AI Betting Predictions: What the Models Promise, What the Bookmaker Keeps

Updated October 2026

Licensed

Available in US

Fast payouts

18+ Only

The number a bookmaker quotes and the chance of the event actually happening are never the same thing. AI betting predictions have not changed that. A model can crunch form, head-to-heads, line-ups, xG, weather and injury data far faster than a human, and it can spit out a probability for every fixture on a Saturday’s coupon. The bookmaker still publishes the odds, and those odds still carry an overround — a margin built into every price that takes a cut off every winning and every losing one. So the work for a punter in 2026 is not finding the cleverest model. It is understanding what a price is doing, what value means once the margin is accounted for, and what staking discipline costs in lost bets before any system breaks even.

This guide is not a tip sheet. It does not name a winner, predict a scoreline, or sell a subscription to a bot that promises to. What it does is walk through how a price is built, how a value bet is judged, why a margin eats into any model the same way it eats into any human pick, and how the UK market around it — the Gambling Commission, GAMSTOP, the safer-gambling tools, the bonus rules and the credit-card ban — shapes what a punter can actually do.

As of 30 October 2026 the data was checked against the Gambling Commission’s public register.

Bankroll and Staking: The Cost a Punter Pays Before the First Bet

AI predictions promise an edge. The cost of getting that edge — and a lot of bettors never count this cost — is bankroll management. Bankroll is the float a punter sets aside for betting, kept separate from rent money, food money and any other commitment. The first rule, before staking plans and Kelly fractions, is that bankroll is money the punter can afford to lose in full. A model that wins 55 per cent of the time at average odds of 2.00 is still expected to lose in any given week, and a streak of 12 losers in a row on a 1/12 system happens to working systems the way a car crash happens to careful drivers.

A staking plan turns that bankroll into the size of each stake. The most common starter is flat staking — the same unit on every selection, usually between one and three per cent of bankroll — which spreads variance and stops a punt from becoming a gamble. The opposite is progressive staking, where the stake climbs after each loss to recover ground in one hit; that scheme looks clever on a spreadsheet until the bankroll hits zero before the recovery comes in, which is most of the time. AI prediction services do not change the maths. A model giving a 60 per cent strike rate is brilliant on paper and still subject to drawdowns that wipe out a flat-staking bankroll if the unit size is too high. The staking plan is what keeps the bettor in the game long enough for the edge to show.

There is also the simple cost of the subscription to the AI service itself, or the time spent feeding it data. A free model is free, but a paid one eats into any edge it provides. A service charging £30 a month has to deliver £30 a month of expected profit before it pays for itself, and that is before the bookmaker’s margin is taken into account. Punters who treat the AI as a labour-saving device and stick to disciplined staking usually do better than punters who treat the AI as a money-printing oracle and abandon the staking plan after a bad run.

Value, Price and the Margin That Eats Both

The single word every punter needs is value. A value bet is one whose odds are higher than the true probability of the event would justify. If a team has a 50 per cent chance of winning and the bookmaker is offering 2.20, the price implies 45.5 per cent — that is a value bet. If the price is 1.80, it implies 55.6 per cent, and even with a 50 per cent true chance the punter is paying for an edge that is not there. AI predictions help with the first half — estimating the probability — and the second half is the punter’s job. The bookmaker’s job is to make sure neither half is easy.

That is the margin. In a perfectly efficient two-way market — a tennis match with no vig — the two outcomes would sum to exactly 100 per cent. They never do. Take the example of a market priced at 1.90 and 1.90. The implied probabilities are 1 ÷ 1.90 = 52.63 per cent on each side, totalling 105.26 per cent. That extra 5.26 per cent is the bookmaker’s margin, sometimes called the overround, and it is built into every selection on the coupon, not slapped on as a fee at the end. The way to read the margin is simple: of every £1 staked across the whole market, the bookmaker expects to keep about 5p, regardless of which side wins.

Take an illustrative two-way market priced at 1.90 / 1.90. The implied probabilities are 1 ÷ 1.90 ≈ 52.6 per cent on each side, summing to roughly 105.3 per cent, so the margin sits in the 5–6 per cent band and the payout share in the 94–95 per cent band. That is a typical spread for a mainstream football 1X2 market at a UK-licensed bookmaker. Three-way markets (home / draw / away) carry a heavier margin because the bookmaker has three outcomes to balance; markets with more runner selections — correct score, first goalscorer, a tricast — carry heavier still, sometimes 12 to 20 per cent. The narrower the margin, the more of every pound staked returns to bettors in aggregate.

