A tipping service in Australia is a business that sells betting selections: someone publishes which horse, team or player to back, usually with a price and a stake beside it, and you pay for the right to see it. The service is not a bookmaker. It takes no bets, sets no prices and holds none of your money, so every tip is a bet you place yourself, at whatever price is left by the time you get there.
You cannot judge the tips in advance, so judge the two things you can check before paying: the shape it is sold in, which decides who pays the tipster and so what the tipster is rewarded for, and whether its record is the kind of record that can be checked at all. Most are not, and the reason is structural rather than dishonest.
The shapes a tipping service comes in
Five shapes cover most of them, and plenty of services run two at once.
| Shape | How the tips reach you | Who pays the tipster |
|---|---|---|
| Subscription feed | A members page, app or email | You, weekly or monthly |
| Discord or Telegram group | A private channel, as messages | You, a joining or monthly fee, sometimes nothing |
| One-off package | A race day, carnival or meeting, sold as a pack | You, once |
| Free tips with bookmaker links | A public feed, a sign-up link beside each tip | A bookmaker, per sign-up or as a share of losses |
| Mentorship bundle | Tips plus lessons, spreadsheets or a staking plan | You, a larger up-front fee |
A monthly subscription you can cancel is a small repeating bet on the tipster, and a bad month is cheap information. A package and a mentorship fee are paid in full before you have any evidence, and a losing week refunds neither. A messaging group has its own problem: the tips and the record are the same posts, and a post can be edited or deleted.
Racing tips and sports tips are sold differently
A horse racing tipping service usually sells by the meeting or the day, and quotes a price that can move sharply in the last minutes before the jump. A sports tipping service tends to sell by the round or the month and leans more on lines and totals. The checks are identical, but racing tips are much harder to get at the advised price, and a tote tip harder still, because a tote approximate is not a price you can take and hold.
What a tipping service is not
Not a bookmaker, so it is not on the register of licensed interactive wagering providers the Australian Communications and Media Authority publishes. Use that register on the bookmaker a tip points you towards: how to check a bookmaker is licensed covers reading it. Not a fund either. A service that asks you to send money for someone else to bet with, or offers a share of profits on a pooled account, is selling something else, and that is the shape Scamwatch and the ACCC publish warnings about: see sports investment scheme scams.
Who pays the tipster, and what that rewards
Three answers, pulling in different directions.
- You pay, by subscription or package. The income depends on you staying subscribed, which rewards results good enough to keep you there.
- A bookmaker pays, either once for each new customer the tipster's link brings in, or as a share of what referred customers lose, month after month. On the second kind the tipster earns most in the months their followers do worst.
- Both, which is common: a subscription on top of bookmaker deals.
None of this makes a tip wrong. It changes what the tipster is paid to do, and the clearest tell is the link: a feed that always quotes the same bookmaker is routing you rather than finding you the price. Affiliate tipsters covers the deal types, the disclosure the ACCC expects beside a paid link, and the ban on activity-based commissions from 1 January 2027.
Free tips cost something: a worked comparison
A bookmaker-funded feed is not free. The cost moves out of a fee and into the price you take, which is measurable rather than a matter of opinion.
A $60 bet at $3.40 returns $204 if it wins. A price 2% shorter is 3.40 x 0.98 = $3.332, so the same $60 returns $199.92, which is $4.08 less, and $4.08 is 2% of $204. Across many tips a service betting near break-even returns roughly what it stakes, so 2% of the return is about 2% of the turnover, which makes it comparable with a fee. Illustrative figures: 12 tips a week at $60 each, a turnover of 12 x $60 = $720.
| What you pay | Weekly cost on $720 of turnover |
|---|---|
| $40 a week, best price on offer taken | $40, which is 40 / 720 = 5.6% of turnover |
| Nothing, linked price 2% shorter | 0.02 x $720 = $14.40 |
| Nothing, linked price 5% shorter | 0.05 x $720 = $36.00 |
| Nothing, linked price 8% shorter | 0.08 x $720 = $57.60 |
The break-even shortfall is the fee over the turnover: 40 / 720 = 5.6%. Under that the free feed is cheaper even though it sends every tip to one bookmaker, and over it the subscription is. The two behave differently as your stake changes: the fee is fixed, so 52 x $40 = $2,080 a year whether you bet or not, while the price cost grows with every dollar you turn over.
Note: this compares two ways of paying and nothing else. Neither sum says whether either service's tips are any good. Are tipping services worth it runs the fee against turnover in full.
Why a published record is almost never verifiable from the outside
A record is written by the party it judges. That alone would not be fatal, since plenty of businesses report their own numbers. What makes it close to unfalsifiable is that three things are invisible to a reader, and each can be worth more than the edge being claimed.
The denominator. You see the tips in the record. You cannot see the tips never published, posted and then deleted, or quietly re-labelled as something other than a tip. Take an illustrative 100 level-stake tips showing 12 units of profit, a return on turnover of 12 / 100 = 12%. If eight losers were dropped from a real 108, the honest figures are 12 - 8 = 4 units on 108 staked, or 4 / 108 = 3.7%. Eight deletions out of a hundred cut the real figure to under a third of the headline, and the record shows no gap.
The price. A record states an advised price. By the time you read it the market has moved, and there is no way from outside to establish that the price was ever available, to anybody, for any stake.
Which record you were shown. You saw the service that advertised to you, or the one at the top of a leaderboard, and a lucky run buys both positions. Services that went badly stop advertising, so the records placed in front of you lean towards the fortunate ones. That is survivorship bias, and it works on honest services too.
