Who are the greatest openers and opening pairs in Tests?

Test opening batters are a special breed. Opening the batting is arguably the toughest task in any form of cricket. Batting on a square turner with eight fielders near the crease is quite daunting, but facing a shiny new ball coming at you at speeds of around 150kph is something else. Especially if you batted in the first 100-odd years of Test cricket, with not much protective gear to speak of.

Opening batters in a Test match have one main task - to walk back together as a pair when the teams go in to lunch. Any batting captain would prefer a score of 75 without loss at lunch to 100 for 1 or 120 for 2. Well, a certain recent English captain excepted.

My analysis of Test opening batters in this article more or less follows the pattern set in the article on ODI opening batters, but deviates in certain areas since there is a clear need to differentiate between success and failure. Also, we need to account for the fact that until recently, Tests were played only in bilateral series. Dead rubbers were always a possibility towards the end of long series, meaning there were times when the intensity you associate with Test matches was missing from the format. Also, if we accept that in ODIs, average and scoring rate need to be given equal weight, in Tests, it is equally clear that we should give more weight to the batting average and less to the scoring rate.

An important caveat here. When I extensively analyse some aspects of the game, I cover long periods - nearly 150 years on the Test front and over 50 years in the ODI game. Conditions, laws, equipment, strategies, and attitudes change across these spans of time. It is easy to say that one cannot compare across various ages. Well, in that case, no comparisons are possible ever.

I have worked on the assumption that the basic requirements of the teams are and have always been the same - win Test matches by dismissing the other team twice and scoring one run more than them. To that end, the responsibilities of the Test opening batters remain the same today as they were in 1877: give your team a great start. Fifty without loss is a very good start: whether it was at The Oval in 1890, Madras in 1962, or the MCG in 2022.

Because we are dealing with 147 years of history, and availability of data for the first half of this period is sketchy, there might be some discrepancies. For instance, a batter might have retired injured and the No. 3 batter might have taken his place - without this fact being recorded on the scorecard. Please allow for minor variations that result from this kind of anomaly. I have corrected data for pairs to the extent possible.

We'll first look at opening batters and then at partnerships. The analysis is current upto the third Test between England and Sri Lanka earlier this month. The cutoff is 2500 runs scored while opening, and 63 batters qualify.

Alastair Cook leads the table of runs scored with a huge aggregate of nearly 12,000, ahead of Sunil Gavaskar by over 2000 runs. They are followed by Graeme Smith and two modern Australian greats - David Warner and Matthew Hayden. Two chalk-and-cheese batters, Virender Sehwag and Geoff Boycott, follow next. The top ten is a veritable who's who of great openers. Hayden and Mark Taylor did not score a single run outside the opening positions, a feat accomplished by only five other openers (with a 2500-run cutoff): Marcus Trescothick, Mike Slater, Bill Lawry, Conrad Hunte and Taufeeq Umar.

Onward to the twin pillars of batting - runs per innings and strike rate. Below is a composite graph with both measures side by side for batters who excelled on both counts. There is a fine distinction to be made on strike rates. In Tests, data for balls played in individual innings is available more consistently than it is for number of balls played at the time of wickets falling. So strike-rate data is more complete for batters than it is for partnerships. In the 2548 Tests considered, a total of 18,550 innings were played. Of these, there is balls-played information for 14,022 innings; this is just over 75%. This information plays a significant part while determining the Batting Index for individual batters.

Herbert Sutcliffe leads the Runs per Innings table with a huge value of nearly 55. It is not a surprise that Jack Hobbs follows in second place, a run and a half behind. Bob Simpson is in third position, at around 52 runs. Len Hutton is the other opening batter to have an RpI value above 50. The next place is occupied by a surprise batter - Virender Sehwag. Despite his penchant for quick scoring, he averaged over 48 runs per innings. Then come a mix of modern and previous generation top-drawer opening batters. The tenth placed batter is Rohit Sharma. The ten batters on the table have RpI values in excess of 47.