A value bet is the punter finding a price that is higher than the true probability the model estimates, once the margin is stripped away. A model giving Arsenal a 60 per cent chance against Brighton is no good if the bookmaker is offering 1.50 (implying 66.7 per cent) — the punter is paying for a price that already overstates Arsenal’s edge. The same model says the price is value if the bookmaker is offering 1.85 (54.1 per cent) and the punter can find a softer line elsewhere. AI helps with the probability; it does not move the price, and the price is what the punter pays.

AI in Practice: Form, Statistics, xG and the Limits of the Model

What a good AI prediction actually does is mundane, and that is part of its value. It reads form tables, results, expected goals (xG), shots on target, possession in final third, set-piece conversion, injury status and suspension lists. It may track weather, referee tendencies, travel distance, days since the last match, the strength of the opposition’s last five fixtures and a hundred other variables. The output is a probability per outcome, sometimes per market. A human tipster with a notebook can do most of this. An AI does it in seconds and updates it as team news breaks.

The limits are equally mundane. A model trained on five seasons of Premier League data has no opinion on a newly promoted side’s tactical shift under a new manager. A model trained on the Premier League is not a model for the Brazilian Série B. A model that uses bookmaker odds as an input is just echoing the market, which is the most efficient pricing engine in the world on public events. The honest framing is that AI predictions are tools, not oracles, and the punter’s job is to know when the tool is the right shape for the question being asked.

A useful test for any AI prediction service is to back-test it on settled markets — does it pick winners at a higher rate than the market-implied probability? — and to forward-test it with a separate paper bankroll before any real money goes near it. A model that cannot beat the market in back-testing will not beat it in real betting, because the conditions are softer in back-testing. A model that beats the market in back-testing but loses money on its staking fees, its subscription costs or the bankroll swings has been correctly priced by the market too.

In-Play and Live Betting: AI’s Hardest Question

In-play betting is the hardest market to predict and the one AI tipsters love to talk about. The argument goes that a model can react faster than a human to a red card, a tactical switch or a shift in momentum, and that the lag between event and price creates a window of value. There is some truth in that, on the part of liquidity — major bookmakers price football within a second or two of a major event — and there is also a lot of marketing.

The bookmaker’s margin on in-play markets is heavier than on pre-match, sometimes 7 to 10 per cent on mainstream football, because the model the bookmaker runs is the same kind of model the punter is paying for, with the same kind of data, only running on faster hardware. The value window, if it exists, is narrow, and a punter paying 7 per cent overround every time they strike is bleeding that margin back out before the edge can show. AI predictions are useful in-play for what they remove — the impulse to back a team that has just scored, which is rarely value — and for what they confirm — that the price after a goal is usually shorter than the new probability justifies. The punter’s edge in-play comes from restraint as often as from speed.

Combining Selections: Accumulators, Bet Builders and the Cost of Multiplying Margins

AI tipsters often sell accumulator bets as a high-payout route. A four-fold accumulator at 2.0, 2.0, 2.0, 2.0 pays 16.00 — a £10 stake returns £160. Each leg has its own margin, and the multiplied price carries the multiplied margin. The overround compounds. A four-fold with a 5 per cent margin per leg has a combined margin in the region of 19 per cent, not 20 — the math is one minus the payout shares — and the payout share drops into the low 80s. A punter has to win all four legs to be paid, and the combined edge the AI has to clear is far heavier than on a one-stand bet.

Bet Builders are a single-payout accumulator on correlated selections from one match — a home win, both teams to score, over 2.5 goals. They pay more than the same selections placed as singles because the bookmaker can correlate the selections and reduce the price accordingly; the margin on a Bet Builder is heavier than on a singles market on the same match, often 8 to 12 per cent. AI predictions on a Bet Builder are usually picking the easiest-looking outcomes and hoping the correlation pricing is wrong. Sometimes it is. Often it is not.

The honest framing: accumulators and Bet Builders are the highest-margin products on the bookmaker’s site, and the highest-stakes lesson for a punter who treats AI predictions as a route to a big payout. They are also the most fun, and betting should be fun before it is anything else. The cost of that fun is a margin that an AI prediction cannot beat by picking the right matches, because the matches are not where the margin sits. Pick carefully.

