Sports betting picks and tipsters: the full research

The research was done by AI agents that open web pages. Some sites refuse them; where that happened we say so. The small numbers point to the source list at the end. Where a figure is our own sum or guess, it is marked “our estimate” and the basis is given.

The claim

Our collection for this scheme holds 11 tweets¹. Only two of them come from accounts that our naming rule allows us to quote, because each tweet promotes a paid picks group whose sales page we opened²⁴. The first tweet states a monthly price and carries a link and a Whop hashtag, and a Whop store of the same name exists¹². The second is headed with the name of a picks group, and a Telegram channel and a Whop store carry that name¹⁴⁶; we did not see a link from the X account itself to either. They are quoted below. View counts are as collected. Dates are worked out from each tweet’s ID number. Tweets are examples of the claim. The page judges the scheme, not the people who tweeted.

“Everyone in our community absolutely fed last week! A $10 unit bettor would be up over $780 in just 7 days! All this for only $30 per month. Come win with us.”

@RoyaltyWagers, 19 June 2026, 362 views¹. The link and hashtags at the end of the tweet are left out.

“TommyBANKERS ANNOUNCEMENT 📢 Over the last 30 days we’ve built an incredible community in here with over 4,000 members - along with that, we finished the World Cup with 12.87 UNITS PROFIT and 6.9% ROI with every single bet tracked publicly in this channel.”

@tommypredicts, 24 July 2026, 4,533 views¹. This quote is cut off at the end: our file holds only the first part of the tweet.

The other nine tweets are from accounts with fewer than 10,000 followers and no sales page that we saw, so we describe them without names. Eight ask the reader to follow, like or comment in return for picks sent by private message; seven of those use one of two near-identical wordings, posted by four accounts with 9 to 197 followers¹. One announces a new group for mixed martial arts picks and says a “funded bettor program” is being worked on¹. None of these nine states a profit figure in the tweet text¹. We did not view the images, videos or threads attached to any of the 11 tweets, and we did not open the private messages.

Because the collection is small, the rest of this page rests on how the method is promoted and measured generally, not on these 11 tweets alone.

What the scheme is

A tipster is someone who tells other people what to bet on. The tips are called picks. A parlay is one bet that joins several picks together and pays only if every one wins. The reader places the bets with their own money at a sportsbook (a betting company). The tipster does not bet for them.

Picks reach readers in three ways in our collection¹:

Free picks sent by private message, in return for a follow, a like or a comment. The tweet shows no price. What is offered after that is not public.
A paid group. The two quoted tweets lead to monthly subscriptions hosted on Whop, a marketplace for paid communities²⁴.
A “funded bettor program”. The buyer pays a fee to take a test on a pretend bankroll. If they hit a profit target without breaking the rules, the firm promises a share of later profits. We explain these below.

Three words in the tweets need a gloss. A unit is the bettor’s standard bet size, so “a $10 unit bettor” bets about $10 at a time. ROI (return on investment) and yield mean the same thing here: profit divided by all the money staked. A 5% yield means $5 of profit for every $100 bet. The closing line is the last price a sportsbook offers before the game starts; it is usually the most accurate price, and bettors who regularly get better odds than the closing line are the ones sportsbooks treat as skilled²⁵²⁷.

The scheme also covers “fixed match” offers: someone claims to know the result of a game in advance and sells it.

Some bettors do beat sportsbooks, and that is measured below. The question is how large the edge is, how much must be staked to turn it into an income, and whether a follower gets the same result as the tipster.

The arithmetic

Step 1: what the tweets give and what they leave out

Only two of the 11 tweets give a number¹. The first gives a profit ($780), a unit size ($10), a period (7 days) and a price ($30 a month), but not the number of bets, the odds, the total staked or any losing week¹. The second gives units won (12.87) and the yield (6.9%), which is enough to work out how much was staked¹.

Step 2: what “$780 in 7 days” means

$780 at $10 a unit is 78 units of profit in one week (our sum: 780 / 10).

To see what that takes, we need a yield. Two are available:

6.9%, the figure in the second tweet¹.
3.97%, the most favourable large-sample average we found: the yield across 1,534,041 tips from 3,484 tipsters on Blogabet, a site that records tips before the event. The study’s summary says: “Tipsters won 3.29% more tips than the implied win probability set by bookmakers and produced averaged yields of 3.97%.”¹¹

Our estimate, from those figures:

At 6.9%: 78 / 0.069 = about 1,130 units staked in the week. That is $11,300 at $10 a unit, or about 160 one-unit bets a day.
At 3.97%: 78 / 0.0397 = about 1,965 units staked. That is $19,650, or about 280 one-unit bets a day.

So a 78-unit week means the bets were much larger than one unit each, or the week was unusually lucky, or the return was far above anything measured over a long record.

The seller’s own long-run claim gives a second test. Its Whop page says “We’re up over 1000 units of profit in 2026”³. That statement is unverified. Taken at its word, and read on 10 October, about 40 weeks into the year, it means about 25 units in an average week (our estimate). The quoted week is about three times the seller’s own average. It is a good week, not a rate.

Step 3: what “12.87 units at 6.9% ROI” means

12.87 / 0.069 = about 186.5 units staked over the tournament (our estimate). A follower betting $10 a unit made about $129 (12.87 x 10). To make $1,000 from the same run the unit would have to be about $78, and for $10,000 about $777 (our estimates).

This claim is modest and its two numbers agree with each other. The difficulty is the length of the record. Joseph Buchdahl, an analyst who checked tipsters’ records for 14 years, worked out that a bettor with a 10% yield at even-money odds (a winning bet pays the same amount as the stake) “will need about 540 even-money betting propositions before I can conclude that something other than chance might be accounting for my performance”¹². About 186 units is well short of that.

Our own check, as an estimate: if the bets were one unit each at about even odds (the tweet says neither), luck alone moves the result by about 13.7 units either way (the square root of 186.5). A profit of 12.87 units is inside that range. A bettor with no skill would do this well roughly one time in six if the odds were fair. Against a typical bookmaker margin of about 4% the same bettor expects to lose about 7 units, and would finish this far ahead roughly one time in 15 (both are our estimates). Either way the run is too short to tell skill from luck. The seller does publish comparisons with the closing price in its channel, which is a better sign than a win-loss record alone⁶.

Step 4: the price a bettor has to beat

Sportsbooks build a margin into their odds. At the common US price written “-110”, a bettor stakes $110 to win $100, and must win 52.38% of bets to break even (our sum: 110 / 210).