Sehwag tops the list of top ten quick scorers by a mile, having scored at almost exactly five runs per over. David Warner follows in second place with a scoring rate of over 4.2 runs. Sanath Jayasuriya hovers around the four-run mark. Then there are a few attacking openers around 3.6. Rohit and Sehwag are the only batters in both top tens. A surprise placement is Mohammad Hafeez, not necessarily known for his attacking batting.

In summary, the standout batters are Sutcliffe and Hobbs for their RpI values and Sehwag and Rohit for being on both tables.

Like for ODIs, I have developed a composite Batting Index that will let us look at the opening batters in the context of the twin values of RpI and strike rate. I have assigned the more important RpI a higher weight by tweaking the formula. The Batting Index is calculated with the formula "RpI + 7.5 * RpO". This ensures that the RpI values get primacy. The RpI component is capped at 55 and the RpO component at 37.5. The fact that there is more than 75% availability of balls data means that I can put in a slightly higher emphasis on strike rate for batters than for opening partnerships.

Sehwag leads, again, by a mile on the index table. There is no doubt that he is helped by a superb scoring rate but the high RpI value plays its part. Warner comes in in second place, some distance away. Hobbs is next, no doubt helped by the high RpI value of nearly 53. Hayden and Rohit are in the next two positions, for reasons similar to Sehwag's. They are followed by Graeme Smith and in turn by Simpson and Sutcliffe from further back, whose well-deserved places owe primarily to their high RpI values. Saeed Anwar and Justin Langer round off the collection of openers with index values of over 70.

Before moving on to partnerships, let me highlight some special performances by opening batters.

- Eight opening batters scored three hundreds in consecutive innings. They are Sutcliffe, Geoff Boycott, Gavaskar (twice), Greenidge, Graham Gooch, Haynes, Langer, and Warner (twice).

- Two opening batters, Sutcliffe and Gooch, scored four hundreds in five innings.

- One opening batter, Gavaskar, scored five hundreds in seven innings.

First, a few numbers on Test opening partnerships to provide some perspective.

- Total number of opening partnerships: 9275

- Average opening partnership: 36.1

- Average team innings score: 272

- Contribution to team total: 13.3%

I have set 50 runs as the mark for an opening partnership to be deemed a successful one. Not that 49 is not a successful one, but we need to set the bar somewhere. In general, a partnership that scores about these many runs will have taken the innings into the second hour of the innings, and between 15 and 20 overs would have been bowled on average. Unlike ODIs, I have not set an alternative "balls played" bar, since the data on balls played per partnership is available for fewer than 45% of Tests.

I have ranked any opening partnership with a dismissal below 20 partnership runs as a failure. The reason is simple. Fewer than ten overs have been bowled in most of these cases, and the match is still in its first hour. That is not enough for the team. Also, I have worked on the expectation that these tough cutoffs will define the opening partnerships whether they are facing totals of 300 or 100. Readers might argue that this pair of rigid cutoffs should be changed to reflect the era when any given match was played. However, I feel strongly that whatever changes there have been in the game, the fundamental expectations placed on opening batters have not changed. Based on these classifications, around a quarter of partnerships are successes and nearly half are failures.

- Successful partnerships: 2323 (25.0%)

- Failed partnerships: 4390 (47.3%)

The cutoff for inclusion is 1500 runs scored while opening as a pair, and 40 pairs qualify.

From here onwards, it is the rare table that Hobbs and Sutcliffe are not on top of. It is mostly a question of how far they are ahead of the rest. In the RpI stakes, they average over 85 runs per innings - this is not a misprint! They are nearly 25 runs ahead of the next placed pair - Neil McKenzie and Graeme Smith. Hobbs comes in again in third place, with another very good batter, Wilfred Rhodes. These two did very well, averaging nearly 60 runs per innings. Then come Simpson and Lawry, Gooch and Mike Atherton, Hutton and Cyril Washbrook, and John Wright and James Franklin, who complete the list of pairs who have exceeded 55 runs per innings. Smith appears in three different pairs. John Wright and Bruce Edgar bring up the rear with an RpI around 30.

