What this ground actually produces, measured across 31 T20 matches of ball-by-ball records rather than described from memory.
The record shows a modest edge to batting first, worth about 13 runs.
Recent matches are scoring lower than the all-time record: 0.00 an over at the death against 8.60 across the full sample, so the older figures understate the difficulty of what happens now.
Average runs and wickets per phase, split by innings. This is where a ground gives itself away: a par score is not one number, it is a shape.
| Phase | Batting first | Chasing | ||
|---|---|---|---|---|
| Runs/over | Wickets | Runs/over | Wickets | |
| Powerplay (ov 1-4) | 7.16 | 1.0 | 7.27 | 0.7 |
| Early middle (ov 5-8) | 6.81 | 1.1 | 6.99 | 0.8 |
| Middle (ov 9-12) | 7.45 | 1.0 | 7.23 | 1.2 |
| Late middle (ov 13-16) | 7.45 | 1.4 | 6.84 | 1.3 |
| Death (ov 17-20) | 8.60 | 2.4 | 5.90 | 1.4 |
Chases stall here. The last four overs go at 8.60 an over batting first and only 5.90 batting second - a 2.69-run collapse in the phase that usually decides a T20.
What the total actually reads at the end of each phase — built from this ground's own measured run rates, not by dividing a final average into equal pieces. This is what "is 60 for 2 after 8 a good position" means here.
| Par score after | 4 ov | 8 ov | 12 ov | 16 ov | 20 ov |
|---|---|---|---|---|---|
| Batting first | 29-1 | 56-2 | 86-3 | 115-4 | 150-7 |
| Chasing | 29-1 | 57-2 | 86-3 | 113-4 | 137-5 |
| Phase | All-time | Last 0 | Change |
|---|---|---|---|
| Powerplay (ov 1-4) | 7.16 | 0.00 | ▼ -7.16 |
| Early middle (ov 5-8) | 6.81 | 0.00 | ▼ -6.81 |
| Middle (ov 9-12) | 7.45 | 0.00 | ▼ -7.45 |
| Late middle (ov 13-16) | 7.45 | 0.00 | ▼ -7.45 |
| Death (ov 17-20) | 8.60 | 0.00 | ▼ -8.60 |
Measured over-to-over variation averages 2.55 runs, which makes this about as predictable as a typical T20 ground. The death overs are the loosest part at 3.19.
A required rate means different things at different grounds. At Newlands, Cape Town the powerplay goes at 7.16 an over batting first, so a chase that is level with the rate after four overs is genuinely level — not ahead, and not behind. The phase table above is what the model uses to judge that, rather than a single national average.
Ball-by-ball records for every T20 played at this ground in the dataset, aggregated per phase and per innings. The same records feed the live win probability on this site, which scores 81.5% across 19,340 predictions on 2,546 matches it had never seen during training — the method and calibration table are on the accuracy page.
No ball-tracking is used: speed, swing and spin measurements are not available at this scale, and these figures are delivery outcomes only.
See live predictions · All 48 grounds · How win probability works