Jump straight to the Frequently asked questions →
User Guide
From registration to everyday use
Football Analytics in brief: Football Analytics is a subscription-based football analytics web application. Using a statistical (Dixon-Coles) goal model, an xG correction, player and line-up effects, ELO-based team strength, market-specific machine learning and an ensemble model, it produces calibrated probabilities for every tracked match, across more than ten betting markets (1X2, double chance, draw no bet, both teams to score, over/under 0.5-4.5 goals, exact score), and an AI summarises in plain language what it sees and why for each estimate. The data is updated automatically every day. Access works on a monthly quota-based subscription (plan); the exact plans and prices are available live on the pricing page. The probabilities and analyses shown are for informational purposes only, do not constitute betting advice, and do not guarantee any future result.
This guide is written for people using Football Analytics as a user or a super user (group administrator). It walks through registration, choosing a plan and paying, inviting teammates, and then covers the tools available in everyday use and the account-related settings.
The guide deliberately does not cover data-management or system-operation (administrator) functions. These are not available at all in the user role; as a super user an Admin link appears in the “My account” section, through which you can manage the members of your own group (see section 4.4) — no other administrator function is reachable from there either.
Football Analytics is a football analysis web application that provides guidance on expected results, goal counts and other betting markets on the basis of statistical models and imported match data. The system processes data on teams, players, starting line-ups and referees. An estimate is not a single calculation: the base estimate built on the Dixon-Coles goal model is followed by an xG correction, player and line-up effects, ELO-based team strength (with the league-strength correction), market-specific machine learning, an ensemble model and finally calibration — the detailed order is described in section 7.11.
Football Analytics offers two roles to end users, one building on the other.
Which menu items you see in the interface depends primarily on your plan, not on your role: a user and a super user in the same group have access to the same analysis tools. What the super user gets in addition is group management as an extra permission — that is, inviting members and reviewing how the group uses the service.
You can create a new account on the Registration page. The process has two parts: first you choose a plan, then you enter your account details.
The next step depends on whether you chose a plan when registering.
If available, you will find a sign in with Google button on both the Registration and the Sign-in page. With this, your e-mail address is automatically considered confirmed, since Google has already verified it. In this case the system sets a random password for your account, which you can replace with your own at any time later on the Account settings / Change password page.
Confirming your e-mail address is mandatory: until this has happened you can neither sign in nor run an analysis. If the confirmation e-mail has not arrived, you can resend it from the sign-in page.
To sign in you need your e-mail address and your password.
If you have forgotten your password, click the Forgotten password link on the sign-in page and enter your e-mail address. The system sends you a link with which you can set a new password.
The system states precisely on the sign-in page what has happened if something does not work. The most common messages, word for word:
Football Analytics offers a choice of several plan types. The exact prices and monthly quotas can always be viewed, fully up to date, on the Choose a plan page; the table below shows the structure of the plans.
| Plan | Monthly analysis quota | Members | Characteristics |
|---|---|---|---|
| BASIC plan | smaller monthly allowance | 1 person | Entry-level plan for individual use. |
| PRO plan | larger monthly allowance | 1 person | The most frequently chosen plan, for regular individual use. |
| ADVANCED plan | largest monthly allowance | several people (group) | This is the plan that makes it possible to create a group: the buyer becomes a super user and can invite further members to use the shared allowance. |
| Extra quota | one-off credit | – | Can only be bought alongside an existing, active plan, if more analyses are needed temporarily. |
You can switch to a larger plan at any time on the Change plan page. If you only need more analyses temporarily, a one-off extra quota can also be bought alongside your existing, active plan, without having to replace the plan itself.
There is no automatic, recurring charge in Football Analytics: your subscription simply expires once the purchased period has elapsed. An approaching expiry is indicated by the header chip on every page (it turns red within 7 days), and on the Profile page a separate badge states how many days are left.
At the top of the My plan page a card shows your current plan and the “Valid until” date (or the fact that it is unlimited). If you have no active plan at the moment, the system draws your attention to this as follows: “You do not currently have an active plan. Choose one below to continue — you can only run analyses with an active plan.”
Below that, every publicly available plan appears as a card: plan name, possibly a “Most popular” highlight, a short description, the price (or “On request” if there is no fixed price), a list of features (typically the monthly quota, for example “{number} analyses / month”), and a “Switch to this / Buy” button.
If you already have an active plan, a separate “Extra quota” section also appears: this makes a one-off, immediate quota credit possible alongside your current plan (it does not replace it), and it can only be used for as long as your plan is valid. This is useful if your allowance is foreseeably going to run out in a given month, but you do not want to move permanently to a larger plan.
The Order page summarises the name and price of the plan you have chosen, then shows the following fields and cards in order:
If you buy a multi-person (group) plan, you automatically become a super user: your plan creates a group to which you can invite further members.
Every account has a primary group: this is where you registered and where your own plan runs. You can only change or buy a plan in your primary group.
