The piece's central argument is that this series has been decided not by the pack that dominates but by the team that converts entries into the 22 faster than the defence can reset — and that dynamic, combined with the pace of the FNB surface, is why the author tips New Zealand by three (27-24). The statistical backbone is striking: South Africa averaged 5.5 points per 22-entry this season but managed 0.7 per entry at Ellis Park while holding 61% territory; Cape Town flipped that entirely, with the Boks' first three entries yielding 17 points despite being outcarried 154-104. The argument isn't that the Springboks can't play here — it's that dry, quick Soweto ground gives New Zealand's wider threats the room Cape Town denied them, and that the Bok defensive structure bends at altitude when ball moves at pace through hands.
The piece frames two specific pressure points. First, the lineout: Codie Taylor dropped from the 23 entirely signals Rennie's acknowledgement of an 82.9%-vs-93.1% throwing gap, and the question of whether Aumua can throw against Marx in front of 90,000 people is identified as the single most important head-to-head of the match. Second, the 55-to-70-minute window — both previous Tests were settled there, and while the Erasmus closing mechanism (Steenekamp, du Toit, de Jager, Hanekom) remains the most reliable in Test rugby, it only works from in front. If Sititi and Lienert-Brown are on the field with the score level and the ball moving, the Boks will be defending rather than closing. Kolbe at fullback on a fast track, and what Feinberg-Mngomezulu does under pressure at altitude, are flagged as the individual variables that could swing three points either way.