The share of all bets that a sportsbook keeps is called the hold. In Illinois in February 2026, trade press reporting the state regulator’s figures put the overall hold at 9.4%, and the hold on parlays at 17.7%: “the house claimed $60.4 million in winnings with a pedestrian 17.7% hold against $340.4 million handle”²⁸. (Handle is the total amount bet. “Pedestrian” here means ordinary for parlays.) Bettors as a group got back about 82 cents of each dollar bet on parlays that month.

Hold is counted in more than one way, and it differs between states and months. The 9.4% above is worked out on adjusted revenue, a figure after deductions²⁸. On gross revenue, before those deductions, Illinois sportsbooks kept 16.33% in May 2026, close to the 15.26% of May 2024 and the 15.05% of May 2025²⁹. The same report sets that against June 2026, during the World Cup, when favourites kept winning: “New York, taking a record handle during the World Cup, held 5.19%. Kansas held 3.90%.”²⁹ Bettors as a group did unusually well in the World Cup month. That is one reason a good World Cup record says little about the next month.

Step 5: the volume needed for an income

Income from betting is yield times money staked.

The table uses five yields. 1.3% is what a peer-reviewed study got by backing the majority view of free tipsters on one site, over 68,339 events⁶³. 2% is roughly the edge against the closing price in 61,979 bets logged by users of a paid value betting tool (1.6% to 2.8%); that is an expected edge, not banked profit¹⁶. 3.97% is the Blogabet average, the most favourable large sample¹¹. 6.9% is the tweet’s figure¹. 10% is the highest long-record yield we saw cited by a tipster-monitoring service, over nearly 1,000 bets; it is the service’s own statement and we did not open the report behind it⁶⁰. All the sums are our estimates.

Yield	Staked each month for $1,000	Staked each month for $10,000	$10 bets a day for $1,000
1.3%	about $76,000	about $760,000	about 253
2%	$50,000	$500,000	about 167
3.97%	about $25,200	about $252,000	about 84
6.9%	about $14,500	about $145,000	about 48
10%	$10,000	$100,000	about 33

Nobody places 84 separate $10 bets a day from a picks group. In practice the money comes from bigger units, and bigger units need a bigger bankroll (the pot of money set aside for betting). One funded bettor firm caps each bet at 2% of the account³³. If a bettor kept each unit to 1% or 2% of their bankroll, which is our assumption and not a sourced rule, a $78 unit means a bankroll of about $3,900 to $7,800, and a $777 unit means about $39,000 to $78,000 (our estimates).

Step 6: the costs

The subscription. The first tweet says $30 a month¹. The seller’s Whop page today shows “Royalty Wagers Premium” at $39.99 a month as the default price, with three other price options that we did not open³. The tweet is from June, so the $30 may be an earlier price or one of those options. At a 3.97% yield a bettor must stake about $756 a month just to cover $30, and about $1,007 to cover $39.99 (our estimates). The fee is paid whether the picks win or not.

Edge that is not really there. An edge measured when a bet is found is often bigger than the edge that lasts. In the 61,979 bets mentioned above, the edge the tool showed at the moment of its alert averaged 3.0% to 9.2%, depending on how much the reference bookmaker, Pinnacle, would accept on that market. The edge measured against the closing price was 1.6% to 2.8%¹⁶. The author writes: “Below a $500 limit, only 28% to 30% of the logged EV survived to the close. Above $2,000, around three quarters survived.”¹⁶ (EV, expected value, is the edge the tool calculated. A low limit marks a thin or early market.) These are estimates of expected edge, not profit, and the data came from the tool’s vendor¹⁶. The source does not measure followers who bet after a tip is posted. Our inference is that a follower faces the same problem and one more: by the time they bet, the price may have moved. That is likely to matter most in early, small and niche markets; the first quoted seller’s page mentions “niche betting markets such as Table Tennis”³.

Tax, for US readers. The Internal Revenue Service says: “Gambling winnings are fully taxable and you must report the income on your tax return.”³⁰ Losses can be set against winnings only by people who itemize their deductions (list them on the tax return instead of taking the standard amount), and never beyond the amount won³⁰. A 2025 paper by the accounting firm KPMG says a law passed in July 2025 will “limit deductions for wagering losses to 90 percent of such losses, effective for tax years beginning after December 31, 2025”³¹. The IRS page we read, last reviewed on 24 September 2026, did not mention that limit³⁰. We did not open the law itself or check whether it has since been changed. In the UK the position is different: the tax authority’s manual treats betting winnings as outside income tax, in the words “There is no tax on a habit.”³²

Stake limits. Covered under “What the rules allow now”. A bettor who wins steadily can have their bet size cut. How often that happens to steady winners is not published.

What people who tried it report

We found no income report, checked against payment records, from an ordinary follower of paid or private-message picks. Not a good one and not a bad one. What exists is data on bettors in general, and a few studies of tipsters.

Bettors in general. Researchers at three US universities followed 717,724 online gamblers through payment records over about five years. The university’s summary quotes a co-author: “Of the more than 700,000 gamblers that we studied, 96% percent appeared to lose money to online gambling.”²¹ The paper itself says “fewer than 5% of potential gamblers withdrew more than they spent, collectively earning more than $100 million”²². Limits: the data cannot separate sports betting from online casino and fantasy sports, it does not see money left in a betting account, and it is a working paper that has not been through peer review²¹²².

The Bank of America Institute looked at its own customers’ payments in 2026 and found: “customers typically recover less than 75 cents for every dollar transferred to online betting platforms.”²³ The report adds that it sees money moving in and out, not final profit, because winnings left on the platform are not visible²³.

The British regulator asked betting companies about 14,923,840 active accounts in 2024. It found “25.42 per cent of active customers are in profit vs. 46.78 per cent of restricted customers”²⁴. This is the most favourable official figure we found. It counts accounts, not people, an account with a single bet counts, and it gives no profit sizes²⁴. It does not show that a quarter of bettors earn an income.

Tipsters on social media. One study tracked every pre-match bet slip posted by three prominent Nigerian betting influencers on X and Telegram: 5,467 slips and about $4.8 million in bets, checked against the sportsbook’s records. “The influencers themselves collectively lost 25.24% on their promoted bets, while a follower who staked the same amount on every tip would lose 38.27% on their investment.”¹⁹ This is the only study we found that measures what a follower would get. Limits: it is a preprint (not peer reviewed), it covers three accounts in one country, and the slips were mostly long-odds bets joined together, with median odds of 265, meaning the middle slip would pay about 265 times the stake¹⁹. It does not describe a service that posts single bets at short odds.

