[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"blog-article-how-to-predict-correct-score-en":3,"mdc--1zwnkc-key":65},{"id":4,"slug":5,"status":6,"section":7,"category":8,"author":9,"publish_date":10,"date_updated":11,"read_time":12,"image":13,"embedded_components":14,"related_calculators":14,"related_articles":15,"title":16,"description":17,"keywords":18,"content":27,"faq":28,"availableLocales":62},"a50821d3-46d6-4bf7-8d36-1ed314ab5983","how-to-predict-correct-score","published","betting","strategies","Evgeniy Volkov","2026-07-19","2026-07-19T06:23:18.848735+00:00",15,"\u002Fimages\u002Fblog\u002Fhow-to-predict-correct-score.webp","[]",[],"How to Predict the Correct Score: Poisson and xG (2026)","How to predict a correct score in football with expected goals (xG) and the Poisson model. A step-by-step method, worked example and calculator.",[19,20,21,22,23,24,25,26],"how to predict correct score","correct score prediction","football score prediction","correct score calculation method","poisson method football","expected goals xG","football score forecast","how correct score is calculated","# How to Predict the Correct Score: Poisson and xG (2026)\n\nI once bet on 2-1 in a match I was completely sure about. It finished 1-1. Frustrating, sure, but what got me afterward was this: when I ran that match through the model, it had flagged 1-1 as the most likely scoreline all along. I went with my gut instead of the numbers. That's exactly what this guide is about: how to predict correct scores with data in 2026, not instinct.\n\nOne thing upfront: you cannot guarantee the correct score, and any site or app claiming otherwise is lying to you. What you *can* do is calculate the probability of each scoreline and find bets where the bookmaker has mispriced things. Below is a step-by-step method using expected goals (xG) and the Poisson distribution, a live worked example with real numbers, and a built-in [calculator](\u002Fbetting\u002Fcorrect-score-calculator).\n\n## TL;DR: How to Calculate the Correct Score in 5 Minutes\n\nShort version: take both teams' expected goals, run them through the Poisson formula, multiply the results together, and you get the probability of each scoreline. You only bet where your probability beats what's baked into the odds. Five steps:\n\n| Step | What you do |\n|---|---|\n| 1 | Gather recent xG and form for both teams |\n| 2 | Calculate expected goals λ using attack and defence strength |\n| 3 | Run the Poisson distribution for 0, 1, 2, 3 goals |\n| 4 | Build the full scoreline grid (multiply the probabilities together) |\n| 5 | Spot the value scorelines and only bet where the edge exists |\n\n### The key numbers\n\nThree things to remember. First: λ (lambda) is a team's expected goals for the match, the engine of the whole calculation. Second: the Poisson formula converts λ into the probability of a specific goal count. Third, and most important for your bankroll: a bet is only worth placing if your probability as a decimal multiplied by the odds is greater than 1. Below 1, skip it.\n\n## Why Correct Score Is So Hard to Call\n\nBefore you start calculating, get a feel for the scale of the problem. A match result gives you three options. Correct score gives you dozens: 0-0, 1-0, 0-1, 1-1, 2-1, and so on. You're picking exactly one from that pile, and a single extra goal kills your ticket.\n\n### Dozens of possible scorelines in one match\n\nEven capping things at 3-3, that's 16 combinations. Real matches sometimes end 4-3 or 5-1. Probability is spread across all of them, so even the single most common scoreline in a typical match only lands around 10–13% of the time, not 40–50%. That's a mathematical ceiling: consistently nailing the correct score is genuinely impossible in principle. The goal is to estimate probabilities more accurately than the bookmaker does.\n\n### Why apps and \"AI predictions\" don't deliver what they promise\n\nSearch \"correct score prediction\" and you'll drown in apps badged with AI: \"deep statistical analysis,\" \"92% confidence.\" The problem is they're black boxes. They don't show the method, don't explain where the numbers come from, and teach you nothing. You can't check their logic and you have no idea what that \"confidence\" is actually based on.\n\nOur approach is the opposite: we show all the maths so you can run the numbers yourself and make your own call. Transparent probability beats an opaque promise. No app or AI guarantees a scoreline, they give you probabilities at best, and below you'll learn to calculate those yourself without any middleman.\n\n## The Method: How to Predict the Score Yourself\n\nStep by step. This is the same method professional models and bookmakers use, just without the black box.\n\n### Step 1. Gather xG and recent form\n\nStart with expected goals (xG) for both teams across their last 5–10 matches: goals scored (xG) and conceded (xGA) separately. Why xG instead of raw goals? Because surface form lies. A team might win 1-0 while their opponent hit three posts, that's luck, not quality. xG measures the quality of chances created, not whether they went in, which makes it a far more honest measure of team strength.\n\n#### Where to find xG\n\nFree sources used by amateurs and professionals alike: Understat, FBref, SofaScore. All carry xG and xGA by team for the current season and recent fixtures. Take the average from the last 5–10 games to capture current form rather than stale season-wide figures. If a team has changed manager or lost their main striker, weight the recent matches more heavily.\n\n### Step 2. Calculate expected goals (λ) using attack and defence strength\n\nNow we convert form into λ for each team. The logic is straightforward:\n\n```\nλ for team = attack strength × opponent defence weakness × league average\nwhere  attack strength       = team's average goals scored ÷ league average\n       defence weakness      = opponent's average goals conceded ÷ league average\n```\n\nIn plain English: if a team scores 1.3 times the league average, their opponent concedes 1.1 times the league average, and the league average for home sides is 1.4 goals per match, then λ = 1.3 × 1.1 × 1.4 ≈ 2.0. You calculate λ separately for the home side and the away side. Across the top leagues, matches average around 2.7 goals in total, typically skewed toward the home side (roughly 1.5 vs 1.2).\n\n### Step 3. Run the Poisson distribution\n\nGoals are rare, independent events, which is why they fit the Poisson distribution well. The probability that a team with an expected goal rate of λ scores exactly k goals:\n\n$$P(k) = \\frac{\\lambda^k \\cdot e^{-\\lambda}}{k!}$$\n\n#### The Poisson formula in plain English\n\nDon't let the notation put you off. Plug in the expected goals λ and the number of goals k (0, 1, 2, 3), and you get the probability of exactly that many goals. Here e is Euler's number (roughly 2.718), and k! is the factorial: 3! = 3 × 2 × 1 = 6. Run the formula for k from 0 to 5, separately for each team.\n\n### Step 4. Build the full scoreline grid\n\nThe probability of a specific scoreline is the product: the home team's goal probability multiplied by the away team's goal probability (assuming independence). Multiply out every combination and you get the [complete grid](\u002Fbetting\u002Fcorrect-score-calculator) of probabilities for all scorelines.\n\n### Step 5. Find the most likely scorelines and the value\n\nThe grid shows which scorelines are most probable. But a probable scoreline is not the same as a profitable bet. A bet only has value if its [fair odds](\u002Fbetting\u002Fodds-converter) sit below what the bookmaker is actually offering. That's what the final section below covers.\n\n## Worked Example: Home λ = 1.6, Away λ = 1.1\n\nLet's run the numbers. Say it's a match where the home side is attack-minded and the visitors are mid-table: we expect 1.6 goals from the hosts and 1.1 from the away side. Run Poisson separately for each team.