If the super user of another group invites you and you already have your own account, no new parallel account is created: your existing account gains secondary membership of the new group. In that case you can use the quota of both groups, and your own primary plan remains unchanged.
Only a super user can send an invitation, only to their own group, and only if there is free space in the plan.
/group; if you do not see it in the menu, type the address directly into your browser).The invitee can accept the invitation by clicking the link in the e-mail they receive. There are two possible cases.
On the My group page the super user can review the primary and secondary members of the group separately, remove a member from the group, revoke an invitation that is still pending, and check via the Activity log menu item how much of the shared quota individual members of the group have used. In addition — via the Admin link in the “My account” section, on the Users screen — they can modify a member's display name and e-mail address, suspend their access, generate a new temporary password for them, and also delete their account. The super user does not learn the members' passwords, and every such operation is logged.
At the top of the page a KPI row shows the status of the group in four figures: “Max. members”, “Active members”, “Pending invitations”, “Free space”.
Above the invitation form the system states the rule on secondary membership in advance: “If an account already exists for the invited e-mail address (whether in another group or with its own plan), no new duplicate account is created — the existing account also gains secondary membership of this group, and its previous group and plan are kept.” If there is no free space, you see this instead of the form: “There is no free space in the group for a new invitation.”
The columns of the Members table are: Username, Name, E-mail, Role, Status (badge: “active” or “suspended”), Joined (date), and a “Remove member” button in each row.
If you have secondary members, a separate table also appears with the title “Secondary members”, with this explanation: “Users joining from another group whose own account and plan are elsewhere — they also consume this group's quota (the quota of the plan expiring soonest is used first).” Its columns are: Username, Name, Their own (primary) group, Status, Joined, and an “End membership” button.
The columns of the Invitations table are: E-mail, Status (badge: “Pending”, “Accepted”, “Revoked” or “Expired”), Created, Expires, and a “Revoke” button for invitations that are still pending.
After signing in, the Home page gives you a quick overview. At the top there is a quick starter headed “Which match are you interested in?”: you pick the home and away team, and the button runs the analysis immediately — this uses one unit of your allowance. For that reason “Match analysis” no longer appears as a separate card in the list below. Beneath it, cards show the other analysis tools actually available to you, grouped exactly like the left-hand menu (Analysis, Daily picks, My favourites) — only Match analysis is missing from the “Analysis” group, because the quick starter above replaces it. The greeting reads: “Welcome back, {your name}!”
The state of your subscription and your quota is not on the Home page but in the header bar of every page, as two compact indicators — so you never have to navigate back to the Home page to see what is left. The language selector is in the same place, at the right-hand edge of the header: switching keeps you on the page you were on, and the system remembers your choice.
In addition, the Home page shows the current daily and weekly AI summary and — if enabled — the card of your referral programme.
The left-hand menu is divided into three main parts.
Most other sections in the menu, related to data handling and system operation, are administrator tools. The administrator may, however, also assign certain measurement and diagnostic menu items — namely Daily odds, Daily hit rate and Analysis generation — to your plan or to your group; in that case a “Measurement / diagnostics” section appears in your menu as well. These menu items are not part of the free trial allowance, and the system does not add them automatically to any plan.
Next to the title of every page you will see the same pictogram as on the Home page card — so you can tell at a glance which tool you are in.
This chapter goes through the tools you can use for your daily work. Exactly which of them is available to you depends on your plan.
The most detailed analysis tool. You select a league or competition, then the home and away team, and start the analysis. The system produces a full analysis: expected result probabilities, goal count estimates, the head-to-head (H2H) record and further, deeper statistical metrics. Under the “Advanced settings” block you can specify how many earlier matches the system should look at (leave it empty for automatic), and you can also choose between the three options of the analysis source preset: “Full ensemble” (the default, using every available piece of data), “Without market odds/ML” (ignoring bookmaker prices entirely — useful if you want an opinion independent of the market, or if odds data is missing) and “Base model + ELO only” (relying on the least external data; the most stable option early in a season or with incomplete line-up data). The preset genuinely produces a different estimate, it does not merely change what is displayed.
The report shows the essentials at the top, and below them, under the heading “Detailed data”, come collapsible blocks (Most likely exact scores, Other outcomes, Form — the last matches of the two teams, Head-to-head matches, Line-up and injuries, Bookmaker prices and differences, Referee and additional indicators, Model details for experts), so you do not have to scroll through everything on a single, very long page. Each block has a short preview to the right of its title, so you can decide whether it interests you without opening it — see section 7.1 for details.
This tool compares the implied probabilities calculated from bookmakers' odds with the model's own estimate. Where the model sees a higher chance of an outcome than the odds suggest, it highlights the opportunity as a so-called value candidate.
Presents the form, key players, goalscoring potential and card risk of the players of a selected team.
Examines how the composition of the starting eleven, the chosen formation and the strength of the players in each position affect the estimate for the outcome of the match, comparing two teams.
With this tool you can retrieve the predictions for the imported matches of the coming days quickly and all at once, arranged into the usual betting market categories (1X2, both teams to score, over/under goals, double chance).