Tipsters who agreed to be checked. Buchdahl verified tipping services from 2001 to 2015 and concluded that “few tipsters are capable of making a profit beyond that which could be expected to happen by chance alone”¹³. In his sample of 48 tennis tipsters and 7,170 bets, the combined yield was -0.94% at the odds the tipsters quoted. 59.1% of their quoted prices beat the closing price, so the group had some information. But, he writes, “Not one of them managed a positive profit expectation (based on this analysis) that was even remotely statistically significant.”¹³ The sample is from before 2016 and covers one sport. Buchdahl sells books on betting, and his site states: “Football-Data’s revenues come from losses accrued by customers of banner-advertised bookmakers.”¹³

Prediction services, tracked in 2026. Tips.GG, a company that ranks tipsters, says it scored over 500,000 football predictions from 165 services at a flat $10 stake from 1 January to 31 August 2026. “Thirty-eight of them (23.0%) ended the period with a positive return on investment.”²⁰ So 127 of the 165 did not. Five services were above 10%, but the leader’s +17.40% rests on 95 predictions, and a website needed only 50 to be counted²⁰. Of the 73 services with 1,000 or more predictions, 4 were in profit. The middle return among the profitable services outside the top five was 2.47%, and seven finished less than one point above zero²⁰. The release says the bookmaker’s margin, 8.19% on average, was left in the odds²⁰. We read this on a site that republished the company’s release, through a tool that summarises pages; Tips.GG’s own page refused our agents. Treat the figures as the company’s.

The positive dataset. The Blogabet average of +3.97% across 3,484 tipsters is the best large-sample evidence that some tipsters beat the posted odds¹¹. Limits: we read only the summary; tipsters choose to post there; the yield is at the odds the tipster quoted, before subscription fees, before prices move and before stake limits; and it is an average, so it does not say what share of tipsters were in profit¹¹.

A second positive sample, with a catch. A peer-reviewed study of free tips on Oddsportal, an odds-comparison site, found: “A strategy of betting when a majority predict an outcome produces average returns of 1.317% for 68,339 events.”⁶³ The returns were worked out at average odds, not the best odds on offer⁶³. The same paper found that “limiting the crowd to tipsters with more experience (more past tips) or more skill (higher historical returns on their tips) does not improve betting returns”⁶³. So the crowd as a whole carried a small edge, and picking tipsters by their past record did not add to it. The paper was accepted in 2018; we did not check whether the result holds today.

Funded bettor programs. We found no pass rate and no payout statistics for any firm. The only reports are customer reviews. One firm’s Trustpilot page has 22 reviews, mostly five stars, with payouts that reviewers say ran from $515 to $7,290³⁹. Another firm’s page has 34 reviews, most of them five stars and a small number one-star⁴⁰. All of these are unverified statements by reviewers. We do not repeat what the one-star reviews say, because we did not ask the company for its reply. Reviews of either kind say nothing about what share of buyers are ever paid.

“Fixed match” offers. Ankura, a consultancy, and U.S. Integrity, a firm that monitors betting for sports bodies, paid to join sellers of fixed matches. Of 27 sites they found on hidden parts of the internet, 5 were active. Of the six closed Telegram channels they joined or tried to join, their report says all “were either unresponsive or took subscription fees and then became unresponsive”⁴¹. They bought tips from two sites. One tip, for $90, lost. The other site’s first tip, for $65, won at odds of 7/4 (a $4 bet wins $7); the team then bought several more from that site and none came in⁴¹. The report concludes that the services “did not show reliable evidence of being linked to match-fixing”⁴¹. The authors add that this does not rule out more private channels⁴¹. Neither firm is a regulator or a court. Buchdahl tested one fixed-match seller’s published record (88.5% winners at average odds of 2.28) and concluded it could not be real, reasoning that true inside information would move the odds and would not be on sale for 50 euros¹². We found no evidence that any fixed-match seller had real inside knowledge. One bought tip winning and the rest losing is what chance alone could produce (our reading).

What the rules allow now

Selling picks is allowed. Whop bans gambling operations on its marketplace but states that the ban “Excludes information, analysis, picks, or other advisory services related to gambling or sports betting.”⁸ Discord’s gambling policy says “Users may not use Discord to coordinate or participate in illegal gambling” and does not mention picks or tips⁵⁶. Following picks is legal wherever the betting itself is legal for the follower.

Almost nobody checks the record. Maryland passed a law in 2023 “Requiring the Maryland Lottery and Gaming Control Commission to identify and license certain independent evaluators to evaluate and rate certain sports wagering content provided by sports wagering experts, sports wagering influencers, and content partners”⁶². It took effect on 1 July 2023⁶². We did not check how it has been used or which sellers it reaches. A 2023 West Virginia bill that “would have created a system to evaluate sports betting handicappers and touts” failed⁵⁵. We found no other US state that licenses or audits picks sellers. Whop’s rule is that “Any direct or implied earnings claims must be accurate, substantiated, and not misleading”⁹, but that is applied after the fact, not before a seller lists. Whop’s terms also let it close a seller’s account “at any time, with or without notice”¹⁰.

The UK standard shows what a checked record looks like. The UK advertising regulator’s guidance says a tipster who claims past profits should “register the recommended bet with a demonstrably independent body before the race”, and “should not state or imply that success is ‘guaranteed’”⁵⁰. It also says results must not be shown as typical and a favourable period must not be picked out⁵⁰. A 2014 ruling shows that even a registered record may not be enough. A tipster service had advertised “Over £2800* profit in just a couple of weeks”. The regulator noted that the firm had sent its tips “to an approved proofer prior to making their tips available to customers” (a proofer is an independent body that records tips in advance) and that the advertised odds were on offer at that moment⁵¹. It still found the claim misleading, because the firm had not shown that the odds stayed available long enough for members to get them, or that two winning weeks were typical: “We therefore concluded the profit claims in the ads had not been substantiated and were misleading.”⁵¹ That ruling concerns only the company named in it, and the guidance applies to UK advertising. It is not a finding about any account quoted here. The guidance gives readers a test: a record kept by the tipster in the tipster’s own channel is not independent checking⁵⁰.

