Best Golf Tipsters 2026: Compared and Ranked

Most "best golf tipster" lists are affiliate pages with invented numbers. This one only includes services whose results are published somewhere you can go and read, and every row links to its source so you can check it rather than take our word for it. Where a service doesn't publish a figure, we leave it blank instead of filling it in.

Ranked by return on investment on tournament betting - outright, each-way and top-finish. Services betting matchups or 3-balls are playing a different game and are listed separately below rather than forced into the same column.

The ROI column is no longer anybody's own word for how they did. Changed 15 September 2026. Where more than one party has published a record, this table adds the profits together, adds the stakes together, and divides once - the return you would have got by following every source at once. Where a service's only evidence is its own claim, that claim still appears and still ranks, but the row says No independent record and the number is the service's, not ours to vouch for.

Why pooled rather than averaged. Averaging two percentages would let a six-month trial count for exactly as much as an eleven-year audited record. Adding the money instead weights each source by what it actually risked. It changes the table a lot: PGA Profit moves from second on the page to sixth, because an eighteen-month independent trial finished at -25% and that now counts; Golf Insider moves up, because an outside trial beat its audited record. We did not pick which way those went.

This table is sorted by ROI, highest first. That is not the same as sorted by best. The sort is arithmetic - it reads one number and puts the biggest at the top. What makes a number worth acting on is the line underneath the name: how many independent parties checked it, over how long, and whether the record was kept in advance or reconstructed afterwards. On this page those two things run in almost opposite directions. The top two figures have no independent backing at all. The two rows resting on two independent sources each sit third and sixth. If you read only the ROI column you will read this table backwards.

Read the Period column before the ROI column. Every service publishes over the window it chooses, and those windows are not the same. A two-year record and a twelve-year record are not the same evidence even when the percentage matches, and a short window during a good run will always flatter whoever is quoting it - us included.

Every number on this page is an upper bound, ours included. A published ROI is what the advised price returned, not what a subscriber banked. Tips move the market - a service with thousands of members shortens its own prices within minutes of sending - and the advised price is the best one on offer at the moment of advice, which is rarely the one still on the board when you get there. Winning accounts get restricted. Each-way terms differ between books and tighten once a tournament starts. None of that is an accusation against anybody listed here; it applies to us on both our sites. It does mean you should read the ROI column as the ceiling of what the selections were worth, not as the floor of what you would have made.

Somebody actually followed three of these for real, and published what happened. Honest Betting Reviews runs forward trials: they subscribe, place the advised bets as they are sent, and report the result at the end. That is a harder test than any back-record, because nothing in it can be reconstructed afterwards. Their three golf verdicts:

• PGA Profit - 18 months to March 2023: -289 points, -25% ROI, 58% of the bank gone, 8% strike rate. Rated FAILED.
• Golf Insider - 6 months to December 2020: +405.7 points, +47% ROI, bank up 81%. Rated PASSED.
• Golf Betting Club - 9 months to February 2025: -17 points, -2% ROI. Rated NEUTRAL.

Now read the ROI column again. PGA Profit has the best-audited long record on this page and the worst independent forward result on it. Both are true: the proofed 23.5% happened, and so did an eighteen-month stretch that cost a real subscriber more than half their bank. Golf Insider's trial went the other way, but six months is six months, and their own proofed record runs at 19.2% rather than 47%. None of these trials is long enough to overturn a multi-year record, and we have not let them move anybody up or down the table. They are simply the best evidence on this page of the gap between a number a service publishes and what it is like to pay for it.

Second place is two years, and one of them carries it. Steve Palmer's 25.8% is published by the Racing Post and we have left it exactly as it stands. But 2024 was extraordinary - 39.8% ROI on the back of 19 winners - while 2025 came in at 10.81%. On 2025 alone he sits below PGA Profit, below Golf Insider and below our own default setting. Racing Post also publishes no audited results table for him; the figures come from the blurb on his own tip articles, which is the publisher vouching for its own tipster. Seven proofed years at 23.5% is stronger evidence than two self-reported years at 25.8% - the sort does not know that, which is why it is written here instead. In fairness to him, that blurb also claims 21 profitable seasons out of 24 - a far longer record than the two years he publishes numbers for, and one nobody outside Racing Post can check either way.

On the drawdown column, and why most of it says "not published". ROI tells you what a service returned. Drawdown tells you how far underwater you'd have been while it did - and it is the number that actually decides whether you're still subscribed at the end. Almost nobody publishes it. Where a figure appears above, it comes from one of three places: Smart Betting Club measured Coley's worst run at 344 points and recommends a 500-600 point bank to survive it, and Golf Betting Club publishes every event's profit and loss, so we added them up ourselves - a peak in June 2023 to a trough in late 2025, 376 points down over roughly two and a half years. GolfForecast publishes its record the same way, event by event, and the same arithmetic gives 261 pts - all of it in the first eight months of 2020, before the run that carries the rest of that row. For PGA Profit, Golf Insider and Steve Palmer there is no drawdown to quote and we have not invented one; the count of losing years is what those sources actually support.

Points aren't comparable between services, so scale them. A 376-point drawdown on Golf Betting Club's 7,129 points staked is a different injury from 376 points on a small bank. Measured against turnover, Golf Betting Club's worst run cost 5.3% of everything it staked, GolfForecast's cost 2.7% and ours cost 4.0% - but note that our figure is a backtest, GolfForecast's is a recalculation, and Golf Betting Club's is what really happened to real subscribers. Those are three different kinds of evidence, it is not a like-for-like comparison, and we are not going to pretend it is.