The result list is grouped first by country and then by competition: each country gets a heading, and under it each competition forms a block headed by the competition name, with the matches listed chronologically within it. Competitions of unknown country (typically international ones) come at the end of the list. This makes the offering easier to scan when you have ticked several leagues at once.
Shows the match-independent statistics of the selected referee, for example the proportion of home wins, draws and away wins, and the average number of cards. This is information only; on its own it is not automatically built into the prediction model.
When you open the page you will not see any data yet: first you have to select a referee. There are two ways to do this, and both lead to the same place.
If the selected referee has an upcoming, not-yet-played match, a direct link also appears below the statistics, taking you to the Match analysis page with the two teams of that match already filled in.
On the results page of an analysis you can save the metrics that matter to you into your digital notebook using the star button, and you can attach your own freely written note as well (up to 500 characters).
You can review your saved entries on the My favourites page. From there you can delete them individually or in bulk, display them in a printable view, or download them as a PDF.
Unlike the other tools, this one is not a manual query but a package compiled automatically every day at 06:00 Hungarian time: from the matches starting in the next 24 hours the system selects the up to five fixtures that reach the threshold set by the Provider on at least one betting market (for example home win, both teams to score, over 2.5 goals, away win).
Clicking the TOP 5 Analysis menu item simply opens the daily package — there is no filter, no date and no competition selection; the system does this for you automatically.
This tool shows every team of a competition in a single table, with the averages of the selected season. Two steps: choose the competition (in the drop-down the country comes first, followed by the name of the competition — so identically named leagues can still be told apart), then the season (only those with matches already played are offered). The table is produced when you click “Show”.
Choosing the competition and the season is free of charge — you can browse the drop-downs freely. Only the “Show” button consumes quota.
Three menu items are not part of the plans by default, but the administrator can assign them separately to your plan or to your group. If you have been given one of them, a “Measurement / diagnostics” section appears in your menu. These are not part of the free trial allowance, and the system does not add them automatically to any plan.
You do not have to start anything: the daily content is produced by itself, in Hungarian time.
The model itself is refreshed on two schedules: a light round runs every day (loading new matches, evaluating completed ones, refreshing odds), while the full learning chain runs weekly — that is when ELO, the feature table, market-specific machine learning, the large backtest and the calibration are rebuilt. That is why the estimates are fresh from day to day, while the internal parameters of the model change only when the weekly measurement proves that the change is better.
This chapter examines the RESULTS pages of the tools introduced in the previous chapter: exactly which figures, cards and tables you get after a run, and what they mean. Its purpose is that you should know not only WHERE to click, but also HOW to read what you see afterwards.
This is the most detailed report, so there is the most to explain here. The very first element of the report is a summary block: it states in one sentence what the model considers the most likely outcome, shows the 1 / X / 2 split on three bar charts, gives a one-sentence AI verdict, and then the main outcomes (most likely score, expected goals, how much data lies behind it). If the match is close, the warning “⚠️ This is a close match.” also appears here — in that case the model sees no meaningful difference between the two teams. Below the block comes the heading “Why does the system think this?”, followed by the eight collapsible blocks of “Detailed data”: Most likely exact scores, Other outcomes, Form — the last matches of the two teams, Head-to-head matches, Line-up and injuries, Bookmaker prices and differences, Referee and additional indicators, Model details for experts — the list below follows this order, working from top to bottom.
From the bookmaker's odds the module calculates the probability priced into an outcome (implied probability), then compares it with the model's own probability estimate. Where the model sees a higher chance than the odds imply, it proposes a value candidate on the basis of the difference between the two (the edge).
After the impact card (Impact / Goal potential / Card risk, see section 7.1) you get ONE sortable table listing all of the team's players. Clicking any column header re-sorts the whole list by that metric — so from the same table you can see who is currently in the best form (sorted by Player score), who has the highest chance of scoring (by Goal potential), and who carries the highest card risk (by Card risk), without switching between separate tables.
The table columns are: Player, Matches, Avg. minutes, Form (0-10), Goals, Assists, On target, Yellow, Red, Form score, Goal potential, Card risk.
An impact card appears for both teams (Starters / Attacking strength / Defensive strength / Coverage, see section 7.1), and below them a table of starting players with the columns: number, Player, Position, Strength score, Attacking strength, Defensive strength, Reliability (how reliable the estimate for that player is).
A separate card is produced for every analysed match: the team names, the league and the time in the header, the Data coverage on the right, and then a marking showing how many “value tips” there are for this match, or that there is “no value (no odds)”.
The card starts with a short KPI row (Peak score on the scoreline grid, Most likely score according to the 1X2, Both teams to score, Over 2.5 goals, Total corners, Total yellow cards, Expected goals, Data coverage), followed by a table by market, with the columns: Market, Recommended tip / distribution, Model %, Odds, Bookmaker's implied chance, Difference from the model, Value.