Regulators have acted on profit claims and on pooled money. Hungary’s competition authority fined a betting-tips service 9 million forints over claims including “You can achieve a 300-400% profit in 30 days with this strategy”, and fined agencies and influencers more than 10 million forints in total for paid promotion that was not disclosed⁵². In the United States, the Securities and Exchange Commission charged the operators of pooled betting funds who, it alleges, “promised investors 250% to 600% returns from pooled investments in sports betting”⁵³. Those are allegations in a 2019 release; we did not check how the case ended. A man who took money to bet for others through a sports betting fund was sentenced to 48 months in January 2025, after pleading guilty to three counts of wire fraud; the Department of Justice’s release uses the word “defrauded” in its title and says he “sent false performance updates claiming substantial investment profits”⁵⁴. Each of these cases concerns only the companies and people named in it. The last two are about handing money to someone else to bet, not about a picks subscription.

X and affiliate links. An affiliate link pays the person who shares it when a reader signs up. Trade press reports that in February 2026 X added “lotteries, social casinos, sports betting, and other gambling-related content” to the categories barred from paid partnerships⁴⁹. X’s own policy pages refused our agents both directly and through a reader, so this rests on trade press, which does not say whether the rule covers a tipster promoting their own paid group.

Winners can be limited. A sportsbook may cut the amount a customer is allowed to bet. The British regulator reports that “operators reported 643,779 accounts restricted in some form – a rate of 4.31 per cent”²⁴. It also found “25.42 per cent of active customers are in profit vs. 46.78 per cent of restricted customers”, with profit measured over the account’s lifetime on settled bets²⁴. From those three figures, about 8% of accounts in profit were restricted, against about 3% of the rest, roughly two and a half times as often (our sum). So about 92% of accounts in profit were not restricted. Most accounts were not in profit at all: 72.54% of active accounts were in loss²⁴. In Massachusetts, the gaming commission’s minutes record that “players who demonstrated a tendency to win had their limit decreased and players who demonstrated a tendency to lose had their limit increased”²⁵. Trade press covering that meeting reports the commission’s finding that “limiting is something that happens to a fraction of a percent of bettors”: 0.64% of bettors were limited and, of those, “57.6% were limited to between 1-24% of the default maximum bet”²⁶.

These figures cover every account that happens to be ahead, including people who were lucky once. No source we found gives the rate for bettors who win steadily on price. The one documented case on this page is the 2017 research team in “The upside”, whose accounts were cut within months¹⁸. So a follower of a tipster who truly wins may not be able to keep betting at the size the income needs, but we cannot say how likely that is.

Funded bettor programs are sold as contests, not betting. One firm states: “Fan Funded is not a casino, sports book or gambling operator and does not accept or place wagers of any type.”³⁵ Another says: “The BankrollU platform is for educational and recreational purposes only.”³⁶ Neither page names a gambling licence³⁵³⁶. A guide on a comparison site says of a “funded” account: “The displayed bankroll is not a deposit in your name.”³⁷ So any payout is a promise from a private company under its own rules. We found no regulator decision on these programs.

Who makes money from it

The seller of the picks. A subscription is paid whether the picks win or lose. The Royalty Wagers product page says: “Participation is voluntary, and Royalty Wagers is not liable for any losses, damages, or consequences resulting from gambling activities.”³ It shows 28 members on the product; the store page shows 142 joined and 27 reviews, all five stars²³. The reviews include statements that the reviewer made a profit, and one says bookmakers have begun to limit the reviewer; none is verified³. The page also states “We’re up over 1000 units of profit in 2026”; we saw no bet-by-bet record on the public page³. Our estimate of the seller’s gross income: 28 members at $39.99, if all are on the default monthly plan, is about $1,120 a month; if all 142 paid the $30 in the tweet it would be $4,260 a month. “Joined” may include free or lapsed members, so the second figure is a ceiling.

The second tweet is headed “TommyBANKERS ANNOUNCEMENT” and speaks of “over 4,000 members” in “this channel”¹. A public Telegram channel named Tommy BANKERS has about 4,000 subscribers (4,016 when we read it) and describes itself as “Free value picks, no paid wall.”⁶ Posts in the channel point readers to a Whop store of the same name⁶. We did not see a link from the X account itself, so the match rests on the name and the member count. The Whop store lists “Platinum Tier £19.99 / month Diamond Tier £39.99 / month” and yearly plans at £199.99 and £384.99; it shows 144 joined⁴. The seller’s page says: “You can cancel your subscription at any time but we unfortunately do not offer refunds.”⁵ In the free channel the seller posts losing bets and comparisons with the closing price, which is more than most of what we saw⁶. By its own World Cup figures, a follower betting £10 a unit made about £129, roughly six months of the cheaper tier or three of the dearer one (our estimate).

How many subscribers a seller needs. Whop lists card processing at “2.7% + $0.30” per sale⁷. A $30 subscription leaves about $28.89 (our sum). So $1,000 a month takes about 35 paying subscribers and $10,000 takes about 347 (our estimates, before payout fees).

The platforms. Whop is paid on every sale⁷. DubClub, a US platform built for selling picks, tells would-be sellers: “2,000,000+ DUBCLUB MEMBERS 11,057,000+ PURCHASES ON DUBCLUB 15+ MILLIONAIRES MADE”⁴². Its refund policy for buyers says “ALL SALES ARE FINAL AND NON-REFUNDABLE”, and adds that DubClub or the seller “may, at our or their sole discretion, offer a refund, discount or credit at any time”⁴². A 2024 report put its payouts at “more than $50 million in payments to cappers” (a capper is a picks seller)⁴⁴. A later press page says “By late 2025, 570 cappers had earned $10,000 or more on the platform”, with 18 past $1 million⁴⁵. These are the company’s own unaudited figures. The total number of sellers is not given, so the share who earn well is unknown.

One seller’s page on DubClub shows the two sides of the sum. We opened it as an example and did not sample sellers at random, so it is not a typical record. The page lists a price “Starting at $25.00 / week or as low as $1.64/Day”, “Subscribers 400”, and a platform-tracked record of “+5.04u 128-127 MLB moneylines · 2026 season”⁴³. (Here u means units, 128-127 is wins and losses, MLB is US professional baseball, and a moneyline is a bet on who wins.) Our estimate: if all 400 are paying, the seller takes in between about $4,600 and $10,000 a week, from $1.64 a day to $25 a week; the count may include other plans or people who do not pay. A $10-unit follower of the tracked record would be up about $50 for the season. That tracked figure covers baseball only. The plan is described as an “All access package for MLB, NFL, NCAAF,NCAAB and NBA plays”, and the seller’s own description says “3 years of 75+ unit success with our model”, which the page does not track and we did not check⁴³.

