Stats Analysis

Which recent Test upset was the most emphatic?

A lowdown on all the major upsets in the last three years

Anantha Narayanan
Where would you rank New Zealand's series-clinching win in Pune in 2024 in the list of greatest Test upsets?  BCCI

Why do we say that a Test has resulted in an upset win for a team? Mostly it is the gut feel of the informed cricket followers. When one has followed cricket long enough, the cricket follower/writer knows intuitively which is an upset. Almost everyone agreed that India were the overwhelming favourites against New Zealand last year at home and when the results turned 180 degrees, everyone concurred that those results were some of the greatest upsets of all time.

Loading ...

But this will not answer the questions: How great was New Zealand's Bengaluru upset? Was it greater than the one by Bangladesh in Mount Maunganui or by West Indies in Brisbane? Was Australia's 2017 win in Pune a greater upset than the Bengaluru one? What about Bangladesh's wins in Pakistan? And so on.

So I set about quantifying the matches with a view to determine what were really the greatest upsets. You could say that I have tried to translate the learned writer/follower's gut feelings and intuitions into numbers. Of course, this has been done using my own collection of derived numbers. So while you can call this an anecdotal article, it's strongly supported by numbers.

I did not want to go beyond the last three years. In any case, I did an article in 2021 that covered the major upsets across the 140-plus years of Test cricket.

In this analysis of recent Tests from 2022 onwards, I have been able to include two other factors of relevance - the availability of key players and the series status. This analysis covers the last 138 Tests. It is a two-stage process: a preliminary selection based on a couple of quantifiable parameters and a detailed final analysis using multiple measures.

The preliminary selection was simple. From the matches in which the weaker teams (based on my Team Strength Index) defeated the stronger teams, a total of 43 Tests qualified. Then I did a detailed analysis of these 43 matches using quite a few other factors and selected 11 Tests for featuring in this article. The factors I used for analysis are described below. The points total up to 100.

Bangladesh had two emphatic wins in Pakistan last year, by ten wickets and six wickets  AFP/Getty Images

1. Team Strength Differential (max 50 points)

This is the difference between the losing (stronger) team's Team-Strength-Index (TSI) and the winning (weaker) team's TSI. Since the preliminary process selected only the Tests in which the weaker teams won, this number is always positive for the selected Tests. Working with the location-based values ensures that the true strength differential is derived.

For instance, the home TSI for Australia in Brisbane was 85.5 and the away TSI for West Indies was 37.8. Thus, the difference of 47.7 is a very positive indication of the huge problems faced by West Indies. In another example, the home TSI for India, in Bengaluru, was 96.1 (the highest ever) and the away TSI for New Zealand was 63.5. The difference is a high 32.6.

2. Winning team's performance in the last ten Tests (high points for poor performances - max 10 points)

What form the teams carry into the Test is very relevant. For a team on an indifferent run leading to the Test to beat a team on a roll is indeed very creditable. I use the location-based - 3.0 (away win), 2.5 (home win), 1.25 (away draw), 1.0 (home draw), 0.0 (loss) - count to determine this factor. The highest value for this factor is 29.0 for West Indies in Melbourne in 1984 into which they carried a sequence of eight away wins and two home wins. The maximum points possible is 30.0.

If the weaker winning team carries a very poor run into this match, they should get a lot of credit and vice versa. At the same time I want to limit the points to 10.0. The formula is given below.

Index points = 10.0*(30.0 - Last_10_points for weaker team)/30.0.

When they played the Mount Maunganui Test in 2022, Bangladesh were in poor form - (6.75 points out of 30.0) so they got 7.8 points (10.0*23.25/30.0) since their win was despite this poor form.

West Indies' win in Brisbane in January 2024 was their first Test win in Australia in 27 years  Getty Images

3. Losing team's performance in the last ten Tests (max 10 points)

For the other (stronger) team, the reverse works. If the stronger losing team carries a very strong run into this match, the winning team should get a lot of credit and vice versa. The formula is given below.

Index points = 10.0*(Last_10_points for stronger team)/30.0.

In the Mount Maunganui Test, New Zealand were in very good form - (21.0 points out of 30.0), so Bangladesh got 7.0 points (10.0*21.0/30.0) since their win was despite the stronger team's excellent form.

It might be confusing, but these measures will become clearer as we go into the narratives of the matches.

4. Margin of win in the match (max 10 points)

This is to distinguish between West Indies, who defeated Australia by eight runs, and Bangladesh, who defeated Pakistan by ten wickets. The win by the weaker team gains more gloss if it is a convincing win rather than a scrape-through. Ten points are allotted for this parameter. The formula is given below.

Index points = 10.0*TPP for the winning team/100.0, where TPP is Team Performance Points.