\n\n### Goal probabilities for each team\n\n| Goals k | P (home, λ=1.6) | P (away, λ=1.1) |\n|---|---|---|\n| 0 | 20.2% | 33.3% |\n| 1 | 32.3% | 36.6% |\n| 2 | 25.8% | 20.1% |\n| 3 | 13.8% | 7.4% |\n\n### Scoreline probabilities\n\nNow multiply the relevant combinations:\n\n| Score | Calculation | Probability |\n|---|---|---|\n| 1-1 | 0.323 × 0.366 | ~11.8% |\n| 1-0 | 0.323 × 0.333 | ~10.8% |\n| 2-1 | 0.258 × 0.366 | ~9.4% |\n| 2-0 | 0.258 × 0.333 | ~8.6% |\n| 0-0 | 0.202 × 0.333 | ~6.7% |\n\nThese figures come from the formula applied to this specific pair of λ values, purely to illustrate the method. They align well with real-world frequencies (1-1 around 11%, 1-0 around 10%, 2-1 around 8%). For your actual match, pull the exact numbers from the calculator.\n\n::chart-generic\n---\ntype: bar\ntitle: Top scoreline probabilities (example: λ 1.6 and 1.1)\nlabels: [\"1-1\", \"1-0\", \"2-1\", \"2-0\", \"0-0\"]\ndata: [11.8, 10.8, 9.4, 8.6, 6.7]\nyLabel: \"Probability, %\"\n---\n::\n\n### Video: Poisson in practice\n\nIf you'd rather watch than read through the formulas, here's a walkthrough of the Poisson method for scoreline prediction:\n\n::lazy-youtube\n---\nidEn: aDY1P4sV-tg\nidRu: RIN91ItfiRk\ntitle: How to predict the correct score using the Poisson method\n---\n::\n\n## What Actually Moves the Needle (and What Doesn't)\n\nThe model is only as good as the inputs you feed it. Here's what genuinely shifts λ and what's just noise.\n\n| Matters | Why | Doesn't matter \u002F trap |\n|---|---|---|\n| Recent xG and xGA (last 5–10 games) | Reflects current form better than the full season | Gut feel and hunches |\n| Home advantage (~+0.3–0.4 goals) | Home sides genuinely score more | \"The team is due\" (gambler's fallacy) |\n| Key injuries and suspensions | Directly change λ | Yesterday's big scoreline as a \"trend\" |\n| Motivation (European spots, relegation, rotation) | Shifts the intensity of play | \"Lucky\" round numbers |\n| Weather and pitch (heavy wind, rain) | Push the total down | Betting the brand, not the match |\n\n### Home advantage, injuries, motivation\n\nThese three factors adjust λ more than anything else. Home advantage adds roughly 0.3–0.4 goals for the hosts. A striker injury cuts their attacking λ: if a team's xG of 1.8 was built around one centre-forward, drop their λ by 0.3–0.4 without him. A meaningless final-day fixture with heavy rotation on both sides makes the game genuinely unpredictable, and the model is weaker here. Better to skip it altogether.\n\n### Traps: \"the team is due\" and gut instinct\n\nClassic gambler's fallacy: \"They haven't scored in five games, they must be due.\" They're not. Past matches don't influence the probability of the next goal. That's textbook [gambler's fallacy](\u002Fblog\u002Fsystem-bet-football-strategy). Round, tidy scorelines like 3-0 get overrated in people's heads, but they're actually rare. Calculate, don't guess.\n\n## Full Match Breakdown\n\nLet's put the method together in one worked example. Say \"Hosts\" (an attack-minded mid-table side) are playing \"Visitors\" (a cautious away team) in a league that averages 2.7 goals per game.\n\n### Calculating λ for both sides\n\nThe hosts score 1.9 at home on average; the league home average is 1.5, so their attack strength is 1.9 \u002F 1.5 = 1.27. The visitors concede 1.4 away from home; the league average is 1.2, so their defensive weakness is 1.4 \u002F 1.2 = 1.17. Expected goals for the hosts: λ = 1.27 × 1.17 × 1.5 ≈ 2.2. Now flip it: the visitors score 1.0 away, their attack strength is 1.0 \u002F 1.2 = 0.83; the hosts concede 1.1 at home, their defensive weakness is 1.1 \u002F 1.5 = 0.73, so the visitors' λ = 0.83 × 0.73 × 1.2 ≈ 0.73.\n\n### Reading the grid and drawing conclusions\n\nWith home λ at 2.2 and away λ at 0.73, the model surfaces 2-0, 1-0, and 2-1 at the top, not flashy 3-2s or 4-1s. Run those λ values through the [calculator](\u002Fbetting\u002Fcorrect-score-calculator), compare the top scorelines against the bookmaker's prices, and bet only where you have an edge. If the model is pointing hard at 2-0 and the bookmaker has it at an inflated price, that's your bet. No magic, just numbers.\n\n## Pre-Bet Checklist for Correct Score\n\nBefore you place the bet, run through this short list. It protects your bankroll better than any system.\n\n- Did you pull xG from the last 5–10 matches rather than the full season?\n- Did you account for injuries and suspensions of key players when setting λ?\n- Did you check motivation: is this a dead rubber with heavy rotation?\n- Did you calculate probability through a model rather than gut feel?\n- Is there value: does your probability as a decimal multiplied by the odds exceed 1?\n- Are you staking a sensible percentage of your bankroll rather than going all-in?\n- Are you mentally prepared for a downswing? That's just how this market works.\n\nIf even one answer is no, skip the match. Selection discipline beats bet volume every time.\n\n## Where the Method Breaks Down: Why a Model Isn't a Crystal Ball\n\nPoisson is a solid foundation, but it has structural weaknesses every professional knows about.\n\n### Poisson Underestimates Draws (and What to Do About It)\n\nThe main flaw: a pure Poisson model systematically underestimates the probability of draws, especially 0-0. In practice, teams defending a lead suffocate the game far harder than goal independence assumes. Professional models apply the Dixon-Coles adjustment, which lifts the probability of low-scoring draws (0-0, 1-1) closer to real-world frequencies. If you're targeting a draw, nudge the weight up slightly, especially in derbies and games between evenly matched sides.\n\n### Red Cards, Game States, and Low-Score Correlation\n\nThe model knows nothing about red cards, and a sending-off flips λ mid-game. It ignores game state entirely: a team sitting on a lead stops creating chances. Goals in reality are slightly correlated, something basic Poisson can't capture. For the central question of whether any model can predict the result, the honest answer is this: a model gives you probabilities, not prophecy. That applies equally to yourself and to any app claiming \"92% confidence.\"\n\n### How Accurate Is the Method, Really?\n\nEven a strong model hits the correct score only 10–14% of the time, because that's the mathematical ceiling of this market. The value isn't in \"hitting\" it. It's in finding overpriced odds. You don't win through frequent hits. You win by consistently betting at prices above fair value over a long enough sample. This is a long game, not a one-match miracle.\n\n## Turning Probabilities Into a Bet (Value Only)\n\nOne rule: only bet when your probability as a decimal multiplied by the odds is greater than 1.\n\nTake the example from our calculation. The model gives P(2-1) = 9.4%, so the fair price is 1 \u002F 0.094 ≈ 10.6. Now check the bookmaker's price. Offering 12.0? Then 0.094 × 12.0 = 1.13, greater than 1, that's [value](\u002Fbetting\u002Fvalue-bet-finder), bet it. Offering 8.0? Then 0.094 × 8.0 = 0.75, less than 1, the edge is against you, pass. Size your stake using the [Kelly Criterion](\u002Fbetting\u002Fkelly-calculator), not instinct.\n\nThe deeper pricing math (implied probability, [bookmaker margin](\u002Fbetting\u002Fimplied-probability), overround) is covered in a separate guide on [correct score odds](\u002Fblog\u002Fcorrect-score-odds-explained). For this framework, the value rule above is all you need.\n\n## Run the Numbers Yourself: Correct Score Calculator\n\nMultiplying out the grid by hand is tedious and easy to get wrong. Our tool does it instantly: enter expected goals (or each team's attack and defence strength) and get the probability of every scoreline alongside fair odds to compare against the bookmaker's price.