People paid for referrals. The TommyBANKERS product on Whop has an affiliate setting of 20% in the page’s data: a person who refers a subscriber is paid a fifth of the price⁵.

Sportsbooks, and tipsters paid by sportsbooks. Some betting companies pay affiliates a share of what referred customers lose. The partner site of one, 1xBet, says: “For RevShare, it ranges from 15-45%”⁴⁶. (RevShare means revenue share.) A 2017 report on UK tipsters put it this way: “Usually this is 30% of the “revenue” from your account. In layman’s terms, that means 30% of your gambling losses go to them.”⁴⁷ The same report quotes a tipster’s answer that nobody is forced to follow his account or bet on his tips⁴⁷. A few weeks before that report, after earlier press coverage, one large UK betting company closed its affiliate program⁴⁸. The Nigerian study describes the deals there in these words: “an influencer’s earnings depend entirely on how much their followers bet, not on whether those bets win or lose”¹⁹. These sources are old, or from other countries. None of the 11 tweets shows a sportsbook link in its text, and we have no evidence that any account in our collection has such a deal¹.

Funded bettor firms. MyBetFund charges $10, $45, $85 or $170 for a test account of $1,000, $5,000, $10,000 or $25,000³³. The buyer must make 10% and then 5%, with “Max stake per bet: 2%”, no more than 10% lost in total, and then gets 70% of profits³³. Another firm, WagerFunding, explains the business plainly: “The firm profits from challenge fees and their share of winning bettors’ profits.”³⁴ It also says: “No real money bets occur on the platform.”³⁴ A review site that lists 14 such firms, and says it earns commission from links to them, states that at every firm listed “you bet simulated capital priced against real odds”³⁸. MyBetFund’s page says something different. Its two test steps are “in a simulated environment”, and then: “Once funded, you bet with real capital and all profits are yours to withdraw.”³³ (The same page sets the bettor’s share of profits at 70%³³.) We could not settle which is right for that firm; its terms did not load.

Our estimate of what the MyBetFund test requires: gaining 15% with bets capped at 2% of the account takes at least 109 maximum-size bets at a 6.9% yield and about 189 at 3.97%, without ever being down 10% on the way. On the largest account, $1,000 a month to the bettor needs $1,429 of profit, which is 5.7% of the account each month. A comparison site that links to these firms models the chance of getting funded, for a bettor who wins 55% of bets at -110, at 14% to 72% depending on the firm, and says this is a model and not an observed pass rate³⁷. The same site says of itself: “BETfunded used to be a prop firm; its challenges are no longer on sale.”³⁷ (A prop firm here is a firm that sells funded-account tests.)

The upside

Sports betting can be beaten. The evidence describes people doing something quite different from copying picks.

A documented method that won. In 2017 three researchers published a strategy, with their code and data, that predicts nothing. It bets when one bookmaker’s odds are well above the market average. In a simulation of 56,435 bets over ten years it returned 3.5%. Then they bet for real: “During that period we obtained an accuracy of 47.% and a profit of $957.50 across 265 bets, equivalent to a 8.5% return”¹⁸. That is about $190 a month over five months (our sum). The paper says each bet was $50, so $957.50 on 265 bets works out to about 7.2% of stakes by our sum, not 8.5%; the paper’s combined figure of 6.2% for 672 practice and real bets fits 7.2%. The figures are the authors’ own, with public data. It ended this way: “A few weeks after we started trading with actual money some bookmakers began to severely limit our accounts, forcing us to stop our betting strategy.”¹⁸ Elsewhere the same paper says “a few months”¹⁸. We did not check whether the method still works in 2026.

A winning minority, in official data. In the US payment study, the fewer than 5% who took out more than they put in collectively gained more than $100 million, and 10 of the 43 people who deposited over $500,000 were net winners²².

A tipster with a long record. Smart Betting Club, a UK service that has monitored tipsters since 2006, cites one with “a 10.1% ROI and 181 points profit over nearly 1000 bets in major NFL markets”⁶⁰. (Points are units.) This is one line of promotional text from a service that sells memberships. We read it through a summary tool, and the full reports are paid⁶⁰. It is the only long-record example we found.

Short-run winners. The Tips.GG count found 38 of 165 services in profit over eight months, five of them above 10%. The best, +17.40%, had 95 predictions, and only 4 of the 73 services with 1,000 or more predictions were in profit²⁰. By the 540-bet test in Step 3, the high figures are too short to count as records.

Places that tolerate winners. One large bookmaker says of itself: “Pinnacle’s strategy is not to limit sharp players, but rather to learn from them.”⁵⁷ (Sharp means skilled.) That is the company’s own statement, and we did not check where it may legally be used. Betfair, a betting exchange where customers bet against each other, charges its biggest winners extra instead of banning them: a report on its 2025 fee says “Exchange users with a 52-week gross profit of less than £25,000 will not pay the new fee”, with a rate of 20% for customers whose 52-week gross profit is between £25,000 and £100,000 and 40% for those above £100,000⁵⁸. That is the fee as announced in December 2024; we did not check the current one. A fee like that exists because some customers win tens of thousands of pounds a year.

The very top. A claim filed in the High Court in England alleges that one betting syndicate (a staffed group that bets as a business) wins about £600 million a year: “The claimant understands its yearly winnings to be in the region of £600 million.”⁵⁹ This is the claimant’s figure in a claim reported on 4 December 2025. It is alleged and untested. The report we read says that no defence had yet been filed and that the defendant had been approached for comment; it prints no reply⁵⁹. We did not check what has happened in the case since.

The top tenth and top hundredth. We found no dataset that gives what the best 10% or 1% of picks followers earn. The nearest markers are the fewer than 5% of net winners in the US payment data²² and the 0.64% of Massachusetts bettors limited²⁶.

A realistic good result. Buchdahl describes the skilled version as “relatively modest yields of perhaps a few per cent”, with “fairly long periods of treading water, or worse still, losing, lasting hundreds and perhaps thousands of bets”¹⁷. Our estimates, built on a yield of 3% to 6%:

Betting on price yourself, at $50 a bet and 50 to 60 bets a month: about $75 to $180 a month. This rests on one 2017 trial that we did not re-test, and in that trial the sportsbooks cut the accounts within months¹⁸.
$1,000 a month: about $16,700 to $33,300 staked every month, across many accounts, with a bankroll that survives long losing runs.
Following a truly skilled paid tipster at $10 units, 120 bets a month, at a 5% yield: about $60 a month before a $30 subscription, so about $30 after it. That assumes the same odds are still there when the follower bets.