Readers will be aware that the winning teams get just above 50 points for very close matches, 60-plus for comfortable wins, and 75-plus for innings/big-run wins. The losing teams receive the remaining points out of 100.0.

5. Series status after the win (max 10 points)

In a two-Test series, a win in the first Test (as in Mount Maunganui) ensures at least a drawn series for the weaker winning team. A win in the second Test enabled the team to draw (Brisbane) or win (Rawalpindi) the series. The teams are assigned points suitably.

In a three-Test series, a win in the first Test (Bengaluru) was an important win but did not guarantee anything. On the other hand, because the first Test was won, a win in the second Test (Pune) guaranteed a series win. A win in the third Test after the series is already decided (Wankhede/The Oval) becomes less relevant. Again, the teams are assigned suitable points.

In series longer than three Tests, early wins do not guarantee anything, but they keep the series alive. We have never come across an upset in a deciding Test in this period.

Left-arm spinner Tom Hartley took 9 for 193 on Test debut to hand England a famous win in Hyderabad last year  BCCI

6. Key player availability in concerned Test (max 10 points)

This relates to the non-availability of a very important/important player for the concerned Test. The emphatic example is Kane Williamson's non-availability for New Zealand in India. The narratives contain suitable explanations.

A total of 11 Tests are featured chronologically. The key numbers are shown first, followed by a brief narrative. The numbers 1/2 indicate that this is the first Test in a two-Test series; 2/2 indicate that this is the second Test in a two-Test series, and so on. At the end, there is a table that shows the Tests ordered on total points.

In Mount Maunganui in 2022, Bangladesh, with an away TSI of 43.3, with a very poor recent form of 6.75, faced a New Zealand team with a high TSI of 80.6 and a very good recent form of 21.0. If this Test was played ten times over, New Zealand should have won nine times. Well, the tenth occurrence is what actually. Bangladesh won, and that too very comfortably by eight wickets. Since this was a two-Test series, Bangladesh assured themselves of at least a drawn series. The net impact of all these factors was a Total Upset Index of 64.9.

The second Test featured is an unusual one in that it features two strong teams. When South Africa played New Zealand in Christchurch in 2022, they were very much the underdogs. Their away batting was in shambles with averages around 20 strewn about. Their away TSI was only 56.7 as compared to the local team's 79. In terms of recent form, both teams were comparable. And they achieved a series draw with this huge 198-run win, which helped move the Upset Index to 46.

The next match featured is also between two top teams. When Australia played India in Indore in 2023, they had already slid to a 0-2 deficit in the four-Test series. Their inexperience meant that they were way behind in the TSI values. Both teams were in reasonable form. But their nine-wicket win was emphatic in this low-scoring game, and they were missing their captain, Pat Cummins. This gave them additional points and their Upset Index was a decent 53.

One more selection that might surprise a few. Pakistan were at decent strength when they played a very strong Sri Lanka in Galle in 2023. Their recent form was awful - one win in ten Tests. They achieved a tough win in a rather high-scoring match and assured themselves of a series share. That they went on to win the next Test by an innings was the icing on the cake. They received an Upset Index score of 36.

The next featured match is a classic. A totally down-in-the-dumps West Indies meeting the mighty Australia at the impenetrable fortress of Brisbane, conceding a TSI deficit of 48, in awful form, against a team in very good form, and trailing the two-match series 0-1. Prior to the match, only the margin of win for Australia was being discussed. But instead West Indies triumphed by eight runs through an inspired spell by the injured Shamar Joseph. They dismissed the match winner, Travis Head, for two golden ducks. Kavem Hodge, playing his first series, batted well in both innings. The Upset index was a very impressive 73.

We are continuing the theme on the surprise results among top teams. When England toured India in 2024, no one expected the visitors would fashion a win out of nowhere through a spinner making his debut. But that's what happened in Hyderabad. Chasing a reasonable target of 231, India were dismissed for 202 through some great bowling by left-arm spinner Tom Hartley. This was after Ollie Pope played one of the greatest Test innings ever. The numbers are supportive of calling this an upset. A decent TSI difference was there. The forms of the two teams were comparable. The Upset Index was 35. That England managed to lose the Test series 1-4 takes the sheen off this win but that is outside the scope of this analysis.

Despite the upset-halo surrounding the Bangladesh win against Pakistan by a big margin of 2-0, the wins are not that great, based on numbers. The series win was huge, more than the sum of the individual wins. As such, I am featuring only the first Rawalpindi Test. The TSIs differed by less than 10% and both teams were coming off indifferent recent form sequences. Pakistan's numbers, even at home, were not that good. The win was by a big margin and the assured saving of the series gave this match an Upset Index of 28. I am featuring this match only to highlight the series win.