The markets this table covers are: Match result (1X2), Double chance (1X, 12, X2), Draw no bet (DNB), Both teams to score, Over/Under 0.5/1.5/2.5/3.5/4.5 goals, Correct score. If that day's measurement shows the corner or card estimate carries information, the list is also extended with the “Total corners” and “Total cards” markets (see section 7.12).
On each card the system also highlights one market with a ★ Best tip marking: priority goes to markets with a value candidate, and among those, to the one with the highest edge — in a tie, the one with the highest model probability is highlighted.
Below the name of the selected referee, four metrics appear, each on its own card: Matches officiated (how many completed matches they refereed), Home / Draw / Away (the proportion in which each outcome occurred in their matches), Avg. yellow/match and Avg. red/match. Beneath the cards you can see for how many of those matches card statistics were available at all (Matches with card data) — if that figure is much lower than the number of matches officiated, the card averages are built on less data.
On a saved entry you see: the names of the two teams, then in one row your username (this only appears in the administrator view), the league, the date of the match, a source marking (“Daily slip”, “TOP 5 Analysis”, “Toto slip (Hungarian)” or “Analysis” — depending on which page you saved it from), and the “expires: {date}” indication. Below that come the metrics you selected (starred) when saving, the list of markets (if there was a market with a value candidate, it appears here too with a green “value” label), and if you wrote your own tip, it appears under the “Your own tip:” label.
The source marking is a real link: “Analysis” takes you back to the Match analysis page with the saved pair of teams already filled in; “Daily slip” takes you to the Daily slip list page (there is no single-match deep link there, because that page handles several matches at once).
Its columns are: Time, User (only in a multi-person group), Feature, Result (a green “OK” or a red “blocked” badge), Content (the subject of the request, with team names rather than database identifiers). At the bottom of the list there is also a summary of how many entries the page is currently showing out of the total.
If this function is switched on, a card titled “AI summary” appears at the top of the Match analysis report: a few sentences in a human tone, putting the ALREADY calculated report data (1X2, expected goals, key metrics, sample size, referee profile) into plain language.
On the TOP 5 Analysis page the matches of the daily package appear in their own, simpler card format — not the same one as on the Daily slip. After the team names, the league, the time and Data coverage, the card shows per-market markers: next to each market examined, the model's percentage and the corresponding threshold, with a green tick if it was met and a red mark if it was not. There is no per-market odds table and no “★ Best tip” highlight on this card. If viewing the daily package deducts quota, the system first indicates this in a summary text and only opens the package after your confirmation.
An estimate is not a single calculation but a sequence of steps building on each other: after the base model come the xG correction, player and line-up data, ELO (with the league-strength correction), market-specific machine learning, the ensemble model, calibration, and finally the alignment of the goal thresholds and the re-synchronisation of the exact-score grid to the final 1X2. The “Effect by layer” table showing the step-by-step breakdown is an internal diagnostic view: available only in the administrator role, at the end of the “Model details for experts” block of the Match analysis.
After each step the table shows the 1 / X / 2 values, how many percentage points the home chance moved compared with the previous step, the expected goals, and in a fourth column what goal difference the current 1X2 actually corresponds to. That last column is what makes the published expected goal difference and what the published 1X2 really states comparable.
The system calculates an expected number of corners and cards for every match. Whether these appear in the analysis is decided separately for each market, by a daily measurement: if the estimate for that market (corners, or cards) is demonstrably worse than simply stating the average for the period, we hide it; otherwise it is shown. The two markets may therefore go different ways, and the estimate can also switch itself back on if it improves. The measurement is made on completed matches, against actual match statistics; the detailed measurement page is available in the administrator role.
The measurement page shows the following for both markets: the mean absolute error, the RMSE, the bias (a positive value means the system systematically overestimates), the actual average, and the hit rate of the published “Over / Under” recommendation.
The Toto slip (Hungarian) menu item shows the upcoming round of the Hungarian Toto: 13 regular matches plus the 14th bonus match, analysed with the same prediction core as the Daily slip — only in a 1 / X / 2 view.
You do not need to download or refresh anything. Every morning the system automatically downloads the upcoming round from the website of Szerencsejáték Zrt., matches the teams against its own database and then analyses them. By the time you open the page the slip is ready — which is why there is no “synchronise” or “run” button on it.
The columns of the table are: the number of the match, the two teams, the 1 / X / 2 probabilities, Sharpness, Data and the Suggested sign. Sharpness shows how clear-cut the tip is (0% = a pure coin toss, i.e. all three outcomes are roughly equally likely; 100% = completely clear-cut). This is NOT the same as the Data column, which tells you how much history the estimate is based on. The two are independent — an estimate built on plenty of data can still be flat.
The Suggested column usually contains a single sign: the most likely outcome. However, if the two most likely signs are within 10 percentage points of each other, the system shows both — the first in green (this is the tip) and the second in grey (this is the insurance). Below the table you can see how many such doubles there are and how many columns would be needed to cover them all. You do not have to play them; it is only an indication of where the estimate is uncertain.
If a match shows “not enough data”, it means there is not enough previous match history in the database for one of the two teams. In such cases the system deliberately does not give a tip: better nothing than something unfounded.
