International Cricket

ICC Rankings Controversy PAK vs BD May 2026: Ratings Points Decoded

Vikram Joshi 19 May 2026 Updated 19 May 2026 ~5 min read ~963 words

ICC rankings dashboard showing Pakistan and Bangladesh ODI ratings

Share this article

Pakistan beat Bangladesh 2-1 in the three-match ODI series at Karachi from May 8 to 13, 2026. The next morning, the updated ICC ODI rankings dropped Pakistan from 108 to 105 ratings points. Bangladesh, the series loser, rose from 89 to 100. Fan reaction in Karachi was immediate. By midday, "#ICCMath" was trending in Pakistan with 240,000 mentions. The ratings, however, did exactly what the published formula says they should. Here is the math.

How the ICC ODI rating formula works

The ICC ratings model is a weighted Elo. Each match has an expected-result calculator based on the gap between the two sides' pre-match ratings. The actual result is compared to expected, and the difference is multiplied by a match-weight factor (full ODI = 1.0) and a series-completion factor. Series wins add a bonus. Each team's new rating is the rolling 36-month average of all rated points, weighted with the most recent 12 months at 100% and the trailing 24 at 50%.

In the PAK-BD case, pre-series ratings were 108 and 89, a 19-point gap. Expected result for Pakistan was a 2.6-1 series win. Actual was 2-1. Pakistan under-performed expectation by 0.6. Bangladesh over-performed by 0.6. The series winner bonus for Pakistan was 1 ratings point. Bangladesh's expectation-beat earned 11 ratings points (because the gap to overcome was wider).

Why the math produced the 'wrong' feeling

The expectation gap matters more than the result. When a 108-rated side plays an 89-rated side, the rating system already prices in a 2.6-1 series win. A 2-1 result is, in the model's eyes, an under-performance by Pakistan even though they won the series. The same model would have raised Pakistan by 1 point on a 3-0 sweep and by 4 points on a 3-0 sweep with a dominant net run rate.

Bangladesh's 11-point jump has two components. First, the expectation-beat. Second, a rolling-window effect. Bangladesh's 2024 ODI numbers (largely poor) are about to drop out of the 36-month window in July. The May 2026 PAK series is therefore weighted more heavily within Bangladesh's near-window than it would be for a side with stronger 2024 data.

The two PCB letter requests

The PCB has not lodged a formal protest, but director of high performance Aaqib Javed sent a private letter to the ICC ratings desk on May 15. The letter, seen by two people familiar, asks two things. First, a published worked example for every series win that lowers a winner's rating, so fans can read the formula in plain prose. Second, an option for boards to receive a 24-hour pre-publication preview of rating changes, to coordinate communications.

The first request is reasonable and likely to be granted. The ICC ratings page has a four-line formula description that hides exactly the kind of expectation-gap arithmetic that produces this result. A worked example would defuse 80% of social-media outrage. The second request is harder. Pre-publication previews of rankings to boards would create a leakable two-tier disclosure regime. The ICC press team is resisting.

What this means for the Test rankings cycle

The Test rankings face a parallel sensitivity. England's recent home-Test wins over a depleted New Zealand have produced smaller ratings gains than the dressing room expected. Pat Cummins, Australia's captain, was asked at a Sydney press conference on May 11 whether the Test ratings "made any sense." His one-line answer ("I don't check them") led the next morning's back pages.

The Test cycle's 36-month rolling window is the same. The expectation-gap arithmetic is the same. The communications problem is the same. A worked-example publication policy for each ratings update would help. So would a one-page glossary on the ICC site. Both are being drafted internally.

What it means

Pakistan's 105 will reverse next series if they win 3-0 against a sub-90 side. Bangladesh's 100 will stick if they maintain 1-2 results against top-100 sides. The deeper problem is communicational, not mathematical. Expect a worked-example publication policy out of the June 18 Dubai meeting. Watch for the next series where a winner loses points. The cycle is monthly, and the math is unforgiving on near-equal sides.