\n\n👉 Open the [correct score calculator](\u002Fbetting\u002Fcorrect-score-calculator) and run through your next match. To understand the market itself and the types of bets available, start with the [correct score betting explained](\u002Fblog\u002Fcorrect-score-betting-explained) guide. To trade scorelines on an exchange, check out [strategies and lay trading](\u002Fblog\u002Fcorrect-score-betting-strategy). The [double chance calculator](\u002Fbetting\u002Fdouble-chance-calculator) and [draw no bet calculator](\u002Fbetting\u002Fdraw-no-bet-calculator) let you compare neighbouring markets, and the [over\u002Funder glossary](\u002Fglossary\u002Fbetting\u002Fover-under) covers totals terminology. All betting tools are in the [betting section](\u002Fbetting).",[29,32,35,38,41,44,47,50,53,56,59],{"answer":30,"question":31},"None of them guarantee anything. Apps that advertise AI predictions treat score forecasting as a black box, showing a confidence percentage with no explanation of the method behind it. The best any model can do, including ours, is assign probabilities. You're better off calculating the probability of each scoreline yourself using the Poisson model and actually seeing the logic, rather than trusting an opaque app.","Which app predicts the score accurately?",{"answer":33,"question":34},"There is no such website. Correct scores are high-variance by nature, and no source gets them right consistently. Instead of looking for a ready-made prediction, look for a transparent probabilistic model: a site that shows you where the numbers come from rather than just printing a scoreline. Our calculator shows the probability of every score using a proper formula.","Which website gives the correct score prediction?",{"answer":36,"question":37},"AI and statistical models produce probabilities, not prophecies. A model built on xG and Poisson distribution estimates the chances of each scoreline fairly well on average across many matches, but it cannot tell you the result of one specific game. A single late goal or red card breaks any prediction. Probability, yes. Guarantee, no.","Can AI predict the outcome of a football match?",{"answer":39,"question":40},"Look for low expected goals on both sides. When both teams have xG around 1.0 to 1.2, 0-0 and 1-1 float to the top. One important caveat: a clean Poisson model systematically underestimates draws, especially 0-0. Professional models add a Dixon-Coles correction that nudges the probability of low-scoring draws closer to real-world frequencies.","How do you predict a draw in the correct score market?",{"answer":42,"question":43},"You can't guarantee you never lose, and anyone promising otherwise is selling a fantasy. What you can do is minimize the bookmaker's edge working against you: only bet when you have value (when your estimated probability is higher than the one baked into the odds), keep your stake a small percentage of your bankroll, and stop chasing attractive scorelines. Discipline and value, not a system.","How do you bet without losing?",{"answer":45,"question":46},"Three steps. First, estimate the expected goals for each team (λ) using attack strength, the opponent's defensive weakness, and the league average. Then run the Poisson distribution to get the probability of scoring 0, 1, 2, 3 goals. Finally, multiply: the probability of a given scoreline equals the home team's goal probability multiplied by the away team's goal probability.","How is the correct score calculated?",{"answer":48,"question":49},"Across top leagues on average, 1-1, 1-0, and 2-1 come up most often, each roughly 8 to 12% of the time. After those come 0-0 and 2-0. The exact order depends on the match: in a defensive pair you see 0-0 and 1-0 at the top, in a high-scoring one it shifts to 2-1 and 2-2. There is no universal answer, it all comes down to expected goals.","What is the most common score in football?",{"answer":51,"question":52},"xG is significantly more reliable than raw form. Results-based form is misleading: a team might win 1-0 with the opposition hitting three posts, and that tells you nothing about their actual quality. Expected goals measure the quality of chances, not luck in finishing. Recent xG over the last 5 to 10 matches gives you a far more honest λ than a string of results.","Do you need xG to predict the score, or is recent form enough?",{"answer":54,"question":55},"Poisson describes goal distributions well on average across many matches, which is why bookmakers use it. But it has real weaknesses: it underestimates draws, ignores red cards, motivation, and game state, and it breaks down on extreme scorelines. It is a solid foundation for estimating probabilities, not a machine for precise predictions.","How accurate is the Poisson method?",{"answer":57,"question":58},"Free sources include Understat, FBref, and SofaScore. All of them carry xG scored and xGA conceded for each team across the season and for recent matches. Use the last 5 to 10 games to capture current form rather than a stale season average. Those numbers feed directly into your attack strength and defensive strength calculation.","Where do I get xG data for my prediction?",{"answer":60,"question":61},"Because there are dozens of possible scorelines and the probability of each one is small. Even the most common score only hits around 10 to 13% of the time. On top of that, the bookmaker margin on this market is higher than average. We break down the full pricing math in our separate guide on correct score odds.","Why are correct score odds so high?",[63,64],"ru","en",{"data":66,"body":67},{},{"type":68,"children":69},"root",[70,78,84,106,112,117,210,217,222,228,233,239,244,250,255,260,266,271,277,282,289,294,300,305,318,323,329,334,739,745,750,756,768,774,787,793,798,804,900,906,911,1028,1033,1039,1045,1050,1057,1063,1068,1185,1191,1196,1202,1215,1221,1226,1232,1237,1243,1254,1260,1265,1305,1310,1316,1321,1327,1332,1338,1343,1349,1354,1360,1365,1386,1407,1413,1418],{"type":71,"tag":72,"props":73,"children":75},"element","h2",{"id":74},"how-to-predict-the-correct-score-poisson-and-xg-2026",[76],{"type":77,"value":16},"text",{"type":71,"tag":79,"props":80,"children":81},"p",{},[82],{"type":77,"value":83},"I once bet on 2-1 in a match I was completely sure about. It finished 1-1. Frustrating, sure, but what got me afterward was this: when I ran that match through the model, it had flagged 1-1 as the most likely scoreline all along. I went with my gut instead of the numbers. That's exactly what this guide is about: how to predict correct scores with data in 2026, not instinct.",{"type":71,"tag":79,"props":85,"children":86},{},[87,89,95,97,104],{"type":77,"value":88},"One thing upfront: you cannot guarantee the correct score, and any site or app claiming otherwise is lying to you. What you ",{"type":71,"tag":90,"props":91,"children":92},"em",{},[93],{"type":77,"value":94},"can",{"type":77,"value":96}," do is calculate the probability of each scoreline and find bets where the bookmaker has mispriced things. Below is a step-by-step method using expected goals (xG) and the Poisson distribution, a live worked example with real numbers, and a built-in ",{"type":71,"tag":98,"props":99,"children":101},"a",{"href":100},"\u002Fbetting\u002Fcorrect-score-calculator",[102],{"type":77,"value":103},"calculator",{"type":77,"value":105},".",{"type":71,"tag":72,"props":107,"children":109},{"id":108},"tldr-how-to-calculate-the-correct-score-in-5-minutes",[110],{"type":77,"value":111},"TL;DR: How to Calculate the Correct Score in 5 Minutes",{"type":71,"tag":79,"props":113,"children":114},{},[115],{"type":77,"value":116},"Short version: take both teams' expected goals, run them through the Poisson formula, multiply the results together, and you get the probability of each scoreline. You only bet where your probability beats what's baked into