To start with, there are some ground rules. This will focus on the top run-scorers in Test cricket batting from positions five to seven, examining their performances based on the score of their team when they came out to bat.

The average partnerships for the first five wickets across Test cricket history have been 37, 39, 42, 42 and 37 runs respectively, i.e. roughly in the range of 30-40 runs per wicket (RpW). So for the purposes of this piece, a team will be deemed in "trouble" (or in a "crisis") when they have lost their wickets for cheaper than 30 runs apiece.

Following on from this definition, we can then examine the performances of individual batters when they enter the crease with their side in trouble, versus in better situations. There are limitations to this approach. A batter could come out at a reasonably healthy 120-3, only to see two wickets fall quickly. This would be counted as having come in in a better situation, even though really what they need to do is bail their team out of trouble.

This method will also count a batter who comes in with his team at 80-3 in the third innings, with his team looking for quick runs with a big lead, as being 'in trouble', and would not consider the batter who comes in on a decent platform but still facing a big deficit. It's impossible to come up with a perfect statistical definition of 'in trouble' or not, so ours is a useful approximation.

Also read: Bangladesh's Test spin attack is the strongest to tour India in the 21st century

A question that might arise here is why No.4 batters have been excluded. This is a subjective decision, made with the following logic: At a runs per wicket of 10, for example, a No.4 batter comes in at 20-2, while a No.7 comes in at 50-5, which is arguably far worse but would be clubbed with the 20-2 situation.

This effect exists to an extent even when taking Nos.5-7 together, but is somewhat mitigated. Again, it is a subjective call to leave out No.4s. In addition, No.8s and below have not been considered as the vast majority of them are bowlers or at best bowling all-rounders, rarely batters or keeper-batters.

Here is an overall look at the records of players with 3000 or more Test runs from the No.5 to No.7 positions:

What do the numbers tell us about middle-order batters in Test cricket?

The immediate observation here is that most players perform better when coming in at higher scores. This is quite natural. If they are coming in with runs on the board, there is less pressure and it is likely that either one or both of the following is/are true: conditions have been at least half-decent for batting, and the opposition's bowling attack is not very incisive.

In addition, the top run-scorers and those with high career averages, also largely perform well when it comes to batting in difficult situations. Again, expected, simply because better players overall are likely to be better players in almost any given scenario as well.

For the second reason, as far as identifying good crisis batters goes, players like Michael Clarke, Shivnarine Chanderpaul, Garfield Sobers and other greats are right up there. Interestingly, Steve Waugh and Allan Border, rated as perhaps Australia's two toughest battlers, averaged a reasonable amount less when coming in earlier.

Read more: It's about time, Virat: Don't let the other three get away

But apart from this, we can also isolate those middle-order batters who were at their best in difficult situations, compared to more favourable ones.

Ten players on our list average at least five points higher when entering at worse than 30 RpW, compared to better than 30 RpW. These are: Andy Flower, Alan Knott, Brendon McCullum, Graham Thorpe, Ian Botham, Dhananjaya de Silva, Jonny Bairstow, Arjuna Ranatunga, Clive Lloyd and Tony Greig.

Not all of these players are bonafide great batters. In fact, only three of them averaged 45 or more in the middle order; Lloyd, Thorpe and Flower. Most of the others are players with relatively unremarkable career records, roughly averaging between 35 and 40.

A fascinating case study is Brendon McCullum's record. He averaged over ten runs per dismissal more in "crisis" situations than other times (45.2 to 33.9). In the space of one home series against India in 2014, he pulled out a couple of breathtaking knocks – scoring 224 from 30-3 in Auckland and 302 from 52-3 in Wellington.

Ten years prior to that, McCullum had managed 143 against Bangladesh in Dhaka, coming in at 139-5. Against Australia in his final Test, he hit a spectacular, counter-attacking, 145 off 79 balls after New Zealand had been reduced to 32-3.

Six of his ten centuries in these batting positions came from entry points of under 30 RpW, including five of his top six highest scores. The other one, 185 against Bangladesh in 2010, came after New Zealand were 158-5 (RpW of 31.6).

