common source of “I forgot” — because the deadline has nothing to do with the match you’re thinking about.
• Deadline = kickoff means: whoever thinks about the 3:30 match at 3:20 can still predict it. That is the moment people actually think about football.
• The objection is real: predict Sunday’s match on Sunday and you already know Saturday’s results. It does you no good — you’re predicting a different match. It doesn’t change your prediction, only your nerves.

What matters is that the deadline is hard and doesn’t depend on the clock in someone’s phone. And that postponed matches take their deadline with them. Get that wrong and you eventually have a matchday on which somebody predicted after kickoff — and an argument that costs you the league.

Invite with a link, into the chat your people already use — WhatsApp, Slack, Microsoft Teams. Every extra hurdle between the link and the first prediction costs you players, and the hurdles are almost always the same: create an account, install an app, go hunting for a confirmation email.

What happens in practice: 20 people in the chat, 14 click, 9 arrive. The five you lose weren’t having a bad day — they had a registration dialog.

• One link, no search. Nobody should have to “find” a league by name.
• Ask for an account as late as possible — ideally only when someone wants to claim their name, not before their first prediction.
• The link has to show something in the preview. A bare link in a group chat gets scrolled past; one with a card that shows your league’s name gets tapped.
• Latecomers must not start with a deficit. Someone joining from matchday 3 is scored from matchday 3 — otherwise they don’t join at all.

This question decides whether your league lives until May. Prediction leagues almost never die of arguments — they die of a forgotten matchday. Miss one and you fall behind; fall behind and you check the table less; stop checking the table and you miss the next one. That is the whole mechanism.

• Reminders. A good reminder comes once, in time, and says what to do. A bad one comes three times and gets swiped away — after which it is silent forever.
• How a missed matchday is handled. Some leagues score a missing prediction as 0 points, some award an automatic substitute prediction. What matters is that you decide up front and say it out loud once. A substitute prediction nobody knew about feels like cheating.
• Dropped scores. Some leagues drop the worst matchday. That takes pressure off — and takes a little of the winner’s credit with it. Also a trade, not a feature.

And one sentence every founder should write down: the matchday on which half the league doesn’t predict is not the matchday your league dies. It is the one on which it had already died — two matchdays earlier, when three people missed one and nobody went and got them.

Three things kill prediction leagues, and none of them is a missing feature.

1. The forgotten matchday. Number one by a distance.
2. The rule changed mid-season. Sooner or later someone decides draws should be worth more. Change that on matchday 12 and you retroactively change who is leading — and nobody forgives that. Lock the rules before matchday 1 and don’t touch them until May.
3. The league nobody can find. It sits behind a login three people have forgotten, in an app two never installed. The table has to be one tap away, or it doesn’t get looked at.

A prediction game with no stakes isn’t gambling — it’s a game for bragging rights. The moment money goes into a pot that gets paid out to the winner, it gets delicate, and at work it certainly will: that is exactly where someone from HR eventually asks.

The simplest way to spare yourself all of it: no stakes, no pot. If your league wants a kitty anyway (the classic: everyone throws in five euros, the winner buys dinner), run it alongside the prediction game — as a list, not as payment processing. We touch no money ourselves: our kitty is an overview of who has paid in, and nothing else.

→ More on this, including the questions IT and HR will ask: Office prediction game

Honestly, because it would be worthless otherwise: you cannot create a league on tipperoo.de yet. We launch for the 2026/27 season. If you need your league now, Kicktipp is the established route — it works today, it’s free, and their own site carries the instructions.

What you take from this guide in that case is still the important part: you make the six decisions above on any platform, and the platform will not make them for you.

And if you can wait: we’re building Tipperoo for exactly this league — ad-free, no betting odds, made for a phone. If you want to move an existing Kicktipp league across, here is what comes with you and what honestly doesn’t: Kicktipp alternative.