Let us now move on to a metric, the strike rate, that is not necessarily important in Test cricket. There is a caveat. Information about balls played at the fall of a wicket is available for 4137 innings out of 9271, which is around 45%. For the rest of the Tests, balls played information is calculated pro rata, based on team data. Sehwag and Gambhir lead the table with a scoring rate of 4.63 in the 87 innings in which they opened the batting. Imagine scoring at this rate, not in the current decade, in which scoring at below four runs per over is frowned upon, but 20 years ago.

An interesting point is that only one of the top ten ten pairs here played their cricket before Test cricket's centenary year. Fredericks and Greenidge played together in the 1970s but managed to score at the very good rate of 3.74. That is, arguably, the most impressive placement on this table. Although it must be said that this is almost totally the scoring rate of the West Indian team, since balls played data is mostly not available for that period. John Edrich and Boycott, and the England of the 1970s at large, were not known for their quick scoring. It is no wonder that they are last on the list, with a low RpO.

The partnership index is calculated by the formula "0.75 * RpI + 5.0 * RpO". This time the values apply to the pairs and not individual batters. Also, the weight for strike rate has been reduced, since we do not have balls-faced data for the majority of matches. The capping is at 63 for RpI and 23 for strike rate.

Who else but Hobbs and Sutcliffe to lead the table, by a mile or so? They are nearly 15 points ahead of McKenzie and Smith, who are followed by those racers, Sehwag and Gambhir. interestingly, Smith appears in no fewer than three pair combinations again; that is some feat indeed. Wright and Edgar are at the bottom with an index of 34.7.

We have already seen that the average percentage of team runs scored by opening pairs is just over 13. However, Hobbs and Sutcliffe scored over 25% of their teams' runs while opening. That is nearly twice the all-time average. Then come Rhodes and Hobbs, with just over 22%. These two pairs are the only ones to exceed 20%. For the years either side of World War I, England really depended on their openers, mainly Hobbs, to score nearly a quarter of their team runs. By now, we are familiar with the Sri Lankan and New Zealand openers who are at the bottom - contributing around 12%.

We now come to the average quality of bowling (ABQI) faced by the opening pair of batters. For once, the two pairs featuring Hobbs are way down on the table. Boycott and Gooch are the top pair. This will not surprise anyone. They faced the great West Indian and Australian pace bowlers of the 1970s and '80s day in and day out. The ABQI is based on the Bowling Quality Index - which ranges from 10 (very poor bowling line-up) to 52.5 (a devastating bowing line-up). West Indies at the Gabba in 1979 were at 52.1 and the Indian team's BQI in Rajkot in 2024 was 52.0 (in both cases, four bowlers averaged under 25 each).

The 40-plus ABQI that Boycott and Gooch faced is frightening. Their courage is obvious and their technique undoubted. It is clear that England's batters of the 1970-1990 period faced the toughest bowling around. This is confirmed by the next two pairs on the table: Atherton-Alec Stewart and Gooch-Atherton. It must be pointed out that many of these English batters were quite reluctant to travel to the subcontinent. Atherton played eight Tests in the region, Boycott eight, Stewart ten, and Gooch 13.

Granted that the two pairs anchored by Hobbs faced ABQI of around 30: the South African attacks of the time were average, as were some of the West Indian bowling line-ups. But that has been the case through the history of Test cricket. Modern batters, say, from 1980 onwards, have faced weaker bowling line-ups through their careers.