the odds. Five steps:",{"type":71,"tag":118,"props":119,"children":120},"table",{},[121,140],{"type":71,"tag":122,"props":123,"children":124},"thead",{},[125],{"type":71,"tag":126,"props":127,"children":128},"tr",{},[129,135],{"type":71,"tag":130,"props":131,"children":132},"th",{},[133],{"type":77,"value":134},"Step",{"type":71,"tag":130,"props":136,"children":137},{},[138],{"type":77,"value":139},"What you do",{"type":71,"tag":141,"props":142,"children":143},"tbody",{},[144,158,171,184,197],{"type":71,"tag":126,"props":145,"children":146},{},[147,153],{"type":71,"tag":148,"props":149,"children":150},"td",{},[151],{"type":77,"value":152},"1",{"type":71,"tag":148,"props":154,"children":155},{},[156],{"type":77,"value":157},"Gather recent xG and form for both teams",{"type":71,"tag":126,"props":159,"children":160},{},[161,166],{"type":71,"tag":148,"props":162,"children":163},{},[164],{"type":77,"value":165},"2",{"type":71,"tag":148,"props":167,"children":168},{},[169],{"type":77,"value":170},"Calculate expected goals λ using attack and defence strength",{"type":71,"tag":126,"props":172,"children":173},{},[174,179],{"type":71,"tag":148,"props":175,"children":176},{},[177],{"type":77,"value":178},"3",{"type":71,"tag":148,"props":180,"children":181},{},[182],{"type":77,"value":183},"Run the Poisson distribution for 0, 1, 2, 3 goals",{"type":71,"tag":126,"props":185,"children":186},{},[187,192],{"type":71,"tag":148,"props":188,"children":189},{},[190],{"type":77,"value":191},"4",{"type":71,"tag":148,"props":193,"children":194},{},[195],{"type":77,"value":196},"Build the full scoreline grid (multiply the probabilities together)",{"type":71,"tag":126,"props":198,"children":199},{},[200,205],{"type":71,"tag":148,"props":201,"children":202},{},[203],{"type":77,"value":204},"5",{"type":71,"tag":148,"props":206,"children":207},{},[208],{"type":77,"value":209},"Spot the value scorelines and only bet where the edge exists",{"type":71,"tag":211,"props":212,"children":214},"h3",{"id":213},"the-key-numbers",[215],{"type":77,"value":216},"The key numbers",{"type":71,"tag":79,"props":218,"children":219},{},[220],{"type":77,"value":221},"Three things to remember. First: λ (lambda) is a team's expected goals for the match, the engine of the whole calculation. Second: the Poisson formula converts λ into the probability of a specific goal count. Third, and most important for your bankroll: a bet is only worth placing if your probability as a decimal multiplied by the odds is greater than 1. Below 1, skip it.",{"type":71,"tag":72,"props":223,"children":225},{"id":224},"why-correct-score-is-so-hard-to-call",[226],{"type":77,"value":227},"Why Correct Score Is So Hard to Call",{"type":71,"tag":79,"props":229,"children":230},{},[231],{"type":77,"value":232},"Before you start calculating, get a feel for the scale of the problem. A match result gives you three options. Correct score gives you dozens: 0-0, 1-0, 0-1, 1-1, 2-1, and so on. You're picking exactly one from that pile, and a single extra goal kills your ticket.",{"type":71,"tag":211,"props":234,"children":236},{"id":235},"dozens-of-possible-scorelines-in-one-match",[237],{"type":77,"value":238},"Dozens of possible scorelines in one match",{"type":71,"tag":79,"props":240,"children":241},{},[242],{"type":77,"value":243},"Even capping things at 3-3, that's 16 combinations. Real matches sometimes end 4-3 or 5-1. Probability is spread across all of them, so even the single most common scoreline in a typical match only lands around 10–13% of the time, not 40–50%. That's a mathematical ceiling: consistently nailing the correct score is genuinely impossible in principle. The goal is to estimate probabilities more accurately than the bookmaker does.",{"type":71,"tag":211,"props":245,"children":247},{"id":246},"why-apps-and-ai-predictions-dont-deliver-what-they-promise",[248],{"type":77,"value":249},"Why apps and \"AI predictions\" don't deliver what they promise",{"type":71,"tag":79,"props":251,"children":252},{},[253],{"type":77,"value":254},"Search \"correct score prediction\" and you'll drown in apps badged with AI: \"deep statistical analysis,\" \"92% confidence.\" The problem is they're black boxes. They don't show the method, don't explain where the numbers come from, and teach you nothing. You can't check their logic and you have no idea what that \"confidence\" is actually based on.",{"type":71,"tag":79,"props":256,"children":257},{},[258],{"type":77,"value":259},"Our approach is the opposite: we show all the maths so you can run the numbers yourself and make your own call. Transparent probability beats an opaque promise. No app or AI guarantees a scoreline, they give you probabilities at best, and below you'll learn to calculate those yourself without any middleman.",{"type":71,"tag":72,"props":261,"children":263},{"id":262},"the-method-how-to-predict-the-score-yourself",[264],{"type":77,"value":265},"The Method: How to Predict the Score Yourself",{"type":71,"tag":79,"props":267,"children":268},{},[269],{"type":77,"value":270},"Step by step. This is the same method professional models and bookmakers use, just without the black box.",{"type":71,"tag":211,"props":272,"children":274},{"id":273},"step-1-gather-xg-and-recent-form",[275],{"type":77,"value":276},"Step 1. Gather xG and recent form",{"type":71,"tag":79,"props":278,"children":279},{},[280],{"type":77,"value":281},"Start with expected goals (xG) for both teams across their last 5–10 matches: goals scored (xG) and conceded (xGA) separately. Why xG instead of raw goals? Because surface form lies. A team might win 1-0 while their opponent hit three posts, that's luck, not quality. xG measures the quality of chances created, not whether they went in, which makes it a far more honest measure of team strength.",{"type":71,"tag":283,"props":284,"children":286},"h4",{"id":285},"where-to-find-xg",[287],{"type":77,"value":288},"Where to find xG",{"type":71,"tag":79,"props":290,"children":291},{},[292],{"type":77,"value":293},"Free sources used by amateurs and professionals alike: Understat, FBref, SofaScore. All carry xG and xGA by team for the current season and recent fixtures. Take the average from the last 5–10 games to capture current form rather than stale season-wide figures. If a team has changed manager or lost their main striker, weight the recent matches more heavily.",{"type":71,"tag":211,"props":295,"children":297},{"id":296},"step-2-calculate-expected-goals-λ-using-attack-and-defence-strength",[298],{"type":77,"value":299},"Step 2. Calculate expected goals (λ) using attack and defence strength",{"type":71,"tag":79,"props":301,"children":302},{},[303],{"type":77,"value":304},"Now we convert form into λ for each team. The logic is straightforward:",{"type":71,"tag":306,"props":307,"children":311},"pre",{"className":308,"code":310,"language":77},[309],"language-text","λ for team = attack strength × opponent defence weakness × league average\nwhere  attack strength       = team's average goals scored ÷ league average\n       defence weakness      = opponent's average goals conceded ÷ league average\n",[312],{"type":71,"tag":313,"props":314,"children":316},"code",{"__ignoreMap":315},"",[317],{"type":77,"value":310},{"type":71,"tag":79,"props":319,"children":320},{},[321],{"type":77,"value":322},"In plain English: if a team scores 1.3 times the league average, their opponent concedes 1.1 times the league average, and the league average for home sides is 1.4 goals per match, then λ = 1.3 × 1.1 × 1.4 ≈ 2.0. You calculate λ separately for the home side and the away side. Across the top leagues, matches average around 2.7 goals in total, typically skewed toward the home side (roughly 1.5 vs 1.2).",{"type":71,"tag":211,"props":324,"children":326},{"id":325},"step-3-run-the-poisson-distribution",[327],{"type":77,"value":328},"Step 3. Run the Poisson