June on the Record: What the Model Got Right, and What It Didn't

Every month we publish this piece for the same reason: if you're going to trust a statistical model, you should be able to see everything it said - not a highlight reel. June 2026 gave us a bit of everything: an outright winner in the Netherlands, another in Canada, and a couple of weekends where our names went home on Friday. All of it is below, taken directly from our public results page.

One housekeeping note before we start: our current top-four selection format was applied retrospectively to events before 4 June 2026, so the earliest entries on the results page reflect that standard applied after the fact. Everything from the KLM Open onwards was published in the normal way.

The wins

KLM Open (DP World Tour, 4 June). Eugenio Chacarra won. When the model flags a player and that player lifts the trophy, it's tempting to shout about it - we'd rather just note it. Chacarra was one of four names we put forward that week, and the data liked his profile more than the market did. Hillier backed it up with 14th. That's the good half of the story; we'll get to the other half shortly.

RBC Canadian Open (PGA Tour, 11 June). This was the strongest single week of the month. Bud Cauley won, and Matt Fitzpatrick - another of our four - finished second. Tommy Fleetwood added an 11th. When the top of a leaderboard looks like your selection sheet, the model has done its job. It won't look like that most weeks, which is exactly why we're pointing it out while it's true.

U.S. Open (18 June). Majors are the hardest events to model - deep fields, brutal setups, small margins. Fleetwood's 11th was a genuinely solid return in that context, and Fitzpatrick's 22nd was respectable. Neither is a headline, and we're not pretending otherwise, but top-25 finishes at a U.S. Open are the sort of quiet, unglamorous results a sound process should produce.

Open d'Italia (DP World Tour, week of 25 June). Joaquin Niemann finished third and Angel Ayora fifth. Two names inside the top five is a good week by any sensible standard, even without a win.

The misses

This section gets equal billing, because a results page that only remembers the good weeks isn't a results page.

The missed cuts. Aaron Rai missed the cut at the RBC Canadian Open - the same week Cauley and Fitzpatrick went first and second. Si Woo Kim missed the cut at the U.S. Open. There's no way to dress up a missed cut and we won't try: the model liked both players, and both were watching the weekend from home. This is what golf does. The same process that put Cauley near the top of our Canadian Open sheet put Rai on it too.

The anonymous finishes. The KLM Open win came with baggage: Antoine Rozner finished 55th and Lindberg 60th. Russell Henley was 65th at the U.S. Open. Patrick Reed (45th) and David Puig (48th) did little at the Open d'Italia. These are the results that quietly vanish from most people's marketing. They're staying on ours.

The Travelers Championship (25 June) deserves its own line because it was neither a win nor a disaster - it was simply ordinary. J.J. Spaun's 7th was the best of it; Patrick Cantlay finished 14th, Rai bounced back from Canada with 30th, and Si Woo Kim recovered from his U.S. Open exit to finish 44th. A perfectly forgettable week. Most weeks are, and it's worth saying so.

What to take from a month like this

Three things, none of them dramatic.

Variance is the whole game. The same model, run the same way, produced a winner and a missed cut in the same field in the same week - twice, if you count Kim's U.S. Open exit alongside Fleetwood's 11th. That's not the model malfunctioning; that's golf. Seventy-odd players make a cut and one wins. Any process that claimed to avoid missed cuts entirely would be lying to you.

Judge the process over months, not weekends. A week like the Canadian Open proves nothing on its own, and neither does a missed cut. What matters is whether, over a long run of events, the players the data flags perform better than the market expects them to. One month - even a decent one - is a small sample, and we'd be the first to say June flattered us in places.

This is why we publish everything. ProPlace is a research tool. We look for spots where the data disagrees with the market, and we show our work - every selection, every finish, every Friday exit, on a public results page you can check yourself. We don't take stakes, we don't tip, and we don't tell you what to do with the analysis. What you're owed in return for your attention is the full record, and this piece is part of keeping that promise.

If you'd like to see how the model reads the upcoming fields, the full research is available with a 7-day free trial - no card required, cancel anytime.

Past performance is not a guide to future results. 18+. Please gamble responsibly. Visit BeGambleAware.org.