Onwards to a very important aspect of opening batting: how successful a pair was in getting the team off to a good start. Hobbs and Sutcliffe succeeded in an amazing two-thirds of the innings they opened in. These are numbers to savour. In a typical Ashes series of five Tests and nine innings, this amazing pair crossed 50 six times. They had a ten-innings sequence of 182, 59, 156, 156, 58, 53, 172, 119, 155, and 85 in the 1926-28 period. That is an aggregate of 1039 runs in seven Tests, arguably the best stretch ever for a pair of batters. Trevor Goddard and Eddie Barlow, and McKenzie and Smith, succeeded around half the times they batted together. McKenzie-Smith have failed a high number of times too, as have Edrich-Boycott. Quite surprisingly, Dean Elgar and Aiden Markram are at the bottom of the table, having succeeded only once in five times. No way to explain this. And the almost makeshift Indian pair of M Vijay-Shikhar Dhawan had a success rate of around 22%.

As expected, Hobbs and Sutcliffe failed only once in six attempts when they batted together. Hutton and Washbrook and Trevor Goddard and Eddie Barlow failed in fewer than 30% of their innings together. Intriguingly, Wright failed about 21% of the time when he partnered Trevor Franklin, but when he partnered Edgar, they failed in half their innings.

Finally, a take on the hundred partnerships scored by pairs while opening. I am not a great fan of emphasising the 100th run, but a hundred is a significant milestone and worth calling attention to. Hobbs and Sutcliffe notched up a century partnership in around 40% of their innings; that is, they averaged three to four hundred stands in each five-Test series they played. Surprisingly, Rhodes and Hobbs are in second place, way behind. They are followed by Chris Rogers and Warner. At the other end of the table, we have Wright and Edgar scoring a single century partnership in the 56 innings they opened in.

Readers may ask why I have not developed a concept of quality of runs scored for Tests, as I have done for ODI openers. The reason is the difficulty in identifying situations clearly. Opening while facing a first-innings score of 100 may not exactly be a comfortable task, since the pitch might be a very difficult one. And facing up to a 400 total may not necessarily be all that difficult - the pitch might be easy to bat on. Also, a final-innings target of 200 might be a lot more difficult than one of 300. In addition, the time available is a significant factor and not enough data is available on that. The bottom line is that there are too many variables. So I developed other comparing points.

As I did last time, let me put down my thoughts on the best Test opener of all time. My first idea was to carry out a structured exercise based on many of these factors. However, it is clear that one name stands head and shoulders above the rest. To not anoint Hobbs as the best ever Test opener would take a lot of courage. Look at his overall numbers. Over 5000 runs at an RpI value of over 50. He was the common partner in two outstanding partnerships: the one with Sutcliffe yielded over 85 runs per innings and the one with Rhodes nearly 60 runs per innings. Maybe the only slight negative is that Hobbs faced bowling line-ups that were not all that great. But the reality is that a batter can only play the bowling attack that is in front of him - and score prodigiously, as in Hobbs case. The other contenders were Sehwag and Sutcliffe.

As far as pairs are concerned, Hobbs and Sutcliffe are so far ahead that they brook no competition. An RpI value that is off the charts - nearly 25 runs over the next best. Imagine how the other teams' bowlers would feel when they see a pair averaging 85 runs each time they bat together, succeeded two out of three times, failed only once in six attempts, and crossed the three-figure mark 15 times in 38 attempts, walk in. They lead on six of the nine tables. If that is not Bradmanesque, I am not sure what is. The closest contenders were Greenidge-Haynes.

The complete data files, including the complete data for all the partnerships played by the 40 selected pairs and all the innings played by the 63 batters, can be downloaded here.

The quirky stats section
In each article, I present a numerical/anecdotal outlier relating to Test and/or ODI cricket. This time the query is, "Which bowlers bowled more overs than they conceded runs and still took six wickets or more". The analyses have been normalised to six-ball overs. The scorecard spell is given in brackets, if different.

It is amazing how many such ultra-economical and ultra-productive spells there have been in Test cricket. No fewer than 13 instances in nearly 150 years. And, surprisingly, reasonably well distributed across the years. Arguably, the most impressive instances are the spells of Mackay and Gibbs, who bowled so many overs, and Lohmann, Lawson and Steyn, who had remarkable performances where they conceded fewer runs than they took wickets.

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.