What the UK Market Wraps Around the Bet

The Gambling Commission is the regulator, and it sits inside the Department for Culture, Media and Sport as a non-departmental public body. Its job is to license the whole commercial market in Great Britain under the Gambling Act 2005 and to enforce the Licence Conditions and Codes of Practice. Northern Ireland is outside the 2005 Act and runs on older law, so a Great Britain licence does not cover a Northern Ireland customer and the Commission is not the regulator there. A bookmaker serving customers in Great Britain needs a Commission licence wherever it is based; a Curaçao, Maltese or Gibraltar licence is not a substitute, and a GB-licensed operator is not the only GB-licensed operator. The Commission’s public register, searchable by business name, trading name, domain or account number, is the whole test of who holds one.

Identity verification happens before a customer can deposit or gamble at a GB-licensed operator — name, address and date of birth, in force since 7 May 2019, including with a free bet or bonus. A credit card is not a valid way to fund gambling in Great Britain, in force since 14 April 2020, and that ban extends to credit-card-funded payments through an e-wallet. Debit cards and bank-based payments are not covered by the ban. A financial vulnerability check applies once a customer’s deposits minus withdrawals exceed £150 in a rolling 30 days, in force since 28 February 2025; the check is a public-record check (bankruptcy, county court judgments and the like), not an income check and not a credit reference check. Financial risk assessments — the affordability checks of the press — have been announced in a staged plan, with no start date.

GAMSTOP is the national multi-operator online self-exclusion scheme, and a licence condition for online operators since 31 March 2020. The exclusion periods are six months, one year, five years, or five years with auto-renewal, and after the minimum period the exclusion stays in place for a further seven years unless the customer asks GAMSTOP to remove it. GAMSTOP covers online sites and apps; betting shops run a separate self-exclusion service. The safer-gambling tools at most GB-licensed bookmakers — deposit limits, time-outs, product blocks, self-exclusion — sit on top of GAMSTOP and are not a replacement for it.

Advertising Rules and the Bonus Cap

Advertising is not licensed by the Commission; it runs through the CAP Code (non-broadcast) and BCAP Code (broadcast), enforced by the ASA, alongside the Commission’s marketing conditions. The CAP Code section 16 sets the social responsibility line — marketing must not be of strong appeal to under-18s, and no one who is or seems to be under 25 may be featured gambling or play a significant role in a gambling advertisement, with an exception for marketing placed where a bet can be made directly. The rule concerns who may appear in the advertisement, not who may bet.

The bonus rules in force since 19 January 2026 cap wagering requirements on bonuses at 10x — a £10 bonus at 50x needed £500 of qualifying bets, and under the cap no bonus can carry more than a 10x playthrough. Mixed-product promotions, where an offer requires gambling on two or more product types such as betting, casino, bingo and lottery (“bet on football to get free spins”), are banned. The cap and the mixed-product ban apply to GB-licensed operators; an offshore brand is not bound by them, and that is part of the cost of using one. A bettor who takes a 50x wagering bonus at an offshore site has fewer of the protections the GB market wraps around the offer.

The whistle-to-whistle TV ban on betting advertisements during live sport before the 9pm watershed, in force since 1 August 2019, is an industry code, not statute; horse and greyhound racing are excluded. The Premier League clubs’ voluntary withdrawal of gambling sponsorship from the front of matchday shirts, from the 2026/27 season, is voluntary. None of this is a guarantee that an offer a punter sees is fairly priced; the price carries its own margin, and the offer carries its own conditions.

Offshore Sites and What a Punter Loses by Using One

Providing gambling facilities in Great Britain without the licence the Act requires is an offence under section 33 of the Gambling Act 2005, aimed at the operator. No offence aimed at a customer who merely bets on an unlicensed site was found, and a punter who does so is not committing a criminal act under that section. What the punter is doing is opting out of the protections the GB market provides: no GAMSTOP coverage, no Commission complaints route, no approved dispute-resolution body, no British consumer redress if funds are withheld, and no obligation on the operator to honour the 10x bonus cap or to run the £150 vulnerability check. A bookmaker refusing a withdrawal at an offshore site is not answerable to the Commission; a GB-licensed bookmaker refusing one is, and that is the difference.

