How do you really play? The app that answers you honestly.

Measure your play against chance, know your real balance and play counter-cyclically. Six tips for Swiss lottery players – with calculated numbers.

Almost every player has a story in their head. "I've been playing the same numbers for years." "I have a good feeling about the 7." "Last year went better." Nice stories. Only: hardly anyone has numbers to go with them. How much have you spent this year? How often were you above chance? Which method actually hit better for you – and from when is that difference even more than noise? No lucky streak answers these questions. Only data. Here are seven tips to help you truly see your own play for the first time. Some of them are uncomfortable. That's exactly why they work. First: What are the others doing? So you know where you stand, here are the solid numbers for Switzerland. 39.4% of respondents have taken part in a lottery in the last twelve months – lotteries are thus by far the most popular form of money gaming in the country. 13.2% play at least monthly. Only 2.2% play more than once a week. Swiss Health Survey 2022, ISGF Report No. 422 So anyone who plays every draw belongs to a small minority. And anyone who plays once a month is completely normal. It gets more interesting with behaviour around large jackpots. Our analysis of over a hundred draws shows a very clear pattern: Draw Jackpot under 10 Mio. Jackpot from 20 Mio. Wednesday 1'081'160 tips 1'693'382 tips Saturday 1'200'187 tips 1'811'583 tips Median values, observed based on published Swisslos amounts. Read the numbers again. With a large jackpot, around 600'000 more tips are submitted – a good 50% more competition for the same pot. That is the most important sentence in this article: The bigger the jackpot, the more people will share it with you in the event of a win. Tip 1: Know your real balance Most players remember the five number win from three years ago. Nobody remembers the 180 francs that have gone into it since. The Spielbuch adds both together: staked, won, balance. Per month, per year, per tip. No estimate, no memory – the actual number. A real example from the app: four recorded tickets, four evaluated draws, CHF 22.00 staked, CHF 0.00 won. Balance: minus 22 francs. That is not a bad streak. That is the lottery. And anyone who knows this number makes better decisions than someone who doesn't. Tip 2: Measure yourself against chance – not against feeling This is the strongest trick in the whole article, and almost nobody knows it. With 6 out of 42, the pure chance expectation is 0.86 main hits per tip. That is not an estimate, that is mathematics: 6 × 6/42 = 0.857. For the lucky number it is 17% , for any prize tier around 2.9% – that is about 1 in 34 per tip. These three numbers are your benchmark. The Trefferanalyse places your real values right next to them. An example from the app: Ø 0.78 hits against a chance expectation of 0.86. Lucky number 15% instead of 17%. Win rate 0% instead of 2.9%. That looks like a bad run. But it isn't – it is exactly the normal range. And that is precisely why the app contains the sentence we like best: "Under 20 comparisons per method, differences are pure chance." Remember this rule. It protects you from the most expensive mistake in the lottery: crafting a strategy out of five good rounds. Tip 3: Test your ideas before you put money in You have a number idea and want to know whether it's any good? Save it as a favourite . Favourites are saved predictions without a stake – they don't count towards your balance, cost nothing, and are still automatically evaluated at every official draw. This way you collect real comparison data over weeks without risking a franc. And if an idea convinces you, one click on "Als Schein übernehmen" turns it into a real ticket. Try it for free before you pay. That's the right way round. Tip 4: Play counter cyclically Now the consequence of the table above. Your number selection does not change your chance of winning . It is 1 : 31'474'716 per tip for the jackpot, always, at every draw. What you can influence is only: how many people you would have to share with. Two levers really work here: The timing. With a smaller jackpot, over half a million fewer people play. The pot is smaller, your share in the event of a win is bigger. The number zone. Millions of tips never go beyond 31 because they represent birth dates. The range 32 to 42 makes up 26.2% of the playing field and is systematically underrepresented. Two numbers from there cost nothing and take two seconds. Tip 5: Budget first. Always. An uncomfortable but important number from the same study: Among lottery players of the last twelve months, 7.9% show risky or problematic playing behaviour – significantly more than the average across all money games. That is why the Spielbuch has a monthly budget that you set yourself, plus optional emails after every evaluated draw and before expiring tickets. The app states clearly what this budget is – and what it is not: "The budget is a guide, not a playing block." The decision remains with you. We just give you the number to go with it. For orientation: A ticket costs at least CHF 5.00 – two tips at CHF 2.50 each. That is the framework, not the goal. Tip 6: Decide for