distribution",{"type":71,"tag":79,"props":330,"children":331},{},[332],{"type":77,"value":333},"Goals are rare, independent events, which is why they fit the Poisson distribution well. The probability that a team with an expected goal rate of λ scores exactly k goals:",{"type":71,"tag":79,"props":335,"children":336},{},[337],{"type":71,"tag":338,"props":339,"children":342},"span",{"className":340},[341],"katex",[343,453],{"type":71,"tag":338,"props":344,"children":347},{"className":345},[346],"katex-mathml",[348],{"type":71,"tag":349,"props":350,"children":352},"math",{"xmlns":351},"http:\u002F\u002Fwww.w3.org\u002F1998\u002FMath\u002FMathML",[353],{"type":71,"tag":354,"props":355,"children":356},"semantics",{},[357,446],{"type":71,"tag":358,"props":359,"children":360},"mrow",{},[361,367,374,379,384,389],{"type":71,"tag":362,"props":363,"children":364},"mi",{},[365],{"type":77,"value":366},"P",{"type":71,"tag":368,"props":369,"children":371},"mo",{"stretchy":370},"false",[372],{"type":77,"value":373},"(",{"type":71,"tag":362,"props":375,"children":376},{},[377],{"type":77,"value":378},"k",{"type":71,"tag":368,"props":380,"children":381},{"stretchy":370},[382],{"type":77,"value":383},")",{"type":71,"tag":368,"props":385,"children":386},{},[387],{"type":77,"value":388},"=",{"type":71,"tag":390,"props":391,"children":392},"mfrac",{},[393,434],{"type":71,"tag":358,"props":394,"children":395},{},[396,409,414],{"type":71,"tag":397,"props":398,"children":399},"msup",{},[400,405],{"type":71,"tag":362,"props":401,"children":402},{},[403],{"type":77,"value":404},"λ",{"type":71,"tag":362,"props":406,"children":407},{},[408],{"type":77,"value":378},{"type":71,"tag":368,"props":410,"children":411},{},[412],{"type":77,"value":413},"⋅",{"type":71,"tag":397,"props":415,"children":416},{},[417,422],{"type":71,"tag":362,"props":418,"children":419},{},[420],{"type":77,"value":421},"e",{"type":71,"tag":358,"props":423,"children":424},{},[425,430],{"type":71,"tag":368,"props":426,"children":427},{},[428],{"type":77,"value":429},"−",{"type":71,"tag":362,"props":431,"children":432},{},[433],{"type":77,"value":404},{"type":71,"tag":358,"props":435,"children":436},{},[437,441],{"type":71,"tag":362,"props":438,"children":439},{},[440],{"type":77,"value":378},{"type":71,"tag":368,"props":442,"children":443},{"stretchy":370},[444],{"type":77,"value":445},"!",{"type":71,"tag":447,"props":448,"children":450},"annotation",{"encoding":449},"application\u002Fx-tex",[451],{"type":77,"value":452},"P(k) = \\frac{\\lambda^k \\cdot e^{-\\lambda}}{k!}",{"type":71,"tag":338,"props":454,"children":458},{"className":455,"ariaHidden":457},[456],"katex-html","true",[459,512],{"type":71,"tag":338,"props":460,"children":463},{"className":461},[462],"base",[464,470,478,484,490,496,502,508],{"type":71,"tag":338,"props":465,"children":469},{"className":466,"style":468},[467],"strut","height:1em;vertical-align:-0.25em;",[],{"type":71,"tag":338,"props":471,"children":476},{"className":472,"style":475},[473,474],"mord","mathnormal","margin-right:0.1389em;",[477],{"type":77,"value":366},{"type":71,"tag":338,"props":479,"children":482},{"className":480},[481],"mopen",[483],{"type":77,"value":373},{"type":71,"tag":338,"props":485,"children":488},{"className":486,"style":487},[473,474],"margin-right:0.0315em;",[489],{"type":77,"value":378},{"type":71,"tag":338,"props":491,"children":494},{"className":492},[493],"mclose",[495],{"type":77,"value":383},{"type":71,"tag":338,"props":497,"children":501},{"className":498,"style":500},[499],"mspace","margin-right:0.2778em;",[],{"type":71,"tag":338,"props":503,"children":506},{"className":504},[505],"mrel",[507],{"type":77,"value":388},{"type":71,"tag":338,"props":509,"children":511},{"className":510,"style":500},[499],[],{"type":71,"tag":338,"props":513,"children":515},{"className":514},[462],[516,521],{"type":71,"tag":338,"props":517,"children":520},{"className":518,"style":519},[467],"height:1.3879em;vertical-align:-0.345em;",[],{"type":71,"tag":338,"props":522,"children":524},{"className":523},[473],[525,530,735],{"type":71,"tag":338,"props":526,"children":529},{"className":527},[481,528],"nulldelimiter",[],{"type":71,"tag":338,"props":531,"children":533},{"className":532},[390],[534],{"type":71,"tag":338,"props":535,"children":539},{"className":536},[537,538],"vlist-t","vlist-t2",[540,723],{"type":71,"tag":338,"props":541,"children":544},{"className":542},[543],"vlist-r",[545,716],{"type":71,"tag":338,"props":546,"children":550},{"className":547,"style":549},[548],"vlist","height:1.0429em;",[551,583,597],{"type":71,"tag":338,"props":552,"children":554},{"style":553},"top:-2.655em;",[555,561],{"type":71,"tag":338,"props":556,"children":560},{"className":557,"style":559},[558],"pstrut","height:3em;",[],{"type":71,"tag":338,"props":562,"children":568},{"className":563},[564,565,566,567],"sizing","reset-size6","size3","mtight",[569],{"type":71,"tag":338,"props":570,"children":572},{"className":571},[473,567],[573,578],{"type":71,"tag":338,"props":574,"children":576},{"className":575,"style":487},[473,474,567],[577],{"type":77,"value":378},{"type":71,"tag":338,"props":579,"children":581},{"className":580},[493,567],[582],{"type":77,"value":445},{"type":71,"tag":338,"props":584,"children":586},{"style":585},"top:-3.23em;",[587,591],{"type":71,"tag":338,"props":588,"children":590},{"className":589,"style":559},[558],[],{"type":71,"tag":338,"props":592,"children":596},{"className":593,"style":595},[594],"frac-line","border-bottom-width:0.04em;",[],{"type":71,"tag":338,"props":598,"children":600},{"style":599},"top:-3.394em;",[601,605],{"type":71,"tag":338,"props":602,"children":604},{"className":603,"style":559},[558],[],{"type":71,"tag":338,"props":606,"children":608},{"className":607},[564,565,566,567],[609],{"type":71,"tag":338,"props":610,"children":612},{"className":611},[473,567],[613,660,666],{"type":71,"tag":338,"props":614,"children":616},{"className":615},[473,567],[617,622],{"type":71,"tag":338,"props":618,"children":620},{"className":619},[473,474,567],[621],{"type":77,"value":404},{"type":71,"tag":338,"props":623,"children":626},{"className":624},[625],"msupsub",[627],{"type":71,"tag":338,"props":628,"children":630},{"className":629},[537],[631],{"type":71,"tag":338,"props":632,"children":634},{"className":633},[543],[635],{"type":71,"tag":338,"props":636,"children":639},{"className":637,"style":638},[548],"height:0.927em;",[640],{"type":71,"tag":338,"props":641,"children":643},{"style":642},"top:-2.931em;margin-right:0.0714em;",[644,649],{"type":71,"tag":338,"props":645,"children":648},{"className":646,"style":647},[558],"height:2.5em;",[],{"type":71,"tag":338,"props":650,"children":654},{"className":651},[564,652,653,567],"reset-size3","size1",[655],{"type":71,"tag":338,"props":656,"children":658},{"className":657,"style":487},[473,474,567],[659],{"type":77,"value":378},{"type":71,"tag":338,"props":661,"children":664},{"className":662},[663,567],"mbin",[665],{"type":77,"value":413},{"type":71,"tag":338,"props":667,"children":669},{"className":668},[473,567],[670,675],{"type":71,"tag":338,"props":671,"children":673},{"className":672},[473,474,567],[674],{"type":77,"value":421},{"type":71,"tag":338,"props":676,"children":678},{"className":677},[625],[679],{"type":71,"tag":338,"props":680,"children":682},{"className":681},[537],[683],{"type":71,"tag":338,"props":684,"children":686},{"className":685},[543],[687],{"type":71,"tag":338,"props":688,"children":690},{"className":689,"style":638},[548],[691],{"type":71,"tag":338,"props":692,"children":693},{"style":642},[694,698],{"type":71,"tag":338,"props":695,"children":697},{"className":696,"style":647},[558],[],{"type":71,"tag":338,"props":699,"children":701},{"className":700},[564,652,653,567],[702],{"type":71,"tag":338,"props":703,"children":705},{"className":704},[473,567],[706,711],{"type":71,"tag":338,"props":707,"children":709},{"className":708},[473,567],[710],{"type":77,"value":429},{"type":71,"tag":338,"props":712,"children":714},{"className":713},[473,474,567],[715],{"type":77,"value":404},{"type":71,"tag":338,"props":717,"children":720},{"className":718},[719],"vlist-s",[721],{"type":77,"value":722},"​",{"type":71,"tag":338,"props":724,"children":726},{"className":725},[543],[727],{"type":71,"tag":338,"props":728,"children":731},{"className":729,"style":730},[548],"height:0.345em;",[732],{"type":71,"tag":338,"props":733,"children":734},{},[],{"type":71,"tag":338,"props":736,"children":738},{"className":737},[493,528],[],{"type":71,"tag":283,"props":740,"children":742},{"id":741},"the-poisson-formula-in-plain-english",[743],{"type":77,"value":744},"The