So for following picks at $10 units, a realistic good result is tens of dollars a month after the fee. For a careful, numerate person betting on price themselves, it is roughly one to two hundred dollars a month, and limits can cut it short. A full income is documented only for people who run betting as a business with capital.

What it takes to compete

They bet on price, not on a feeling about the game. Every documented winner above bets only when the odds on offer are better than the true chance, and measures that against the closing line¹⁷¹⁸. Even that is not enough on its own. An economist’s analysis of NBA bets found that “it is only the top 10 percent of moneyline bets that managed to turn CLV into meaningful profits”⁶¹. (CLV, closing line value, is how much better your odds were than the closing price.) The edge has to be bigger than the sportsbook’s margin, which was about 4.5% in that sample⁶¹.

They have capital. The table in Step 5 is the cost of entry.

They keep long records and judge them slowly. About 540 even-money bets are needed before a 10% yield can be told from luck, and more at longer odds¹².

They plan around limits. The 2017 researchers were stopped within months¹⁸. In British data about 8% of accounts in profit were restricted, against about 3% of the rest (our sum from the regulator’s figures; profit measured over the account’s lifetime)²⁴. The rate for steady winners is not published. A bettor who is limited needs other accounts, an exchange, or a bookmaker that accepts winners.

How to tell early which side of the line you are on. These are checks a reader can make before paying anything:

Is the tipster’s record registered before each event with a body that gains nothing from the tipster’s sales? That is the UK regulator’s standard⁵⁰. A record in the seller’s own channel does not meet it.
How long is the record? A week or one tournament carries almost no information. If two bettors differ in skill by 2% on even-money bets, the weaker one still has a 38.19% chance of being ahead after 100 bets¹²¹⁴.
What yield does it show? The long records we found cited sit between a few percent and about 10%¹¹¹⁷⁶⁰. Buchdahl flags claimed win rates of 80% to 90% at around even odds as impossible¹⁵.
Does the record include the odds and the stake for every bet, and the losing days?
Do not rely on a past record alone. In one large sample, choosing tipsters by their past returns did not improve results⁶³.
Work out your own number: your monthly stakes times a yield of 3% to 6%, minus the subscription. If the result is small or negative at the stakes you can afford, the picks being good does not change it.
If you do bet, write down the odds you actually got and the closing odds for each bet. If you are not beating the close after a few hundred bets, the evidence says to expect the average bettor’s result¹³⁶¹.

Signs that point the other way: a profit quoted for one week, a promise of certainty, a “fixed” result for sale, or a request to send money for someone else to bet.

What we did not verify

The tweets were not re-opened on X. Text, views and follower counts come from our file, and the second quoted tweet is cut off there. We did not view any tweet’s images, videos or threads.
A link from either quoted X account to the Whop store or Telegram channel of the same name. The match rests on the names, the price in the first tweet and the member count in the second.
The records behind the two quoted tweets: the Telegram bet log, the Discord group and the “over 1000 units” statement. The unit figures are the sellers’ own.
The price of Royalty Wagers when the tweet was posted, its three other price options, whether Whop member counts include free or lapsed members, and whether Whop reviews come from paying members.
What the eight private-message accounts send after a reader follows, likes or comments; the mixed martial arts group and its planned funded bettor program.
X’s own rules on gambling promotion, paid partnerships and repeated posts. Its policy pages refused our agents directly and through a reader.
Whop’s full fees, and how long it may hold a suspended seller’s money (our tool summarised that passage, so we left the figures out).
The Blogabet paper beyond its summary, and the Tips.GG study at its source.
Whether the 2018 finding on free Oddsportal tips still holds, and the 2017 value betting method in 2026. The 8.5% return printed in that 2017 paper does not match its own stake and profit figures.
What followers lose to price moves after a tip is posted. The value betting data measures edge at the alert against the close, not a follower’s delay.
How often steady, price-driven winners are limited. The official figures cover all accounts in profit.
How Maryland’s 2023 law on evaluating betting content has been applied, and which sellers it covers.
How the DubClub seller in our example compares with other sellers there, and that seller’s three-year claim.
The Illinois and Massachusetts numbers against the regulators’ own files; we read trade press for them, and did not check what the Illinois “adjusted” figure deducts. One of our agents found no Massachusetts figure in a different trade report of the same meeting.
The US 90% limit on deducting betting losses, in the law itself or on an IRS page, and whether it has since been changed.
Current affiliate pay at US-licensed sportsbooks. One company’s rate was read at source; the 30% figure is from 2017 UK reporting.
Funded bettor programs: pass rates, payout records, full terms, whether any “funded” account places real bets, and any regulator ruling. Trustpilot reviews, positive and negative, are unverified, and we did not ask any firm for its reply to them.
Any court or regulator document on “fixed match” sellers. A reported US guilty plea over the sale of supposed inside information was seen only in trade press, because the Department of Justice page refused our agents, so it is left out.
How the 2019 SEC case ended, and the High Court claim beyond one news report from December 2025, including its current state. The current Betfair fee.
Any US state or federal action against a picks seller over win-record claims. Our searches found none, which is not proof that none exists.
Which sportsbooks or exchanges a reader may lawfully use; that depends on country and US state.
What followers of a paid, independently checked tipster actually earn after fees, late odds and limits. No source measures it; the figures in “The upside” are our estimates.

