Analysis · 1,870,356 deliveries
Spin is cheapest exactly where T20 uses it least.
At the death, spin concedes 0.82 runs an over fewer than pace — the largest gap of any phase — and takes wickets at the same rate. Teams bowl it 15.7% of the time.
The gap widens as the innings ends
Spin is cheaper than pace in every phase. What's odd is the shape: the advantage is largest at the death, where teams use spin least.
| Phase | Spin econ | Pace econ | Spin saves | Spin wkt % | Pace wkt % | Spin bowls |
|---|---|---|---|---|---|---|
| powerplay | 6.93 | 7.53 | −0.60 | 4.00 | 4.33 | 27.6% |
| middle | 7.27 | 7.99 | −0.72 | 4.79 | 4.86 | 55.2% |
| deathwidest | 8.63 | 9.45 | −0.82 | 8.50 | 8.38 | 15.7% |
Wicket % is bowler-credited dismissals per legal ball. Run-outs excluded.
“That's just Narine and Rashid”
The obvious objection: captains only trust elite spinners at the death, so of course they look good. That's a real effect — death spin is more concentrated. The top 20 spinners bowl 24.1% of death spin, against 13.7% for pace.
So we controlled for it — comparing only bowlers with 200+ death balls on both sides. Specialists against specialists:
The gap is 0.84 — essentially unchanged from the raw 0.82. It isn't only the elite spinners.
And it isn't one good season
Spin was cheaper at the death in 14 of 15 seasons. Usage never passed 19.8%.
Which kind of spin
Finger spin holds up best. Every spin type is cheaper at the death than either kind of pace.
What this does and doesn't show
- This is observational, not causal. It does not prove a team bowling more death spin would concede less. Captains choose when to bowl spin, and they may pick favourable moments — a new batter, a turning pitch — that we can't control for.
- What it does show is that the gap survives the specialist control and holds across 14 of 15 seasons. If death spin were purely elite-bowler selection, the like-for-like comparison should have closed it. It didn't.
- Around a fifth of deliveries are bowled by players with no published bowling style. Those are excluded rather than guessed.
- Scope: men's T20 with six-ball overs. This study covers the 1,870,356 deliveries across 9,781 matches where the bowler's type is known. Super overs excluded.
Citing this
Free to quote with a link to cricpickems.com/study/spin-at-the-death. The underlying splits are on the bowler and venue pages. Method notes on about.