The battle for Group H's realistic second spot (Spain permitting). Cape Verde's central-compact-then-counter identity against Saudi's possession-leaning mid-block makes Saudi the default ball-carrier — historically not their best costume. Pina's screening against Al-Juwayr's free movement is the game's axis; Livramento and Mendes breaking at a Saudi back four that holds a moderate line is Cape Verde's clearest route in any game this group.
Key Duel
Kevin Pina vs Musab Al-Juwayr — the screen against the wanderer.
Watch For
Al-Dawsari's last-stand tournament; Cabral's progressive carries; nerves — for both, likely the biggest match in federation history.
Margin Notes
Cape Verde would be among the smallest nations ever to reach a knockout round (verify framing). Donis would hit five competitive games in charge here.
Selection notes were pre-baked June 11 and are verified day-of in the edition, not here — anything marked “verify” must be confirmed before it is load-bearing.
The Market
Odds & best bet
model projection: 2.44 total goals · over 1.5 70% · over 2.5 44% · over 3.5 23%
Market odds versus the model: implied probability, our probability, and the edge per selection
Market
Selection
Odds
Implied
Ours
Edge
Total goals
Over 2.25
1.93
50%
44%
-5.8%
Total goals
Under 2.25
1.93
50%
56%
+5.8%
Total goals
Over 2.5
2.05
45%
44%
-1.3%
Total goals
Under 2.5
1.71
55%
56%
+1.3%
Asian handicap
Cape Verde -0.25
1.93
49%
39%
-10.2%
Asian handicap
Saudi Arabia +0.25
1.89
51%
61%
+10.2%
Best bets — top 2, ≥5% edge
1. Saudi Arabia +0.25 (Asian handicap) @ 1.89, edge +10.2% — best price 1.89 (bovada). Flat 1u, paper record.
2. Under 2.25 (Total goals) @ 1.93, edge +5.8% — best price 1.93 (betonlineag). Flat 1u, paper record.
same-match picks are correlated — they tend to win and lose together; the units record swings accordingly.
market snapshot Jun 12, 7:00 PM ET · median odds across books, de-vigged multiplicatively · totals / handicap / BTTS are model-priced from the score matrix — the Opta overlay covers W/D/L only · no logged consensus prediction — 1X2 edge not computed.