Every team of the competition appears in its own row, with the per-match averages of the season:
Your plan provides a monthly analysis quota. Every analysis actually run consumes one unit of this allowance, and in the Daily slip every single selected and analysed match counts as a separate unit. In the Toto slip and the TOP 5 Analysis a single operation uses several units — in those cases the system displays the exact number before running it, and deducts nothing until you accept.
On the Profile page you can review your username, role and group details, the current state of your quota, the details of your subscription (plan name, start and expiry date, status), and your past invoices and payments. From here you can change your display name, e-mail address or password, and this is also where you can initiate the deletion of your account.
More precisely, working from top to bottom, you see this:
After entering your current password you can set a new one, which must be at least 10 characters long and must contain a letter, a number and a special character; in addition, the system rejects passwords that are too common and easy to guess. After a successful password change the system signs you out, so you have to sign in again with the new password.
Under the Activity log menu item you can view your own quota-consuming queries: the time, the feature used and the result. As a super user you also see the usage of all members of your group in the same place, broken down by user, which helps you review who has used how much of the shared allowance.
The columns of the table are: Time, (User — only appears if you are a member of a multi-person group), Feature, Result, Content. In the Result column a green “OK” or a red “blocked” badge indicates whether the request ran or was rejected (for example because of a lack of quota).
When you click the Sign out button, the system asks for confirmation (“Are you sure you want to sign out?”) in order to avoid signing out accidentally.
You can initiate the deletion of your account on the Profile page. The system sends a confirmation link by e-mail, valid for 24 hours; the deletion becomes final when you click the link.
There can be several reasons, and the system always names the exact one. The most common: your account is not yet e-mail verified (“This account has not been e-mail verified yet…” — check your inbox, including the spam folder, or request a new link); you are already signed in on another device or browser, because a user or super user account can only have one active session open at a time (“This account is currently in active use on another device…” — sign out there, or wait a few seconds); or, after too many failed attempts, the system has temporarily locked the account for a few minutes (“Too many failed sign-in attempts.”). If your account or group has been suspended, the system states this explicitly too, and in that case an administrator needs to step in.
If there has been no click or activity in your account for 10 minutes, the system signs you out automatically for security reasons (“You were signed out automatically after 10 minutes of inactivity.”). This is not a fault — simply sign in again to continue.
No. With a user or super user account only one active session can be open at a time. If you want to sign in on another device, sign out of the first one first.
If this option is available, yes — in that case e-mail verification is automatic (Google has already confirmed it), and you receive a random password which you can later replace with your own on the Profile page.
You can sign in, look around the interface, and you also get 20 free trial analyses so you can try the system for real. You can split these 20 requests freely across the “Daily use” menu items (Match analysis, Bookmaker prices, Player form, Starting line-up impact, Daily slip, Toto slip, TOP 5 Analysis, Referee profile, Team statistics). After that the analysis functions stay locked until you buy a plan — the menu items remain visible, but their content does not open.
Without a plan you get a total of 20 free trial analyses, and you can use them on any “Daily use” menu item — Match analysis, Bookmaker prices, Player form, Starting line-up impact, Daily slip, Toto slip, TOP 5 Analysis, Referee profile, Team statistics — split however you like. Every analysis actually run takes one unit; in the Daily slip each selected match counts as a separate unit, so there the 20 trials give you 20 matches. The trial allowance is one-off and lifetime-scoped: it does not restart monthly, and once used up you can only continue by buying a plan. Deleting your account and registering again with the same e-mail address does not grant a new trial allowance either — we keep a record of this (solely to prevent abuse, in the form of an unreadable hash), see clause 2.8/a of the Privacy Notice. For the same reason, at most two accounts without a plan can be created from one network (IP address); this limit does not apply to registration with a plan.
The menu items tied to your plan become unavailable until you buy a plan again on the Change plan page. There is no automatic renewal; extending is always a manual step — the system never charges your card automatically.
With a card payment the plan is activated shortly after the successful transaction: the system checks the payment status every few minutes, and you can also start the check yourself with a button on the order page. With a bank transfer, activation only happens once the transfer has arrived and the system or an administrator has approved it — you will receive an e-mail about this.
Extra quota is a one-off, immediate quota credit alongside your EXISTING plan — it does not replace it, and it can only be used until your plan expires. It is worth considering if you need more analyses only temporarily (for example over a busy weekend), rather than on a lasting basis.
Yes. Every selected match that is actually analysed counts as an individual unit against your monthly quota — if you select 20 matches, that can use up to 20 units at once.
Because by default no competition is ticked in the filter — this is deliberate, so that a large query consuming a lot of quota does not start on its own. Tick the competitions you are interested in (or use the Select all button), then click Refresh list.
Of the up to five matches appearing in the daily package, only those that reach the high reliability (confidence) level set by the Provider deduct quota — the remaining items can be viewed within the package free of charge. If no item reaches that level, viewing the entire daily package is free.