Enforcement is disruption, not blocking. The Commission issues cease-and-desist notices, refers sites to search engines and works with registrars, hosts and payment providers to make an offshore site harder to find and harder to fund. Court-order DNS blocking powers have been given or are about to be given; ISP blocking is not in operation as a routine measure, and a punter using a VPN to reach an offshore site is not, on its own, an offence. The framing is honest: licensed means checks, limits, GAMSTOP, the bonus cap and the credit-card ban, and also identity verification before the first bet; unlicensed means none of those protections and a real risk of losing funds if the operator decides not to pay out.

Tax: What a Punter Pays and What the Operator Pays

The punter pays no tax on betting winnings in Great Britain. That is general information, not tax advice, and a professional gambler should ask HMRC or an adviser. General Betting Duty is the operator’s tax on gross profits from betting, currently 15 per cent, and from 1 April 2027 a remote betting rate of 25 per cent applies within it, while remote bets on UK horse racing, bets on self-service betting terminals in licensed premises, pool betting and in-person betting stay at 15 per cent. Until 1 April 2027 the remote rate stays at 15 per cent; whether operators pass duty on through odds is not established, and the punter is not charged duty. The Commission’s statutory levy is a separate charge on operators, not a tax on punters. Spread betting is regulated by the Financial Conduct Authority, not the Commission, and spread betting losses can exceed deposits.

A Worked Look at a Two-Way Market

Take an illustrative two-way market priced at 1.90 / 1.90 (fractional 9/10). The implied probabilities are 1 ÷ 1.90 on each side, summing to roughly 105.3 per cent. The bookmaker’s margin, the overround, sits in the 5–6 per cent band; the payout share, the proportion of every pound staked that returns to bettors in winnings, sits in the 94–95 per cent band. That is a typical spread for a mainstream football 1X2 market at a UK-licensed bookmaker.

The cost to a punter over many bets is straightforward. A flat-staking punter placing £10 on a hundred such markets at a true 50/50 strike rate will, in aggregate, lose the margin — roughly £5 of every £100 staked, before any edge the punter or the AI has. A model giving 55 per cent on the same market still has to beat that 5 per cent margin before any profit shows, which it does on paper but only if the model’s probability is correctly calibrated. An over-confident model that gives 65 per cent on a 50/50 event loses money despite being “right more often than the half”, because the price implies a higher probability than the model gives and the bettor is still paying the margin.

That is the cost AI predictions face in real betting, and it is why no tip — AI, human, or otherwise — is ever certain. The margin eats the edge before the punter sees it, and the staking plan decides what is left.

A Look at the Reviewed Bookmakers

Each of the brands below holds a Gambling Commission remote licence for Great Britain, and each is presented as a reviewed bookmaker rather than a recommendation. The price, the offer, the in-play product and the safer-gambling tools differ from brand to brand, and the gap between a brand’s marketing and its terms is where most of the cost of any welcome bonus sits.

The verdict under each block is the page’s own judgement of the brand for the use case the block establishes — what kind of punter it suits, and what the punter pays for that fit.

BetGoodwin — Racing-First Pricing and a Modest Welcome Setup

BetGoodwin pays Best Odds Guaranteed on all UK and Irish horse and greyhound racing, up to £1,000, and that is the spine of the offer for a punter whose weekend is built around the racing pages. The brand is run by Goodwin Racing Limited and operates under a Commission remote licence at account 2927. The site is racing-led, and a punter looking for football in-play depth or a deep Bet Builder catalogue is shopping in the wrong aisle. For a racing punter who wants BOG on every UK and Irish meeting without a stake ceiling baked into the headline, BetGoodwin earns its place on the shortlist.