On Betting Gods' number, and why it is not measuring the same thing. Their tipster describes the service himself as "each-way opportunities, matchups, and placings alongside outrights". Every other row in this table is tournament betting only. A matchup is a near coin-flip priced around evens; an outright is a 100/1 shot that loses forty weeks running. Blending them into one percentage is not dishonest, but it is not answering the same question as the rows above, and a 12.6% built partly on evens shots is a different achievement from 12.6% on outrights. Two other things to weigh: the window is about nineteen months, the shortest on the page by years, and Betting Gods both sells the subscription and keeps the record. What the row does have is volume - 3,756 published bets - and a drawdown they publish rather than one we had to derive.

On PGA Profit's number, and why it moved. Three figures exist. Their own site advertises 2,137 points at 30%. Smart Betting Club, who received the selections before each event, report roughly 1,400 points at 23.47% over seven years. Honest Betting Reviews followed the tips forward for eighteen months to March 2023 with their own money and finished 289 points DOWN, at -25%, an 8% strike rate and a bank 58% smaller than they started with.

The 30% is the service's own and is not counted. Pooling the other two - 1,111 points of profit from about 7,121 staked - gives +15.6%, and that is the number in the column. It is lower than the audited figure and far higher than the trial, which is what pooling is for: neither source gets to be the whole story. The honest summary is that a real audited edge and a real eighteen-month losing run both happened, and a subscriber joining in October 2021 got the second one.

On Ben Coley. He gives his tips away free at Sporting Life and publishes every advised bet, year by year, which is more transparency than most paid services manage - ten published years running to +2,757 points. We quote Smart Betting Club's 23.06% rather than that total because Sporting Life doesn't publish what was staked, and an ROI needs both halves. Read that 23.1% as a 2017-21 figure, not a current one. Smart Betting Club's piece is dated July 2022, and its 1,279 points is precisely Coley's Sporting Life total for 2017 to 2021 (316.45 + 50.73 - 145.12 + 363.56 + 692.51 = 1,278.13). So the window ends on his best year, +692.51 in 2021. He has been profitable every year since - +585, +312, +255, +136 - but the trend is downward, and no ROI can be quoted over those years because the stakes are not published. Note also that SBC tracks him rather than proofing him: he's a newspaper columnist, not a service submitting selections for audit. His own estimate is 20-25%, which we've left out of the column for the same reason we discount Racing Post's numbers for Palmer.

We are fifth of nine on our own list, and our two rows are the least-checked on the page. The sort put us there and we have left it alone. Both of our entries carry No independent record, for the same reason Steve Palmer's does: nobody outside this company has audited either of them. Golf Tipsters' figure is a walk-forward backtest, GolfForecast's is a recalculation from stored odds, and the rule we apply to everyone else applies here - our own word does not count towards a score, so neither number is pooled with anything. If an independent reviewer ever trials either service, that figure goes in the column and ours gives way to it.

The number at the top of this table is ours, and it is the weakest evidence on the page. GolfForecast's +43.0% is nearly double anything else here, and a strict ROI sort puts it first. Read what it is before you read what it says. It is self-reported - our site, our arithmetic, nobody independent has checked a line of it. It is also recalculated: GolfForecast re-derives every past tip from the odds it stored at the time, because it never kept a log of the tips it actually sent. Each event is scored under the rules that were live when it ran, so it is not a backtest either - but it is not a proofed record, and it is not the same evidence as seven years of selections a third party received in advance. On the standard this page applies to everybody else, that basis loses to a proofed figure however big the number is. Treat first place here as what a sort does, not as a verdict we are asking you to accept. One more thing worth knowing: those are UK each-way bets. If you bet at a US sportsbook you cannot place them at all.

On Data Golf, now that it is in the table

It is in the table from 15 September 2026, and it is still not the same product. Data Golf bet matchups, 3-balls and finishing position - not the tournament outrights every other row is built on. It sits in the table because a reader comparing golf betting services will look for it, and leaving it out told them less than including it with the difference stated plainly. Read its row as "what this analytics operation returned on the bets it actually made", not as a like-for-like against a tipster selling outrights.

The figure is our arithmetic over their own published bet log, not a number they quote: 9,831 bets across 2019-21, +214.3 units from 5,458.5 staked, a pooled +3.9%. Per year that is +2.90%, +4.51% and +5.01%. Their log publishes the stake and the return on every single bet, which is more than any other service here does, so the arithmetic is checkable by anyone who wants to redo it.

Two things that catch people out. The stakes are variable, not flat - a bet count is not a stake count, and dividing profit by the number of bets gives a materially different and wrong answer. And the total hides the spread: 72-hole matchups ran +5.7%, then -9.0%, then +12.5% across those three years, while one-round matchups ran -1.2%, +3.4%, +1.1%. Finishing-position bets carried the record at +8.1% while one-round matchups, the largest block of stake, returned +1.9%. A single market over a single season is exactly the kind of figure a tipster page would otherwise quote at you.

They stopped posting bets at the start of 2022, so this is a closed record rather than a live one. It is a serious analytics operation and worth your time. (their results)

Services advertising ROI figures above 100% have been left off entirely. Those numbers exist, but they come from short windows chosen after the fact, and putting them in the same column as a twelve-year record would make this table worse rather than more complete.