No. If the system cannot find enough data for an analysis, it says so in place of the report, and the request is not counted against your quota — this applies equally to every tool (Match analysis, Bookmaker prices, Player form, Line-up impact, Daily slip, Toto slip, Referee profile, Team statistics).
No. Repeating a request for the same match on the same day, in the same tool, does not deduct further quota — after the first request that is actually charged, repeating it that day is free. The system also indicates this at the top of the analysis. Importantly, this applies per tool: the same match in Match analysis, in the Daily slip and in the TOP 5 Analysis counts as three separate requests, because each produces a different calculation and report. The benefit applies per user: if a fellow group member requests the same match, that counts as a separate request. For team-level tools (Player form) the same rule applies to the team.
If you have both a primary and a secondary group membership, the system always consumes the quota of the plan that expires SOONER first. With a group plan it may also be the case that a single member may only use a specified share of the group allowance on their own (Individual allowance, see section 9.1) — this protects the other members from one person using up the whole group's allowance.
The system indicates that no quota is available and offers you the option of changing plan or buying extra quota, with a “Buy a plan / continue payment” button.
Choosing the competition and the season is free of charge — you can browse the drop-downs freely, and nothing is deducted for that. Quota is consumed by the “Show” button: viewing one season of one competition costs 1 query, regardless of how many teams it contains — a single deduction gives you the data of every team in the competition. If you view the same competition and season again on the same day, no further unit is deducted; this is indicated above the button. The benefit applies per user: the same query run by another member of your group counts as a separate, quota-bearing request.
If a super user invites you to another group and you accept the invitation with your existing account, your account gains secondary membership of the new group, while your own primary plan and data remain unchanged.
24 hours. If you do not accept it in time, the invitation expires and the super user has to send a new one.
With a primary membership this affects your own account in relation to that group; with a secondary membership (“End membership”) only the secondary link ceases, and your own primary account and plan remain untouched.
Because they answer two different questions, and this is NOT a fault. “Most likely single score” shows which of ALL the possible final scores (0-0, 1-0, 2-1 and so on) has the greatest chance on its own — this is often a draw (for example 1-1), even if one of the teams is a clear favourite. The reason is simple: a draw concentrates on a few specific scores (0-0, 1-1, 2-2), whereas wins are spread across many (1-0, 2-0, 2-1, 3-1 and so on), so a single draw cell easily outranks any single winning scoreline.
“Most likely score according to the 1X2”, by contrast, does not look at the whole grid: the 1X2 first decides which outcome is the most likely (home win, draw or away win), and only within that outcome does the system look for the exact score with the highest probability.
From this follows the most important difference: the score according to the 1X2 can never contradict the match's main verdict. If the 1X2 makes a home win the most likely outcome, this figure shows a home win too — even if the peak of the grid is a draw at the same time. The sum of a group and the probability of individual scores are two separate questions, and the two KPIs deliberately answer them separately.
Two different things can get mixed up here: “expected goals” and the most likely specific score. Expected goals is an average — if the model says 2.6-0.7, it means that playing many similar matches would produce this many goals ON AVERAGE. The result of a single match can differ substantially, and this is not a flaw in the model but the nature of football: even a heavy favourite often wins 1-0.
The single most likely score is therefore almost always close — not because the model is unwilling to state the difference, but because the probability is spread out. The chance of a 4-0 stays small even if the chance of a home win is 80%: that 80% is distributed across many different scores (1-0, 2-0, 2-1, 3-0, 3-1 and so on), and none of them ends up dominant.
If, however, you see that the expected goal difference and the 1X2 figures do not agree with each other, the Match analysis page issues a separate warning. Both that warning and the step-by-step breakdown (section 7.11) are internal diagnostic views: they are available in the administrator role.
From the corner/card measurement (section 7.12). It compares the estimate with the actual match statistics on completed matches, and it does not show only the average error — that would not be interpretable on its own — but also what the error would be if the system simply stated the average for the period every time. If the estimate cannot beat that deliberately naive baseline, it carries no match-specific information, and the system hides that market from the analysis. The decision is taken separately for each market, every day, so corners and cards may go different ways and the estimate can also switch itself back on. The detailed measurement page is available in the administrator role.
The two tools work with different sensitivity thresholds: the Daily slip already marks a market as “VALUE” at a model-versus-odds difference (edge) of 0.5%, whereas the standalone Bookmaker prices page only does so from 3%. If you need more detailed, stricter filtering, look at the Bookmaker prices page or the Match analysis report.
The difference between goals actually scored and statistically expected goals (xG). A positive value indicates that the team is currently scoring MORE than the quality of its shots would justify (which may fall back over time); a negative value indicates underperformance (goals may still be to come).
It shows how much historical data, and how complete, the estimate in question is built on. Where the marking is “weak”, there is little data behind the figure, so it is better treated with caution.
No. The AI summary only puts the ALREADY calculated report data into human language; it does not change a single figure and does not feed back into the prediction — the report always states this at the bottom of the text.