Betway — Free Bets With a Tight Qualification

Betway’s welcome setup for new UK and Ireland customers is £40 in free bet tokens, four tokens of £10 each, credited on settlement of a first bet of £10 or more at minimum odds of 2.0. The expiry is 7 days, and deposits must come from a debit card. The qualification is tight — minimum odds of 2.0 means the qualifying bet is not a shoo-in, and the 7-day expiry on the tokens means a slow punter loses them — and the debit-card-only deposit rules out the e-wallet route some punters prefer. For a punter happy with a single qualifying bet at evens or above and willing to use the tokens inside a week, the deal does what it says. The cost is the discipline of the clock.

bet365 — In-Play Depth, Streaming and a Wide Feature Set

bet365 is the in-play reference. In-play betting and live streaming of events worldwide sit at the centre of the product, alongside Bet Track, Match Live and a Live Golf Tracker for following bets in play. Hillside (UK Sports) ENC operates under a Commission remote licence at account 55148. The brand is the deepest single-product bookmaker on the GB market, and the cost of that depth is a heavier price for casual punters — a wider feature set is a wider marketing surface, and the welcome offer terms need reading with the same care as anywhere else. For a punter who bets in-play, wants live streaming and is willing to navigate a deeper site, bet365 is the obvious shortlist pick.

William Hill — BOG, WHTV and a Long-Standing Racing Product

William Hill pays Best Odds Guaranteed on UK and Irish horse racing from 8 AM on the day of the race, on both single and multiple bets, and the rule is explicit: it does not apply to ante-post, tote or pari-mutuel bets. WHTV streaming requires a funded account or a bet in the last 24 hours, and live racing requires a bet of £1 or more on a qualifying race; Betting TV shows selected UK and US horse racing and UK greyhounds free. The brand is the broadest mainstream bookmaker on the GB market, and the cost is a heavier price for niche markets than a specialist. For a racing punter who wants the BOG from 8 AM and a free-to-air racing feed, William Hill is the standard pick.

Lottoland — A Lottery-Side Hybrid

Lottoland’s UK app combines casino, lottery betting, sports betting and scratchcards, and the licence for customers in Great Britain sits on the brand. Commission account 38991 is held by EU Lotto Limited. The product is wider than a pure sportsbook, and a punter whose bet is on a sportsbook alone is paying for lottery and casino features they may not use. For a punter who mixes sports with lottery and wants a single wallet, Lottoland is a fit; for a punter whose betting is football-only, the broader product is a distraction.

VBet — Bet Builder and Live Streaming on a Mid-Sized Site

VBet’s app is published by SCGO Limited and offers Bet Builder, pre-match and in-play betting, live streaming and a live calendar with results and stats. Commission account 44662 is held by SCGO Limited. The brand is mid-sized in the GB market, and the in-play product is the centre of the offer rather than a bolt-on. For a punter whose weekend is built around in-play and live streaming, VBet is on the shortlist; for a punter whose main interest is ante-post horse racing or a deep football outright market, the depth is shallower than at bet365 or William Hill.

Midnite — A Newer Site With a Tight Welcome

Midnite’s site cites a Commission licence at account 101839, held by Midnite Limited, and shows in-play events, an acca builder, popular accas and price boosts with maximum-stake limits. The sports welcome offer is for new customers who bet £5 or more on pre-match odds of EVS (2.0) or greater within 14 days of sign-up, with £25 in free bets and 50 free spins credited, valid for 7 days. The qualification is light — £5 at evens is easy to place — but the 7-day expiry on the free bets and the spins is tight, and the maximum-stake limits on price boosts cap the upside on the marketed feature. For a punter who wants a quick qualifying bet and a short turnaround on the tokens, Midnite fits; for a punter who likes to use tokens over a fortnight, the clock is the cost.

Sky Bet — Free Bets With a Low-Value Qualifier

Sky Bet’s welcome offer for new customers is £30 in free bets, credited as three £10 free bet tokens after a deposit of at least £5 and a first single or each-way bet of at least 5p at odds of 1/1 or greater. The qualifier is the lightest on the page — 5p at 1/1 — and that is the spine of the offer for a punter who does not want to stake £10 or more to claim free bets. The brand is football-led and the in-play product is broad on the Premier League and the Championship. The cost is the small stake, which is also the marketing trick: 5p is a free-bet-trigger with no real exposure on Sky Bet’s side.

Tote — Pool Betting and a Different Pricing Model

Tote is the pool-betting brand on the GB market, and the pricing model is different from a fixed-odds bookmaker: punters bet into a pool and the dividend is declared after the race, with the takeout — the equivalent of a margin — built into the pool. For a punter who likes the pool-betting format and the chance of a bigger dividend on a fancied horse, Tote is the fit; for a punter who wants a fixed price before the off, the product is the wrong shape.