Poisson formula in plain English",{"type":71,"tag":79,"props":746,"children":747},{},[748],{"type":77,"value":749},"Don't let the notation put you off. Plug in the expected goals λ and the number of goals k (0, 1, 2, 3), and you get the probability of exactly that many goals. Here e is Euler's number (roughly 2.718), and k! is the factorial: 3! = 3 × 2 × 1 = 6. Run the formula for k from 0 to 5, separately for each team.",{"type":71,"tag":211,"props":751,"children":753},{"id":752},"step-4-build-the-full-scoreline-grid",[754],{"type":77,"value":755},"Step 4. Build the full scoreline grid",{"type":71,"tag":79,"props":757,"children":758},{},[759,761,766],{"type":77,"value":760},"The probability of a specific scoreline is the product: the home team's goal probability multiplied by the away team's goal probability (assuming independence). Multiply out every combination and you get the ",{"type":71,"tag":98,"props":762,"children":763},{"href":100},[764],{"type":77,"value":765},"complete grid",{"type":77,"value":767}," of probabilities for all scorelines.",{"type":71,"tag":211,"props":769,"children":771},{"id":770},"step-5-find-the-most-likely-scorelines-and-the-value",[772],{"type":77,"value":773},"Step 5. Find the most likely scorelines and the value",{"type":71,"tag":79,"props":775,"children":776},{},[777,779,785],{"type":77,"value":778},"The grid shows which scorelines are most probable. But a probable scoreline is not the same as a profitable bet. A bet only has value if its ",{"type":71,"tag":98,"props":780,"children":782},{"href":781},"\u002Fbetting\u002Fodds-converter",[783],{"type":77,"value":784},"fair odds",{"type":77,"value":786}," sit below what the bookmaker is actually offering. That's what the final section below covers.",{"type":71,"tag":72,"props":788,"children":790},{"id":789},"worked-example-home-λ-16-away-λ-11",[791],{"type":77,"value":792},"Worked Example: Home λ = 1.6, Away λ = 1.1",{"type":71,"tag":79,"props":794,"children":795},{},[796],{"type":77,"value":797},"Let's run the numbers. Say it's a match where the home side is attack-minded and the visitors are mid-table: we expect 1.6 goals from the hosts and 1.1 from the away side. Run Poisson separately for each team.",{"type":71,"tag":211,"props":799,"children":801},{"id":800},"goal-probabilities-for-each-team",[802],{"type":77,"value":803},"Goal probabilities for each team",{"type":71,"tag":118,"props":805,"children":806},{},[807,828],{"type":71,"tag":122,"props":808,"children":809},{},[810],{"type":71,"tag":126,"props":811,"children":812},{},[813,818,823],{"type":71,"tag":130,"props":814,"children":815},{},[816],{"type":77,"value":817},"Goals k",{"type":71,"tag":130,"props":819,"children":820},{},[821],{"type":77,"value":822},"P (home, λ=1.6)",{"type":71,"tag":130,"props":824,"children":825},{},[826],{"type":77,"value":827},"P (away, λ=1.1)",{"type":71,"tag":141,"props":829,"children":830},{},[831,849,866,883],{"type":71,"tag":126,"props":832,"children":833},{},[834,839,844],{"type":71,"tag":148,"props":835,"children":836},{},[837],{"type":77,"value":838},"0",{"type":71,"tag":148,"props":840,"children":841},{},[842],{"type":77,"value":843},"20.2%",{"type":71,"tag":148,"props":845,"children":846},{},[847],{"type":77,"value":848},"33.3%",{"type":71,"tag":126,"props":850,"children":851},{},[852,856,861],{"type":71,"tag":148,"props":853,"children":854},{},[855],{"type":77,"value":152},{"type":71,"tag":148,"props":857,"children":858},{},[859],{"type":77,"value":860},"32.3%",{"type":71,"tag":148,"props":862,"children":863},{},[864],{"type":77,"value":865},"36.6%",{"type":71,"tag":126,"props":867,"children":868},{},[869,873,878],{"type":71,"tag":148,"props":870,"children":871},{},[872],{"type":77,"value":165},{"type":71,"tag":148,"props":874,"children":875},{},[876],{"type":77,"value":877},"25.8%",{"type":71,"tag":148,"props":879,"children":880},{},[881],{"type":77,"value":882},"20.1%",{"type":71,"tag":126,"props":884,"children":885},{},[886,890,895],{"type":71,"tag":148,"props":887,"children":888},{},[889],{"type":77,"value":178},{"type":71,"tag":148,"props":891,"children":892},{},[893],{"type":77,"value":894},"13.8%",{"type":71,"tag":148,"props":896,"children":897},{},[898],{"type":77,"value":899},"7.4%",{"type":71,"tag":211,"props":901,"children":903},{"id":902},"scoreline-probabilities",[904],{"type":77,"value":905},"Scoreline probabilities",{"type":71,"tag":79,"props":907,"children":908},{},[909],{"type":77,"value":910},"Now multiply the relevant combinations:",{"type":71,"tag":118,"props":912,"children":913},{},[914,935],{"type":71,"tag":122,"props":915,"children":916},{},[917],{"type":71,"tag":126,"props":918,"children":919},{},[920,925,930],{"type":71,"tag":130,"props":921,"children":922},{},[923],{"type":77,"value":924},"Score",{"type":71,"tag":130,"props":926,"children":927},{},[928],{"type":77,"value":929},"Calculation",{"type":71,"tag":130,"props":931,"children":932},{},[933],{"type":77,"value":934},"Probability",{"type":71,"tag":141,"props":936,"children":937},{},[938,956,974,992,1010],{"type":71,"tag":126,"props":939,"children":940},{},[941,946,951],{"type":71,"tag":148,"props":942,"children":943},{},[944],{"type":77,"value":945},"1-1",{"type":71,"tag":148,"props":947,"children":948},{},[949],{"type":77,"value":950},"0.323 × 0.366",{"type":71,"tag":148,"props":952,"children":953},{},[954],{"type":77,"value":955},"~11.8%",{"type":71,"tag":126,"props":957,"children":958},{},[959,964,969],{"type":71,"tag":148,"props":960,"children":961},{},[962],{"type":77,"value":963},"1-0",{"type":71,"tag":148,"props":965,"children":966},{},[967],{"type":77,"value":968},"0.323 × 0.333",{"type":71,"tag":148,"props":970,"children":971},{},[972],{"type":77,"value":973},"~10.8%",{"type":71,"tag":126,"props":975,"children":976},{},[977,982,987],{"type":71,"tag":148,"props":978,"children":979},{},[980],{"type":77,"value":981},"2-1",{"type":71,"tag":148,"props":983,"children":984},{},[985],{"type":77,"value":986},"0.258 × 0.366",{"type":71,"tag":148,"props":988,"children":989},{},[990],{"type":77,"value":991},"~9.4%",{"type":71,"tag":126,"props":993,"children":994},{},[995,1000,1005],{"type":71,"tag":148,"props":996,"children":997},{},[998],{"type":77,"value":999},"2-0",{"type":71,"tag":148,"props":1001,"children":1002},{},[1003],{"type":77,"value":1004},"0.258 × 0.333",{"type":71,"tag":148,"props":1006,"children":1007},{},[1008],{"type":77,"value":1009},"~8.6%",{"type":71,"tag":126,"props":1011,"children":1012},{},[1013,1018,1023],{"type":71,"tag":148,"props":1014,"children":1015},{},[1016],{"type":77,"value":1017},"0-0",{"type":71,"tag":148,"props":1019,"children":1020},{},[1021],{"type":77,"value":1022},"0.202 × 0.333",{"type":71,"tag":148,"props":1024,"children":1025},{},[1026],{"type":77,"value":1027},"~6.7%",{"type":71,"tag":79,"props":1029,"children":1030},{},[1031],{"type":77,"value":1032},"These figures come from the formula applied to this specific pair of λ values, purely to illustrate the method. They align well with real-world frequencies (1-1 around 11%, 1-0 around 10%, 2-1 around 8%). For your actual match, pull the exact numbers from the