yourself whether the AI plays along This is the point most users overlook at first – and the most important switch in the whole app. You have the choice between eight mathematical methods. And this choice determines whether the learning AI has any say at all: Choose a single method – frequency analysis, Poisson, Markov chains, Monte Carlo, Bayes, Fourier cycles or regression – then it calculates "deliberately pure and unchanged" . No weighting, no influence. You see exactly what you selected. Choose "Ensemble Alle " , then the AI mixes all methods according to their current rating. Only here does the learned weighting take effect. Playing with AI or without AI – your decision, for every single prediction. How the AI learns Every official draw flows in. After each draw, the AI compares the suggestions of all eight methods with the numbers actually drawn. Those that were close gain more influence on the next recommendation. Those that missed, less. Automatically, week after week – you don't have to do anything. Scoring is by points: maximum 100 per draw, of which 85 for the main numbers and 15 for the lucky number. Pure chance is 14.6 points. That is the baseline against which every method must be measured. This is what the ranking looks like after 43 evaluated draws: Method Ø Points Ø Hits Lucky number Influence 1 Fourier cycles 17.05 0.88 30% 20.7% 2 Ensemble Alle 15.31 0.88 19% Mixture 3 Bayes inference 15.25 0.95 12% 16.3% 4 Markov chains 14.94 0.91 14% 15.6% 5 Poisson 14.61 0.88 14% 14.9% 6 Monte Carlo 13.68 0.74 21% 13.2% 7 Regression 11.34 0.63 16% 9.7% 8 Frequency analysis 11.32 0.65 14% 9.6% Chance baseline: 14.6 points · 0.86 hits · 17% lucky number. The influence values of the seven single methods add up to exactly 100%. And now the honest classification Here we could now write: "Fourier cycles leads – play Fourier " We won't. Because we did the math. With 43 draws, the measurement error for the point score is around ±1.9 points. Fourier's lead over chance is 2.4 points – that is barely 1.3 measurement errors. For the main hits, the entire field from 0.63 to 0.95 lies within two measurement errors around the chance line. Fourier's lead comes almost entirely from the lucky number: 30% instead of 17%. On its own, that would be striking. But eight methods are compared simultaneously – and when you measure eight candidates, the probability is around 15% that one looks this good purely by chance. Translated: After 43 draws, the ranking is not yet proof, but a snapshot. That is precisely why the app itself calls this phase "warm up phase". And that is precisely why the last place is never fully switched off – so that a method can catch up again after a weak phase and the AI doesn't get stuck in a single track. Incidentally, the app itself writes the decisive sentence in its own explanation: "The ranking shows which method has been closest in the real draws so far — none of them can predict the next draw." The benefit of the ranking is therefore not that it reveals the winning method to you. It is that you can see what happens – instead of having to trust a black box. Tip 7: Let the AI contradict you Most providers build an AI that agrees with you. Ours is built to contradict you. The KI Einordnung calculates before every draw what is realistic – and says sentences that no marketing person would write voluntarily: "The model value per tip is 2.09 francs and thus below the stake of 2.50 francs; a tip loses money on average." "The series of 42 consecutive draws without a rank 1 winner is officially documented and does not change this chance." And the KI Synthese summarises your own week – without sugar coating. It even explains to you that your own gamification level means nothing: "The level and the streak of active days are pure app gamification from your usage days – they have nothing to do with lottery luck." An app that tells you its own badges are meaningless. That is the standard we set for ourselves. And what the app explicitly CANNOT do Because honesty only counts when it is also uncomfortable: The biorhythm provides no numbers. It is classification, not a number generator. Moon phase, Jupiter hour and retrogrades award only confidence points. They do not change a single drawn ball. The oracle only flows in with Premium and only weighted 1.3 times. No method in the world changes the draw probability. 1 : 31'474'716 remains 1 : 31'474'716. What our analyses actually optimise is the expected number of co winners – that is, the possible payout amount. Not the chance of winning. Anyone who promises you otherwise is selling you something. Your player profile awaits You don't need a subscription or a credit card to start. Record a ticket, save a few favourites, and after a few draws you will see for the first time in black and white how you really play. Most are surprised. Almost all play more relaxed afterwards. → Register for free and start your player profile https://lottoprognose.ch/auth?tab=register → Submit tips at Swisslos https://www.swisslos.ch/de/swisslotto/einzeltipps/spielen.html tipp The next draw is on Wednesday, 14 October – deadline 19:00, announced jackpot CHF 34.6 million. Enough time to know beforehand who you are when you play. If you don't play, you can't win. Good luck. 🍀

Zurück zum Blog