Sources

Does It Pay collection of 11 tweets on this scheme, collected 2026-10-06.
Whop, Royalty Wagers store page. https://whop.com/royaltywagers (also https://whop.com/discover/royaltywagers/). Page as shown on 2026-10-10. Read at source on 2026-10-10.
Whop, Royalty Wagers Premium product page. https://whop.com/royaltywagers/royalty-wagers-premium. Page as shown on 2026-10-10. Read at source on 2026-10-10.
Whop, TommyBANKERS products page. https://whop.com/tommy-bankers/products. Page as shown on 2026-10-10. Read at source on 2026-10-10.
Whop, TommyBANKERS Platinum Tier page, including the affiliate setting in the page’s data. https://whop.com/tommy-bankers/platinum-tier-af. Page as shown on 2026-10-10. Read at source on 2026-10-10.
Telegram, TommyBANKERS public channel. https://t.me/s/tommybankers (subscriber count at https://t.me/tommybankers). Posts dated 2026-09-09 to 2026-10-05. Read at source on 2026-10-10.
Whop, fees page. https://docs.whop.com/fees. Undated. Read at source on 2026-10-10, through a tool that returns an extract.
Whop, prohibited products and services. https://whop.com/prohibited-products-and-services/. Undated. Read at source on 2026-10-10.
Whop, what is not allowed on Whop. https://docs.whop.com/trust-and-safety/trust-safety-overview/what-is-not-allowed-on-whop. Undated. Read at source on 2026-10-10.
Whop, terms of service. https://whop.com/tos/. Last updated 2026-10-09. Read at source on 2026-10-10.
University of St. Gallen repository, record for the Blogabet tipster dataset and paper (Gruettner, Wambsganss and Back, European Journal of International Management, 2021). https://www.alexandria.unisg.ch/259455. Record dated 2020; paper 2021. Read at source on 2026-10-10 (summary only).
Joseph Buchdahl, “Luck and skill in sports betting”, football-data.co.uk. https://www.football-data.co.uk/blog/luck_skill_sports_betting.php. 2013-05-30. Read at source on 2026-10-10, through the r.jina.ai reader.
Joseph Buchdahl, “Beat the market”, football-data.co.uk. https://football-data.co.uk/blog/beat_the_market.php. 2016-01-26. Read at source on 2026-10-10.
Joseph Buchdahl, “Sports betting winners and losers”, football-data.co.uk. https://football-data.co.uk/blog/sports-betting_winners_and_losers.php. 2013-12-10. Read at source on 2026-10-10.
Joseph Buchdahl, “Tipster verification”, football-data.co.uk. https://football-data.co.uk/blog/tipster-verification.php. 2020-04-07. Read at source on 2026-10-10.
football-data.co.uk, “Pinnacle’s Limit as a Confidence Meter: How Much of Your Value Survives to the Close?”, on 61,979 logged value bets (data from the tool vendor ValueBetFactory). Shown as the newest article on the blog index, where it will not stay; the page shows no date or byline, and we take the author to be the site’s owner, Joseph Buchdahl. https://www.football-data.co.uk/blog/. 2026, data to October 2026. Read at source on 2026-10-10.
Joseph Buchdahl, “The Wisdom of the Crowd” (updated). https://www.football-data.co.uk/The_Wisdom_of_the_Crowd_updated.pdf. 2017-06-23. Read at source on 2026-10-10.
Kaunitz, Zhong and Kreiner, “Beating the bookies with their own numbers”, arXiv 1710.02824. https://arxiv.org/abs/1710.02824 (quotes from version 2, https://arxiv.org/pdf/1710.02824v2). Version 1 dated 2017-10-08, version 2 dated 2017-11-11. Read at source on 2026-10-10.
arXiv preprint 2604.08251, a study of three Nigerian betting influencers’ slips. https://arxiv.org/abs/2604.08251 (body at https://arxiv.org/html/2604.08251v1). 2026-04-09. Read at source on 2026-10-10 (abstract direct; body through an extract tool).
Coincub, republished release of the Tips.GG index of 165 football prediction services. https://coincub.com/?p=14530. 2026, covers 1 January to 31 August 2026. Secondary.
University of California, press release on the study of 717,724 online gamblers. https://www.universityofcalifornia.edu/news/legalized-gambling-increases-irresponsible-betting-behavior-especially-among-low-income (also https://today.ucsd.edu/story/legalized-gambling-increases-irresponsible-betting-behavior-especially-among-low-income-populations, 2024-07-23). 2024-07-24. Read at source on 2026-10-10; secondary for the paper.
Taylor, McCarthy and Wilbur, “Online Gambling Policy Effects on Tax Revenue and Irresponsible Gambling” (working paper, copy hosted by gamblingharm.org). https://gamblingharm.org/wp-content/uploads/2025/09/Online-Gambling-Policy-Effects-on-Tax-Revenue-and-Irresponsible-Gambling.pdf. 2024-06-18. Read at source on 2026-10-10 by one agent; the same address refused another.
Bank of America Institute, report on online betting platforms. https://institute.bankofamerica.com/content/dam/economic-insights/online-betting-platforms.pdf. 2026-09-01. Read at source on 2026-10-10.
Gambling Commission (Great Britain), “Commercial restrictions by betting operators”. https://www.gamblingcommission.gov.uk/blog/post/commercial-restrictions-by-betting-operators. 2025-07-23. Read at source on 2026-10-10.
Massachusetts Gaming Commission, minutes of the open meeting of 30 September 2025. https://massgaming.com/wp-content/uploads/Meeting-Minutes-9.30.25-OPEN.pdf. 2025-09-30. Read at source on 2026-10-10.
CasinoBeats, report on the Massachusetts Gaming Commission’s data on limited bettors. https://casinobeats.com/2025/10/01/massachusetts-gaming-commission-sportsbooks-notify-bettors-limited/. 2025-10-01. Secondary.
CDC Gaming, brief on the same meeting. https://cdcgaming.com/brief/winners-face-betting-limits-losers-made-vips-mass-gaming-commission-learns/. 2025-10-02. Secondary.
InGame, report on Illinois sports betting figures for February 2026. https://www.ingame.com/?p=47813. 2026-04-14. Secondary.
RG.org, report on Illinois sports betting revenue for May 2026. https://rg.org/news/gambling-industry/illinois-may-2026-sports-betting-revenue-report. Published 2026-07-15, updated 2026-07-17. Secondary.
Internal Revenue Service, Topic no. 419, Gambling income and losses. https://www.irs.gov/taxtopics/tc419. Last reviewed 2026-09-24. Read at source on 2026-10-10, through the r.jina.ai reader.
KPMG, paper on gambling losses under the 2025 US tax law. https://kpmg.com/kpmg-us/content/dam/kpmg/pdf/2025/gambling-losses-under-one-big-beautiful-bill.pdf. 2025. Secondary.