calculator.",{"type":71,"tag":1034,"props":1035,"children":1038},"chart-generic",{":title":1036,"type":1037},"{\"Top scoreline probabilities (example\":\"λ 1.6 and 1.1)\",\"labels\":[\"1-1\",\"1-0\",\"2-1\",\"2-0\",\"0-0\"],\"data\":[11.8,10.8,9.4,8.6,6.7],\"yLabel\":\"Probability, %\"}","bar",[],{"type":71,"tag":211,"props":1040,"children":1042},{"id":1041},"video-poisson-in-practice",[1043],{"type":77,"value":1044},"Video: Poisson in practice",{"type":71,"tag":79,"props":1046,"children":1047},{},[1048],{"type":77,"value":1049},"If you'd rather watch than read through the formulas, here's a walkthrough of the Poisson method for scoreline prediction:",{"type":71,"tag":1051,"props":1052,"children":1056},"lazy-youtube",{"idEn":1053,"idRu":1054,"title":1055},"aDY1P4sV-tg","RIN91ItfiRk","How to predict the correct score using the Poisson method",[],{"type":71,"tag":72,"props":1058,"children":1060},{"id":1059},"what-actually-moves-the-needle-and-what-doesnt",[1061],{"type":77,"value":1062},"What Actually Moves the Needle (and What Doesn't)",{"type":71,"tag":79,"props":1064,"children":1065},{},[1066],{"type":77,"value":1067},"The model is only as good as the inputs you feed it. Here's what genuinely shifts λ and what's just noise.",{"type":71,"tag":118,"props":1069,"children":1070},{},[1071,1092],{"type":71,"tag":122,"props":1072,"children":1073},{},[1074],{"type":71,"tag":126,"props":1075,"children":1076},{},[1077,1082,1087],{"type":71,"tag":130,"props":1078,"children":1079},{},[1080],{"type":77,"value":1081},"Matters",{"type":71,"tag":130,"props":1083,"children":1084},{},[1085],{"type":77,"value":1086},"Why",{"type":71,"tag":130,"props":1088,"children":1089},{},[1090],{"type":77,"value":1091},"Doesn't matter \u002F trap",{"type":71,"tag":141,"props":1093,"children":1094},{},[1095,1113,1131,1149,1167],{"type":71,"tag":126,"props":1096,"children":1097},{},[1098,1103,1108],{"type":71,"tag":148,"props":1099,"children":1100},{},[1101],{"type":77,"value":1102},"Recent xG and xGA (last 5–10 games)",{"type":71,"tag":148,"props":1104,"children":1105},{},[1106],{"type":77,"value":1107},"Reflects current form better than the full season",{"type":71,"tag":148,"props":1109,"children":1110},{},[1111],{"type":77,"value":1112},"Gut feel and hunches",{"type":71,"tag":126,"props":1114,"children":1115},{},[1116,1121,1126],{"type":71,"tag":148,"props":1117,"children":1118},{},[1119],{"type":77,"value":1120},"Home advantage (~+0.3–0.4 goals)",{"type":71,"tag":148,"props":1122,"children":1123},{},[1124],{"type":77,"value":1125},"Home sides genuinely score more",{"type":71,"tag":148,"props":1127,"children":1128},{},[1129],{"type":77,"value":1130},"\"The team is due\" (gambler's fallacy)",{"type":71,"tag":126,"props":1132,"children":1133},{},[1134,1139,1144],{"type":71,"tag":148,"props":1135,"children":1136},{},[1137],{"type":77,"value":1138},"Key injuries and suspensions",{"type":71,"tag":148,"props":1140,"children":1141},{},[1142],{"type":77,"value":1143},"Directly change λ",{"type":71,"tag":148,"props":1145,"children":1146},{},[1147],{"type":77,"value":1148},"Yesterday's big scoreline as a \"trend\"",{"type":71,"tag":126,"props":1150,"children":1151},{},[1152,1157,1162],{"type":71,"tag":148,"props":1153,"children":1154},{},[1155],{"type":77,"value":1156},"Motivation (European spots, relegation, rotation)",{"type":71,"tag":148,"props":1158,"children":1159},{},[1160],{"type":77,"value":1161},"Shifts the intensity of play",{"type":71,"tag":148,"props":1163,"children":1164},{},[1165],{"type":77,"value":1166},"\"Lucky\" round numbers",{"type":71,"tag":126,"props":1168,"children":1169},{},[1170,1175,1180],{"type":71,"tag":148,"props":1171,"children":1172},{},[1173],{"type":77,"value":1174},"Weather and pitch (heavy wind, rain)",{"type":71,"tag":148,"props":1176,"children":1177},{},[1178],{"type":77,"value":1179},"Push the total down",{"type":71,"tag":148,"props":1181,"children":1182},{},[1183],{"type":77,"value":1184},"Betting the brand, not the match",{"type":71,"tag":211,"props":1186,"children":1188},{"id":1187},"home-advantage-injuries-motivation",[1189],{"type":77,"value":1190},"Home advantage, injuries, motivation",{"type":71,"tag":79,"props":1192,"children":1193},{},[1194],{"type":77,"value":1195},"These three factors adjust λ more than anything else. Home advantage adds roughly 0.3–0.4 goals for the hosts. A striker injury cuts their attacking λ: if a team's xG of 1.8 was built around one centre-forward, drop their λ by 0.3–0.4 without him. A meaningless final-day fixture with heavy rotation on both sides makes the game genuinely unpredictable, and the model is weaker here. Better to skip it altogether.",{"type":71,"tag":211,"props":1197,"children":1199},{"id":1198},"traps-the-team-is-due-and-gut-instinct",[1200],{"type":77,"value":1201},"Traps: \"the team is due\" and gut instinct",{"type":71,"tag":79,"props":1203,"children":1204},{},[1205,1207,1213],{"type":77,"value":1206},"Classic gambler's fallacy: \"They haven't scored in five games, they must be due.\" They're not. Past matches don't influence the probability of the next goal. That's textbook ",{"type":71,"tag":98,"props":1208,"children":1210},{"href":1209},"\u002Fblog\u002Fsystem-bet-football-strategy",[1211],{"type":77,"value":1212},"gambler's fallacy",{"type":77,"value":1214},". Round, tidy scorelines like 3-0 get overrated in people's heads, but they're actually rare. Calculate, don't guess.",{"type":71,"tag":72,"props":1216,"children":1218},{"id":1217},"full-match-breakdown",[1219],{"type":77,"value":1220},"Full Match Breakdown",{"type":71,"tag":79,"props":1222,"children":1223},{},[1224],{"type":77,"value":1225},"Let's put the method together in one worked example. Say \"Hosts\" (an attack-minded mid-table side) are playing \"Visitors\" (a cautious away team) in a league that averages 2.7 goals per game.",{"type":71,"tag":211,"props":1227,"children":1229},{"id":1228},"calculating-λ-for-both-sides",[1230],{"type":77,"value":1231},"Calculating λ for both sides",{"type":71,"tag":79,"props":1233,"children":1234},{},[1235],{"type":77,"value":1236},"The hosts score 1.9 at home on average; the league home average is 1.5, so their attack strength is 1.9 \u002F 1.5 = 1.27. The visitors concede 1.4 away from home; the league average is 1.2, so their defensive weakness is 1.4 \u002F 1.2 = 1.17. Expected goals for the hosts: λ = 1.27 × 1.17 × 1.5 ≈ 2.2. Now flip it: the visitors score 1.0 away, their attack strength is 1.0 \u002F 1.2 = 0.83; the hosts concede 1.1 at home, their defensive weakness is 1.1 \u002F 1.5 = 0.73, so the visitors' λ = 0.83 × 0.73 × 1.2 ≈ 0.73.",{"type":71,"tag":211,"props":1238,"children":1240},{"id":1239},"reading-the-grid-and-drawing-conclusions",[1241],{"type":77,"value":1242},"Reading the grid and drawing conclusions",{"type":71,"tag":79,"props":1244,"children":1245},{},[1246,1248,1252],{"type":77,"value":1247},"With home λ at 2.2 and away λ at 0.73, the model surfaces 2-0, 1-0, and 2-1 at the top, not flashy 3-2s or 4-1s. Run those λ values through the ",{"type":71,"tag":98,"props":1249,"children":1250},{"href":100},[1251],{"type":77,"value":103},{"type":77,"value":1253},", compare the top scorelines against the bookmaker's prices, and bet only where you have an edge. If the model is pointing hard at 2-0 and the bookmaker has it at an inflated price, that's your bet. No magic, just numbers.",{"type":71,"tag":72,"props":1255,"children":1257},{"id":1256},"pre-bet-checklist-for-correct-score",[1258],{"type":77,"value":1259},"Pre-Bet Checklist for Correct Score",{"type":71,"tag":79,"props":1261,"children":1262},{},[1263],{"type":77,"value":1264},"Before you place the bet, run through this short list. It protects your bankroll better than any system.",{"type":71,"tag":1266,"props":1267,"children":1268},"ul",{},[1269,1275,1280,1285,1290,1295,1300],{"type":71,"tag":1270,"props":1271,"children":1272},"li",{},[1273],{"type":77,"value":1274},"Did