So that instead of one very long page showing everything at once, you can open what you are actually interested in. Under “Detailed data” there are eight blocks: Most likely exact scores, Further outcomes, Form, Head-to-head matches, Line-up and injuries, Bookmaker prices and deviations, Referee and additional metrics, and Model details for experts. The data behind it is exactly the same — it is just easier to find your way around.
It highlights ONE row of the per-card market table: priority goes to markets with a value candidate, and among those, to the one with the highest edge — in a tie, the one with the highest model probability is highlighted. This is not the same as the “VALUE” badge — a match can have several “VALUE” markets, but only one of them can carry the “★ Best tip” highlight.
Yes, you can delete them individually or in bulk on the My favourites page. Entries are also deleted automatically 7 days after the date of the match, so if you want to keep one, download it as a PDF or print it before it expires.
You can always see your own requests. If you are a super user, you can also see the requests of ALL members of your group, broken down by user — this helps you keep track of how the shared quota is being used. If you are an ordinary user in a group, you only see your own requests, not those of other members.
They fall into two groups. The first cannot be turned off, because it is part of how your account and subscription work: e-mail address confirmation, password reset, payment confirmation, invoice, group invitation, and the warnings before your plan expires (7, 3 and 1 days beforehand). The second group is reminder-type and means at most one e-mail a week: if you do not have a plan yet, or if you have not signed in for a week. You can turn this off at any time on the Profile page, on the “Notification e-mails” card — turning it off does not affect the first group, so you will still receive your invoice and the expiry warnings.
Yes. After you click the link in the confirmation e-mail, the deletion is permanent and cannot be undone. Your personal data are archived, and your invoices remain in place because of statutory retention obligations, without any personal link.
24 hours. If you do not click it within that time, the deletion request expires and you have to start it again on the Profile page if you still want to delete your account.
| Version | Effective | What changed |
|---|---|---|
| 1.14 | 27 August 2026 | A full review of the four documents against the system's actual behaviour, and the correction of every difference found. The Analysis generation window is today and the preceding day, not the two preceding days. The guide now states which competitions the daily lists draw on. The KPI formerly called “Most likely win” is “Most likely score according to the 1X2”: the 1X2 decides the outcome first, so it can never contradict the match's main verdict. The corner and card estimates are shown or hidden per market on the basis of a daily measurement. The Daily slip has a limit of 40 matches per run, deducts quota match by match, and no longer has the quota card described earlier. The TOP 5 card has its own, simpler format. Card payments are activated shortly after the transaction, not instantly. The Terms and Conditions and the Privacy Notice were aligned with the code on partial deduction, the referral notification, the retention of the IP hash, the session time limits, the temporary password, the recipients of e-mails, the cookie table and the online withdrawal button. |
| 1.13 | 22 August 2026 | The free trial allowance now also has a per-network limit: at most two accounts without a plan can be created from one IP address. Registration that includes a plan is not affected. The reason is the same as for the e-mail-based record: without it, the one-off trial allowance could be taken again and again with new e-mail addresses from the same machine. The IP address itself is never stored, only a salted, irreversible hash of it (Terms and Conditions clause 5, Privacy Notice clause 2.8/a). The interface font size was also raised: body text from 13.5 to 16 pixels. |
| 1.12 | 22 August 2026 | The guide, the Terms and Conditions, the Privacy Notice and the FAQ are now available in German as well, alongside Hungarian and English. Outgoing e-mails (e-mail verification, password reset, order confirmation, payment and expiry notices, weekly reminders, account deletion) are also sent in German to recipients whose account language is set to German. The "Toto slip" menu item has been renamed "Toto slip (Hungarian)": Toto is a Hungarian betting game, and the menu did not say which country's round it shows. |
| 1.11 | 21 August 2026 | Documentation brought in line with how the system actually works. The corner and card estimates are now shown or hidden separately for each market, on the basis of a daily measurement — they are not permanently switched off (7.12). The step-by-step "Effect by layer" breakdown and the corner/card measurement page are internal diagnostic views, available in the administrator role; the warning about a discrepancy between the expected goal difference and the 1X2 still reaches every user (7.11). The administrator can also assign certain measurement menu items (Daily odds, Daily hit rate, Analysis generation) to a plan or a group; these are not part of the trial allowance (5.2). The Match analysis report is divided into expandable blocks, not tabs (FAQ). The Billing menu item requires a separately granted access right, it does not come with the super user role (5.2). |
| 1.10 | 12 August 2026 | The free trial allowance available without a plan has been raised from 3 to 20 analyses. Everything else about it is unchanged: it can be spent on any “Daily use” menu item, every analysis actually run takes one unit (in the Daily slip and the Toto, one per match), and the allowance is still one-off and lifetime-scoped — it does not renew monthly, and deleting the account and registering again with the same e-mail address does not grant a new one. The new number is reflected in chapter 8, in FAQ 10.1 and 10.2, on the landing page and in clause 5 of the Terms and Conditions. |