Paddy Power — Bet Builder+ and the Marketing-Heavy Welcome

Paddy Power’s app offers Bet Builder+, Popular Bet Builders, Super Sub (a bet that rolls over to the replacement if a selected player is substituted), expanded Win More markets and player and ACCA stats. Paddy Power and Betfair sit under the same group. The brand is the most marketing-heavy bookmaker on the GB market, and the welcome offer terms need reading with the same care as anywhere else. For a punter who likes Super Sub and a deep Bet Builder+, Paddy Power earns its place; for a punter who dislikes the marketing volume, the brand is the wrong fit. Paddy Power and Betfair are not independent operators; both run under PPB Counterparty Services Limited.

Responsible Gambling and Where to Get Help

GAMSTOP is the national multi-operator online self-exclusion scheme. The exclusion periods are six months, one year, five years, or five years with auto-renewal, and after the minimum period the exclusion stays in place for a further seven years unless the customer asks GAMSTOP to remove it. The scheme covers online sites and apps; betting shops run a separate self-exclusion service. Every GB-licensed online operator is a GAMSTOP member and is required to honour an active exclusion. An offshore brand is not.

The safer-gambling tools at most GB-licensed bookmakers — deposit limits, time-outs, product blocks, self-exclusion — sit on top of GAMSTOP and are not a replacement for it. A deposit limit, set by the punter, is the cheapest, simplest self-control a BTUE has on. A financial vulnerability check applies once a customer’s deposits minus withdrawals exceed £150 in a rolling 30 days; the check is a public-record check, not an income check. A time-out is a short break; self-exclusion is a longer one; GAMSTOP is the longest, and the only one that covers every GB-licensed site at once.

Where to get help:

• GamCare runs the National Gambling Helpline on 0808 8020 133, free and confidential, 24 hours a day.
• GambleAware funds treatment and research, with a website and live chat.
• NHS gambling clinics offer specialist treatment in several regions of England.
• Gamblers Anonymous runs peer-support meetings across the UK.

The honest framing is that betting is not an income and an AI prediction is not a wage. The bookmaker’s margin is built into every price, the staking plan costs money before any edge shows, and the help lines are listed because the day someone needs them is the day they wish they had been.

Frequently Asked Questions

Can AI help with betting?

AI is a tool that reads form, statistics, line-ups, xG and a hundred other variables, and it can estimate a probability for an outcome faster than a human. It does not change the bookmaker’s margin, the staking plan’s discipline or the fact that no outcome is certain. AI helps a punter who already has a staking plan and a model for value; it does not replace either.

Can AI help with sports betting?

Yes, in the same way it helps with betting more broadly — by estimating probability from data, which is the half of value betting the punter cannot do at the keyboard. The bookmaker still sets the price, the price still carries a margin, and the punter still has to find a price that is higher than the AI’s probability implies.

Can AI predict sports betting?

A model can predict probabilities; it cannot predict outcomes in the sense of naming winners. A 60 per cent model is right 60 per cent of the time and wrong 40 per cent, and the bookmaker’s margin eats into both sides. Anyone selling a model that names winners is selling certainty that no model can deliver.

What is a value bet and why is no tip ever certain?

A value bet is one whose odds imply a lower probability than the true chance of the event. A bookmaker offering 2.20 on a 50/50 event is offering value (the price implies 45.5 per cent); a bookmaker offering 1.80 is not (the price implies 55.6 per cent). No tip is ever certain because the bookmaker’s margin is built into every price, the model’s probability is an estimate not a fact, and an event with three or more outcomes can lose in any of them.

What is the bookmaker’s margin and how can it be seen in the odds?

The margin is the overround built into the price — the gap between 100 per cent and the sum of the implied probabilities of every outcome. A market at 1.90 / 1.90 sums to roughly 105.3 per cent, which is a margin in the 5–6 per cent band. The margin is what the bookmaker keeps of every pound staked, regardless of which outcome wins.

How is the return on a bet worked out from the stake and the odds?

In decimal odds, return equals stake multiplied by odds. A £10 stake at odds of 3.75 returns £37.50, of which £27.50 is profit. In fractional odds, a £10 stake at 5/1 returns £60 (£50 profit plus the £10 stake). For an accumulator, the odds of the legs multiply into one price, and the same formula applies. Free-bet stakes are usually not returned with a win.

Content created by the «signupbetsguide.com» team