Canada’s opener at BMO Field in Toronto, second game of the tournament. The hosts get a winnable spot to start, the kind of fixture you have to bank when you’re a co-host trying to build momentum through a group.

Canada qualified automatically as a host nation, so like Mexico the friendlies-vs-competitive trade-off applies. The squad’s identity has been built around speed in transition (Davies, David), with a possession base that has improved match by match through the cycle. They flopped at Qatar 2022 with zero points; the project is now four years older.

Bosnia & Herzegovina are the awkward draw the Elo system makes look easy. Rating ~1452 reflects a brutal stretch since 2020 — failed qualifications, generational handover from the Pjanić/Džeko era — and may overstate the gap to a Canada side they would historically be much closer to. They concede a lot in friendlies and look listless in dead rubbers, then suddenly turn out a coherent performance in a competitive opener.

Model view. Canada 92% / 8% / 0.4%, xG 2.60–0.05. The 92% is too high. The independent-Poisson framework breaks down at this Elo gap; the away xG floor of 0.05 is mechanical, not real. In any honest read this is a 75–80% Canada match, and BiH will create at least one or two real chances.

Model vs market. The mispricing here is in the over/under markets, not the moneyline. Canada ML will be ~1.20–1.25, the market will be efficient. Where to look:

• BTTS-Yes above ~2.40 — model says 5% (don’t trust that), realistically 25–30%, only fire if odds reach 3.50+
• Canada -1.5 AH is where the volume should go if odds get above ~1.95 (model implies ~67% but actual probably 55–60%, so price-dependent)
• Over 2.5 at 1.85+ — Canada at home historically gets to 3 against weaker sides

Calibration note. This match is the perfect candidate to revisit after Day 1 results: if BiH look better than 1452 suggests, the model’s Group B downstream predictions need a manual rating override.