Risk: Betting involves risk. A record that survives every check describes the past only: past results do not predict future results, and an edge that was real can fade. See responsible gambling for limits and support.
Who would have to vouch for a record, and why that decides it
The fields a complete record carries are set out in how to check a tipster record. The question this page is asking is narrower, and it is the one that decides whether a record can be verified at all: for each thing a record asserts, who is in a position to confirm it?
| What the record asserts | Who could confirm it | What a reader can do alone |
|---|---|---|
| The tip went out before the event started | Whoever stamped the time, if that is not the tipster | Nothing, unless the post carries a timestamp from outside |
| The advised price was really on offer | The bookmaker or exchange named, at that minute | Nothing after the fact, unless a price history for that exact selection and book happens to exist |
| Every tip is in the record, losers included | Only the tipster, unless the tips are numbered | Read the numbering for gaps |
| The stake is the one advised, not a revised one | Only the tipster, unless the original post cannot be edited | Check an unedited channel, where there is one |
| Settlement is right, voids and scratchings included | Official results, which anybody can read | Recount every settled tip |
| The closing price quoted is the market's | The closing market, which anybody can read | Recompute closing line value, bet by bet |
Two of the six a reader can check outright, three only if the service happened to choose a format that leaves evidence, and one, whether the price was ever on offer, not at all. That is the structural answer, and it is not an accusation: the assertions most worth checking are the ones a reader has no standing to check unless the service chose a format that left the evidence behind, so "trust the record" collapses into "trust the tipster" however the record is presented.
The last row still earns its place, because closing line value settles faster than profit does and is the easiest of the headline figures to average away. A monthly average can be carried by one tip: a single bet at plus 40% and nineteen at zero average 40 / 20 = plus 2% for the month, which reads as a service that keeps beating the market and is nothing of the sort. Per bet, you see the spread instead of the mean, which is why a record has to show it bet by bet rather than month by month. Fake tipster red flags covers what tends to arrive instead of a record that can be checked.
Strike rate, ROI, turnover and CLV: why ROI flatters most
Four figures, four questions. A service quoting one of them is choosing which question to answer.
| Figure | The question it answers | How it is flattered |
|---|---|---|
| Strike rate = winners / settled tips | How often the tips win | Tip short prices: at an average winning price of $1.35, a 70% strike rate still returns 0.70 x 1.35 - 1 = minus 5.5% per $1 |
| ROI = profit / total staked | What a dollar staked returned | Weight the stakes, drop the losers, quote advised prices, or stop the clock after a good month |
| Turnover = total staked | How much money the sample covers | A handful of large stakes can be most of it |
| CLV = advised price / closing price - 1 | Whether the tips beat the market's last price | Average it over a month, or choose a closing price that suits |
ROI flatters most easily, because it is a ratio whose numerator and denominator are both chosen by the seller, and a weighted staking plan lets one bet carry the whole number.
Two illustrative services each report 200 settled tips, $20,000 of turnover and $1,200 of profit, so both show 1,200 / 20,000 = 6%. Service A staked $100 flat on all 200. Service B staked between half a unit and 10 units, and one 10-unit tip of $1,000 at $6.00 won, making 1,000 x (6.00 - 1) = $5,000 by itself. The other 199 tips therefore lost 5,000 - 1,200 = $3,800 on the remaining $19,000, a return of minus 3,800 / 19,000 = minus 20%. Same headline, and one is 200 pieces of evidence while the other is one.
ROI also carries no sample size, so 6% over 40 tips and 6% over 4,000 print identically (sample size in betting shows how many a figure that size needs), and it can be quoted against a betting bank rather than total stakes, which is a different figure again (ROI vs yield).
What no record can tell you
Even a record whose six assertions all check out leaves three questions open, and they decide what following a service does to your account. What happens next is one: the market a method beat is made of people who adjust, so past results do not predict future results, and that holds for a clean record as much as a dressed-up one. Your own result is another, because you bet after the tip, at the price left by then, in the accounts you hold, at stakes your own bank allows (copy betting risks covers what else moves in between). The third is whether it was skill at all: a record is evidence about an edge, never proof of one, and at long prices it takes a great many tips before the evidence outweighs the luck.
Where B337's copy betting fits
Copy betting on B337 is sold as the Tipping Pro plan, and the part that answers this page is where the figures come from. Each feed's strike rate, ROI, turnover and closing line value are worked out by B337 from the tips its tipster published through the platform and how those bets settled, beside how many tips have settled and when the feed started. The tipster does not type the numbers in, and nothing is counted from screenshots, spreadsheets or tips posted somewhere else. Publishing a feed needs approval from the B337 team.
The limits are the point. A record tracked this way is not a forecast, not your result, and not an audit, a rating or a recommendation, and it covers only the tips published through B337. Any signed-in account, a free one included, can browse the feeds and read those records before paying.
A subscription shows a feed's tips as they are published and places no bets by itself. Each tipster sets the weekly price of their own feed, shown before checkout. Copying is a separate step: you link one of your own bookmaker accounts, and the bot places a copy sized by your own settings, never the tipster's. The bot runs on your own computer, so a tip published while that computer is off, asleep or offline is missed. Tipping Pro is arranged with the team, and bets placed through B337 use credits, a per-bet usage charge, so add those to the fee side of the comparison above.
Risk: Betting involves risk. A tracked record covers past tips only, past results do not predict future results, and a copy can go on at a shorter price than the tip, or not at all. Many bookmakers restrict automated betting in their terms, and a bookmaker can limit your stakes, void bets or close your account. See responsible gambling for limits and support.
For free and confidential support call 1800 858 858 or visit gamblinghelponline.org.au.