you pull xG from the last 5–10 matches rather than the full season?",{"type":71,"tag":1270,"props":1276,"children":1277},{},[1278],{"type":77,"value":1279},"Did you account for injuries and suspensions of key players when setting λ?",{"type":71,"tag":1270,"props":1281,"children":1282},{},[1283],{"type":77,"value":1284},"Did you check motivation: is this a dead rubber with heavy rotation?",{"type":71,"tag":1270,"props":1286,"children":1287},{},[1288],{"type":77,"value":1289},"Did you calculate probability through a model rather than gut feel?",{"type":71,"tag":1270,"props":1291,"children":1292},{},[1293],{"type":77,"value":1294},"Is there value: does your probability as a decimal multiplied by the odds exceed 1?",{"type":71,"tag":1270,"props":1296,"children":1297},{},[1298],{"type":77,"value":1299},"Are you staking a sensible percentage of your bankroll rather than going all-in?",{"type":71,"tag":1270,"props":1301,"children":1302},{},[1303],{"type":77,"value":1304},"Are you mentally prepared for a downswing? That's just how this market works.",{"type":71,"tag":79,"props":1306,"children":1307},{},[1308],{"type":77,"value":1309},"If even one answer is no, skip the match. Selection discipline beats bet volume every time.",{"type":71,"tag":72,"props":1311,"children":1313},{"id":1312},"where-the-method-breaks-down-why-a-model-isnt-a-crystal-ball",[1314],{"type":77,"value":1315},"Where the Method Breaks Down: Why a Model Isn't a Crystal Ball",{"type":71,"tag":79,"props":1317,"children":1318},{},[1319],{"type":77,"value":1320},"Poisson is a solid foundation, but it has structural weaknesses every professional knows about.",{"type":71,"tag":211,"props":1322,"children":1324},{"id":1323},"poisson-underestimates-draws-and-what-to-do-about-it",[1325],{"type":77,"value":1326},"Poisson Underestimates Draws (and What to Do About It)",{"type":71,"tag":79,"props":1328,"children":1329},{},[1330],{"type":77,"value":1331},"The main flaw: a pure Poisson model systematically underestimates the probability of draws, especially 0-0. In practice, teams defending a lead suffocate the game far harder than goal independence assumes. Professional models apply the Dixon-Coles adjustment, which lifts the probability of low-scoring draws (0-0, 1-1) closer to real-world frequencies. If you're targeting a draw, nudge the weight up slightly, especially in derbies and games between evenly matched sides.",{"type":71,"tag":211,"props":1333,"children":1335},{"id":1334},"red-cards-game-states-and-low-score-correlation",[1336],{"type":77,"value":1337},"Red Cards, Game States, and Low-Score Correlation",{"type":71,"tag":79,"props":1339,"children":1340},{},[1341],{"type":77,"value":1342},"The model knows nothing about red cards, and a sending-off flips λ mid-game. It ignores game state entirely: a team sitting on a lead stops creating chances. Goals in reality are slightly correlated, something basic Poisson can't capture. For the central question of whether any model can predict the result, the honest answer is this: a model gives you probabilities, not prophecy. That applies equally to yourself and to any app claiming \"92% confidence.\"",{"type":71,"tag":211,"props":1344,"children":1346},{"id":1345},"how-accurate-is-the-method-really",[1347],{"type":77,"value":1348},"How Accurate Is the Method, Really?",{"type":71,"tag":79,"props":1350,"children":1351},{},[1352],{"type":77,"value":1353},"Even a strong model hits the correct score only 10–14% of the time, because that's the mathematical ceiling of this market. The value isn't in \"hitting\" it. It's in finding overpriced odds. You don't win through frequent hits. You win by consistently betting at prices above fair value over a long enough sample. This is a long game, not a one-match miracle.",{"type":71,"tag":72,"props":1355,"children":1357},{"id":1356},"turning-probabilities-into-a-bet-value-only",[1358],{"type":77,"value":1359},"Turning Probabilities Into a Bet (Value Only)",{"type":71,"tag":79,"props":1361,"children":1362},{},[1363],{"type":77,"value":1364},"One rule: only bet when your probability as a decimal multiplied by the odds is greater than 1.",{"type":71,"tag":79,"props":1366,"children":1367},{},[1368,1370,1376,1378,1384],{"type":77,"value":1369},"Take the example from our calculation. The model gives P(2-1) = 9.4%, so the fair price is 1 \u002F 0.094 ≈ 10.6. Now check the bookmaker's price. Offering 12.0? Then 0.094 × 12.0 = 1.13, greater than 1, that's ",{"type":71,"tag":98,"props":1371,"children":1373},{"href":1372},"\u002Fbetting\u002Fvalue-bet-finder",[1374],{"type":77,"value":1375},"value",{"type":77,"value":1377},", bet it. Offering 8.0? Then 0.094 × 8.0 = 0.75, less than 1, the edge is against you, pass. Size your stake using the ",{"type":71,"tag":98,"props":1379,"children":1381},{"href":1380},"\u002Fbetting\u002Fkelly-calculator",[1382],{"type":77,"value":1383},"Kelly Criterion",{"type":77,"value":1385},", not instinct.",{"type":71,"tag":79,"props":1387,"children":1388},{},[1389,1391,1397,1399,1405],{"type":77,"value":1390},"The deeper pricing math (implied probability, ",{"type":71,"tag":98,"props":1392,"children":1394},{"href":1393},"\u002Fbetting\u002Fimplied-probability",[1395],{"type":77,"value":1396},"bookmaker margin",{"type":77,"value":1398},", overround) is covered in a separate guide on ",{"type":71,"tag":98,"props":1400,"children":1402},{"href":1401},"\u002Fblog\u002Fcorrect-score-odds-explained",[1403],{"type":77,"value":1404},"correct score odds",{"type":77,"value":1406},". For this framework, the value rule above is all you need.",{"type":71,"tag":72,"props":1408,"children":1410},{"id":1409},"run-the-numbers-yourself-correct-score-calculator",[1411],{"type":77,"value":1412},"Run the Numbers Yourself: Correct Score Calculator",{"type":71,"tag":79,"props":1414,"children":1415},{},[1416],{"type":77,"value":1417},"Multiplying out the grid by hand is tedious and easy to get wrong. Our tool does it instantly: enter expected goals (or each team's attack and defence strength) and get the probability of every scoreline alongside fair odds to compare against the bookmaker's price.",{"type":71,"tag":79,"props":1419,"children":1420},{},[1421,1423,1428,1430,1436,1438,1444,1446,1452,1454,1460,1462,1468,1470,1476],{"type":77,"value":1422},"👉 Open the ",{"type":71,"tag":98,"props":1424,"children":1425},{"href":100},[1426],{"type":77,"value":1427},"correct score calculator",{"type":77,"value":1429}," and run through your next match. To understand the market itself and the types of bets available, start with the ",{"type":71,"tag":98,"props":1431,"children":1433},{"href":1432},"\u002Fblog\u002Fcorrect-score-betting-explained",[1434],{"type":77,"value":1435},"correct score betting explained",{"type":77,"value":1437}," guide. To trade scorelines on an exchange, check out ",{"type":71,"tag":98,"props":1439,"children":1441},{"href":1440},"\u002Fblog\u002Fcorrect-score-betting-strategy",[1442],{"type":77,"value":1443},"strategies and lay trading",{"type":77,"value":1445},". The ",{"type":71,"tag":98,"props":1447,"children":1449},{"href":1448},"\u002Fbetting\u002Fdouble-chance-calculator",[1450],{"type":77,"value":1451},"double chance calculator",{"type":77,"value":1453}," and ",{"type":71,"tag":98,"props":1455,"children":1457},{"href":1456},"\u002Fbetting\u002Fdraw-no-bet-calculator",[1458],{"type":77,"value":1459},"draw no bet calculator",{"type":77,"value":1461}," let you compare neighbouring markets, and the ",{"type":71,"tag":98,"props":1463,"children":1465},{"href":1464},"\u002Fglossary\u002Fbetting\u002Fover-under",[1466],{"type":77,"value":1467},"over\u002Funder glossary",{"type":77,"value":1469}," covers totals terminology. All betting tools are in the ",{"type":71,"tag":98,"props":1471,"children":1473},{"href":1472},"\u002Fbetting",[1474],{"type":77,"value":1475},"betting section",{"type":77,"value":105}]