HM Revenue and Customs, Business Income Manual BIM22017. https://www.gov.uk/hmrc-internal-manuals/business-income-manual/bim22017. Updated 2026-08-04. Read at source on 2026-10-10.
MyBetFund, home page. https://mybetfund.com/. Undated. Read at source on 2026-10-10, directly and through the r.jina.ai reader.
WagerFunding, “What is funded sports betting”. https://wagerfunding.com/what-is-funded-sports-betting/. Footer dated 2025. Read at source on 2026-10-10.
Fan Funded, home page. https://fanfunded.io/. Footer dated 2026. Read at source on 2026-10-10.
BankrollU, challenges page. https://www.bankrollu.com/challenges. Footer dated 2026. Read at source on 2026-10-10.
BETfunded, home page and guide “How sports prop firms work”. https://betfunded.com/ and https://betfunded.com/en/guides/how-sports-prop-firms-work/. Checked date on site 2026-10-03. Read at source on 2026-10-10; secondary for the firms it describes.
Prop Firm Reviews, funded betting review. https://propfirmreviews.net/funded-betting-review/. Updated September 2026. Secondary.
Trustpilot, reviews of FBMO. https://www.trustpilot.com/review/fbmo.net. Latest review read 2026-06-04. Read at source on 2026-10-10.
Trustpilot, reviews of Wagerfunding. https://www.trustpilot.com/review/wagerfunding.com. Latest review dated 2026-08-23. Read at source on 2026-10-10, through the r.jina.ai reader by one agent; the site refused a direct fetch and a later reader attempt.
Ankura and U.S. Integrity, “Match-Fixing on the Dark Web” white paper. https://ankura.com/insights/match-fixing-on-the-dark-web-a-white-paper-by-ankura-and-u-s-integrity. 2024-01-25. Read at source on 2026-10-10.
DubClub, partner sign-up page and refund policy. https://dubclub.win/partners/signup/ and https://dubclub.win/refund-policy/. Sign-up page as shown on 2026-10-10; refund policy updated 2024-01-26. Read at source on 2026-10-10.
DubClub, RipsPicks seller page. https://dubclub.win/ripspicks/. Page as shown on 2026-10-10. Read at source on 2026-10-10.
TechNexus, repost of press coverage of DubClub. https://www.technexus.com/dubclub-makes-it-easier-for-sports-gamblers-to-find-and-follow-betting-experts/. 2024-09-17. Secondary.
YesPress, press page citing DubClub figures for October 2025. https://yespress.io/lewis-burik. Undated, about late 2025. Secondary.
1xBet partner program. https://1xpartners.com/. Page as shown on 2026-10-10. Read at source on 2026-10-10, through the r.jina.ai reader.
Vice, “How Twitter tipsters are profiting on losing bets”. https://www.vice.com/en/article/how-twitter-tipsters-are-profiting-on-losing-bets/. 2017-10-01. Secondary.
Decision Marketing, report on Sky Betting & Gaming closing its affiliate program. https://www.decisionmarketing.co.uk/news/sky-betting-gaming-axes-affiliate-marketing-scheme. 2017-09-06. Secondary.
CasinoBeats, report on X’s paid partnerships policy. https://casinobeats.com/2026/02/18/x-updates-paid-partnerships-policy-ban-gambling-promotions/. 2026-02-18. Secondary.
Advertising Standards Authority (UK), advice on betting and gaming tipsters. https://www.asa.org.uk/advice-online/betting-and-gaming-tipsters.html. 2025-05-22. Read at source on 2026-10-10, through a tool that returns an extract.
Advertising Standards Authority (UK), ruling on Isiris Racing Services. https://www.asa.org.uk/rulings/Isiris-Racing-Services-A13-251637.html. 2014-08-13. Read at source on 2026-10-10.
Hungarian Competition Authority (GVH), press release on case VJ/23/2021. https://gvh.hu/en/press_room/press_releases/press-releases-2023/heres-a-useful-tip-from-the-gvh-there-is-no-such-thing-as-extra-safe-sports-betting. 2023-04-25. Read at source on 2026-10-10.
US Securities and Exchange Commission, litigation release 24585. https://www.sec.gov/litigation/litreleases/lr-24585. 2019-09-04. Read at source on 2026-10-10.
US Department of Justice, Eastern District of Missouri, press release on a sports betting fund sentence. https://www.justice.gov/usao-edmo/pr/man-who-defrauded-investors-sports-betting-fund-sentenced-48-months-prison. 2025-01-22. Read at source on 2026-10-10, through the r.jina.ai reader.
CDC Gaming, brief on West Virginia bill HB 3232. https://cdcgaming.com/brief/west-virginia-bill-to-audit-betting-touts-fails-maryland-effort-still-alive/. 2023-03-19. Secondary.
Discord, gambling policy explainer. https://discord.com/safety/gambling-policy-explainer. 2024-03-15. Read at source on 2026-10-10.
Pinnacle, “Winners welcome at Pinnacle”. https://www.pinnacle.com/betting-resources/en/educational/winners-welcome-at-pinnacle. 2024-03-21. Read at source on 2026-10-10, through the r.jina.ai reader.
Racing Post, report on the Betfair Exchange fee for 2025. https://racingpost.com/news/britain/betfair-exchange-to-introduce-new-commission-system-for-2025-as-premium-charge-is-dropped-a7wbg0v4GCAJ. 2024-12-18. Secondary (read through the r.jina.ai reader; the site refused a direct fetch).
Brighton and Hove News, report on a High Court claim about a betting syndicate. https://www.brightonandhovenews.org/?p=220813. 2025-12-04. Secondary.
Smart Betting Club, tipster profit reports page. https://smartbettingclub.com/tipster-profit-reports-sbc/. Reports dated 2022 to 2024. Secondary.
Karl Whelan, post on closing line value in NBA moneyline bets. https://www.karlwhelan.com/?p=2595. 2026, undated as read. Read at source on 2026-10-10.
Maryland General Assembly, Senate Bill 621 of 2023, “Gaming - Sports Wagering - Independent Evaluation of Sports Wagering Content” (approved by the Governor, Chapter 638). https://mgaleg.maryland.gov/mgawebsite/Legislation/Details/sb0621?ys=2023RS. Approved 2023-05-16, effective 2023-07-01. Read at source on 2026-10-10.
Brown and Reade, “The Wisdom of Amateur Crowds: Evidence from an Online Community of Sports Tipsters”, European Journal of Operational Research, accepted manuscript, DOI 10.1016/j.ejor.2018.07.015. https://ueaeprints.uea.ac.uk/67529/4/Accepted_manuscript.pdf. Accepted 2018. Read at source on 2026-10-10.

Corrections

If you are quoted or named on this page and think something is wrong, want your reply shown beside it, or want to be removed, write to [email protected]. We aim to reply within 14 days, and always within one month. Factual errors are corrected with a dated note here. A sentence that is seriously disputed comes down while we check it.

Change log: (empty)