India were expected to comfortaly win the Bengaluru Test against New Zealand. A decent New Zealand, bereft of Williamson, their best batter by a mile, were up against India's batting might and spin excellence. Despite having an away TSI of 63, New Zealand found themselves 33 short in this regard: the Indian home TSI of 96 is the highest ever - through batting averages of 69, 50, 53, and 56, backed by bowling averages of 20, 21, 22 and 15.

Everyone knows what happened. Overconfident to a fault, India opted to bat first on a green pitch, scored 46, conceded 402, recovered well, but slipped from 408 for 3 to 462 all out, and lost the match by eight wickets. India were in much better form and this resulted in an Upset Index of nearly 63. A well-justified number indeed.

"The first Test loss was a fluke. We will now have a square-turner and turn the series around for a 2-1 win." This was the host team's sentiment when they played New Zealand in the second Test in Pune. The Tamil Nadu offspinning duo, Washington Sundar and R Ashwin, ran through New Zealand for 259 and talks were on for a 350-400 score. Except, Mitchell Santner happened and India conceded a lead of over 100 runs. New Zealand repeated their first-innings score and the target of over 350 meant that Santner wove his magic again and India lost the Test, series, and a place in the WTC final.

The TSI difference was still big (but less than Bengaluru), the recent form went a little nearer. So, despite the big win and the series victory, the Upset Index was slightly down to 59.

In the third "dead rubber", India were missing Jasprit Bumrah and New Zealand played without Santner (in addition to Williamson). The Upset Index dropped slightly but I decided not to feature the match since there are other matches jostling for places. The Upset Index for the third Test was at a decent level of 36 points.

Down in the dumps after the 0-3 hammering by New Zealand, India reached Australia in November 2024 with minimal expectations. And the first Test was in Australia's fortress, Perth. Given these factors, very few expected India to do well. And to top it all, Rohit Sharma and Shubman Gill were not available. Australia were way ahead in TSI comparisons, India's three home defeats had pulled their recent form down a little. But they won big and earned an Upset Index value of 51 points. That they went on to lose three of the next four Tests has little bearing on this analysis.

Finally a match from this year. Pakistan had beaten West Indies in the first Test in Multan comfortably. They made the same mistake India did, creating pitches that turned square from day one. And then the West Indian spinners out-bowled their Pakistani counterparts in the second Multan Test. Jomel Warrican had a memorable match, taking nine wickets and scoring 54 invaluable runs - 36 of these from No.11 in the first innings. West Indies won by 120 runs and squared the series. They were way behind in the TSI comparisons but both teams were going through rough patches. The comfortable win meant that the Upset Index was decent - around 54.

Based on the nuanced compilation of all the relevant facts, this table indicates that the greatest upset was West Indies' eight-run win over Australia in Brisbane. The shock waves this win produced were seismic. For a period of time, this result promised to scuttle Australia's WTC chances. I am quite sure that there will be very few discordant notes on this selection. Some distance behind is Bangladesh's defeat of New Zealand in Mount Maunganui. Completely unfancied, a much weaker side, spearheaded by fast bowler Ebadot Hossain, fashioned this amazing win. In third place is New Zealand's win in Bengaluru, the contrast between 46 and 402 making it fantasy land. Their Pune win follows next, since that was the series-winning Test and proved that the first Test win was not an aberration. The top five is rounded off by the recent most unexpected Multan miracle performed by the down-in-the-dumps West Indies over Pakistan.

Since this is an article featuring upsets in Test cricket, I have to mention the series results as well. Without going into too much detail, I would like to say that the greatest upset in a series in nearly 150 years of Test cricket is the 3-0 demolition of India by New Zealand in 2024. There have been upset series wins previously - India over West Indies and England in 1971, New Zealand at home over West Indies in 1980, Zimbabwe in Pakistan in 1998, Sri Lanka in England in 2014, Sri Lanka in South Africa in 2019, India in Australia in 2020-21, and so on. However, never before has a 3-0 expectation been converted to an emphatic 0-3 result. The only other series that could even be thought of in similar terms is Sri Lanka's 2-0 win over South Africa in 2019. It looks like there is enough stuff on this theme for a future article.

The quirky stats section

Two lists of teams that lost unexpectedly and those who won unexpectedly - presented in chronological order. The other teams' first-innings scores are also provided.

Scoring 500, and losing (In the first two innings of the match).

Scoring below 100, and winning (In the first two innings of the match).

Talking Cricket Group
Any reader who wishes to join my general-purpose cricket-ideas-exchange group of this name can email me a request for inclusion, providing their name, place of residence, and what they do.

Email me your comments and I will respond. This email id is to be used only for sending in comments. Please note that readers whose emails are derogatory to the author or any player will be permanently blocked from sending in any feedback in future.

Anantha Narayanan has written for ESPNcricinfo and CastrolCricket and worked with a number of companies on their cricket performance ratings-related systems