| 1.9 | 7 August 2026 | Alignment pass: several parts of the guide no longer matched the product. The Home page chapter (5.1) described the large “Subscription” and “Quota” cards, which are gone — both figures now sit in the header bar of every page, and the chapter describes them there. The menu chapter (5.2) now lists the actual three blocks (Analysis / Daily picks / My favourites) and mentions the page-title pictograms. The TOP 5 chapter (6.8) still said the daily package is compiled at 12 noon — it has been generated at 06:00 for some time. Menu items are now called by their current names throughout (Referee profile, Toto slip). The quota chapter (8) no longer says the free same-day repeat applies to “the same match”: it now states what counts as one unit in each tool — match, team, competition-season pair, match official or the whole Toto round. The Referee profile now consumes quota (6.6), which the guide previously did not state. |
| 1.8 | 7 August 2026 | New chapters on the Team statistics menu item (6.9 and 7.14). The tool shows every team of a competition in a single table, with the season averages: goals for and against, xG, corners, yellow and red cards, plus a form arrow comparing the average points of the last 5 matches with the earlier ones. In the competition drop-down the country now comes first, so identically named leagues can be told apart. Choosing the competition and the season is free; the “Show” button costs 1 query for one season of one competition — regardless of how many teams it contains — and the cost is stated above the button. Viewing the same competition and season again on the same day is free. The chapters also explain that a dash means missing data, not zero. |
| 1.7 | 6 August 2026 | New section on the Toto menu item (7.13): the weekly Toto round is downloaded and analysed automatically every morning, so there is no manual sync or run button on the page. The section explains the Sharpness and Data columns, the single vs. double suggested sign, and the token cost: the round costs as many tokens as it has analysable matches, the exact number is shown BEFORE you run it, and viewing it again on the same day is free. If your balance does not cover it, nothing is started and nothing is deducted. The quota chapter (8) now states that multi-unit operations always show the cost up front. |
| 1.6 | 5 August 2026 | The “Most likely score” KPI has been renamed and its meaning has changed: it is now “Most likely win”, and it shows the NON-DRAW scoreline with the highest probability — for either the home or the away team (7.1 and FAQ). Previously it showed the peak within the largest 1X2 group, which was misleading: a single specific away scoreline can outrank the best home one even when a home win is more likely in aggregate. The “Confidence level” KPI has been renamed “Data coverage” — the value is unchanged, but the old name looked like a measure of how reliable the estimate was, whereas it measures the amount of data (7.1). A new page reports the accuracy of the corner and card estimates, including a baseline comparison (7.12). On the Match analysis page an expandable “Effect by layer” table shows how much each calculation step changes the final result (7.11). The exact-score grid now satisfies the 1X2 probabilities, the goal thresholds and BTTS at the same time. |
| 1.5 | 3 August 2026 | Automated notification e-mails went live: plan expiry warnings 7, 3 and 1 day beforehand, plus a weekly reminder for users without a plan and for users who have not signed in for a week — the latter two can be switched off on the Profile page (FAQ 10.6). The free trial allowance is now documented: without a plan you get 3 trial analyses in total, usable on any “Daily use” menu item, one-off and for the lifetime of the account (chapter 8, FAQ 10.1 and a new 10.2 entry); the daily TOP 5 Analysis package is now generated at 06:00 in the morning (previously 12:00), and the window is also 06:00-06:00 (6.8 / 7.10); the TOP 5 menu item is visible without a plan too, with locked content. |
| 1.4 | 1 August 2026 | Free repeat within the same day: the same match, in the same tool, on the same day does not deduct further quota (chapter 8 and a new FAQ entry in 10.2); Daily slip: the competition list now shows only those competitions that have matches on the selected days, and matches are grouped by competition (6.5 / 7.5); Match analysis: a new “Own estimate vs market consensus” comparison on the Bookmaker prices tab (7.2). |
| 1.3 | 31 July 2026 | A short, self-contained summary paragraph in the introduction (also for AI search engines); two FAQ answers (sign-in error, scoreline grid vs most likely score) expanded so that they are understandable on their own without cross-references; English introduction text added (previously the Hungarian introduction appeared on the English page as well). |
| 1.2 | 30 July 2026 | Match analysis split into tabs (7.1); technical-term tooltips documented; Bookmaker prices “Best candidate” card and risk warning (7.2); Player model merged into ONE sortable table (7.3); Daily slip quota balance shown before selection and “★ Best tip” highlight (6.5 / 7.5); Referee profile link to the upcoming match (6.6 / 7.6); Favourites with real back-links and separate highlighting of expiry within 2 days (6.7 / 7.7). |
| 1.1 | 26 July 2026 | TOP 5 Analysis chapter added (6.8 / 7.10 / 8 / FAQ 10.2); Daily slip updated (no competition is pre-selected by default, the country is shown next to each competition, and the result card now also shows “Most likely score” alongside “Peak score on the scoreline grid”). |
| 1.0 | 24 July 2026 | Initial release — the public, web-viewable version of the user guide. |
The operation is running. Please wait for the result.
The page is loading. Please wait and do not navigate away.