Where McCullum differs from the likes of Thorpe, Lloyd and Flower is that he was not as prolific when situations were more in his favour. This is a similar trend seen in the records of Ian Botham, Alan Knott, Arjuna Ranatunga and Jonny Bairstow. Tony Greig and Dhananjaya de Silva lie somewhere in between these two categories.

These sorts of records raise a couple of interesting questions as to the reasons behind them. As always, they will vary on a case-by-case basis, but the batter's own style of play could be a major part.

Middle-order batters more comfortable against pace bowling might be able to negotiate them and then thrive when they are forced in early, whereas coming in later may make them more vulnerable to spinners in operation at that point.

For a moment, let us focus on Dhananjaya de Silva, the most recent entrant to the 3000-run club in these batting positions. Along with Andy Flower, he is one of just two players on our long list of 52 who has had to begin 60 per cent of their innings from a team RpW of 30 or worse:

For the longest time, Flower was Zimbabwe's lone warrior, and often their only batter of genuine Test quality. While he averaged 51 in the format, his fellow top seven batters across his career managed a measly 27. Over two-thirds of his innings came when Zimbabwe were under 30 RpW at his entry.

It's possible he learnt over time to hone the skill of riding out tough periods, having been put in them so often. His career-best score came in one of these situations – in 2000, from 61-3 against India in Nagpur, Flower smashed an unbeaten 232 with Zimbabwe following on, to help his side secure a draw.

So far, de Silva's Test career for Sri Lanka has followed a similar strain. On Sri Lanka's recently-concluded tour of England, his two best knocks of 74 and 69 came after his side hat slipped to 40-4 and 91-4. In Bangladesh earlier this year, he scored 102 & 108 in the same Test match after Sri Lanka had been reduced to 47-4 & 67-4.

Ten of de Silva's 12 Test centuries have come in batting positions 5-7. Eight of those have been from entry points under 30 RpW. For the combination of volume of runs, and outstanding performances in difficult situations, Sri Lanka's captain is one of the best around today.

But apart from just responding to the question we have asked, there is another benefit to this line of analysis. Batters with good records in difficult situations may not just possess the mental fortitude required to deal with them. There may be something in their technique or style of play that lends itself to conditions in those periods, meaning there are avenues to explore with regards to moving them up the order as well.

As an example, Australia's Cameron Green started out batting primarily at No.6 and No.7 in Test cricket, and averaged 42 at entry points worse than 30 RpW, compared to 29 at better ones.

His move to No.4 earlier this year paid off almost instantly for Australia, with a stunning 174 not out against New Zealand in just his fourth outing. It's too early to say that this phenomenon is any sort of predictor of success up the order, but it may be an indicator that a batter is suited to coming in early.

Among currently active players, there are a few who come out looking quite solid when splitting their records in the manner employed here. de Silva is one, Pakistan's Mohammad Rizwan (1769 runs) is another – averaging 54.2 at entry points under 30 RpW compared to 33.2 above. Travis Head (2914 runs) and Litton Das (2320 runs) are also marginally better when entering at lower scores.

So...who is Test cricket's greatest crisis batter?

But now, back to the central question. Who is Test cricket's greatest crisis batter? Like most cricketing questions, it depends quite a lot on who is asking.

If you prefer those who excelled with the bat, crisis or not, Andy Flower probably comes out on top, with Michael Clarke, Shivnarine Chanderpaul, AB de Villiers and Garry Sobers also in the mix. Graham Thorpe and Clive Lloyd are also in the conversation, from earlier in the piece.

If you prefer more of a focus on crisis "specialists", Brendon McCullum is among the top picks, perhaps alongside Ian Botham and Jonny Bairstow. And if you want to give some weight to how often they performed rescue acts, Dhananjaya de Silva comes into the frame, as well as maybe Tony Greig and Dinesh Chandimal.

Or maybe according to you, it is Kurtis Patterson for no other reason than the fact that he averages 144 in Test cricket. After a point, the answer is immaterial. The process is more important, and this is a step towards attempting a more objective look at the question.

Follow Wisden for all cricket updates, including live scores, match stats, quizzes and more. Stay up to date with the latest cricket news, player updates, team standings, match highlights, video analysis and live match odds.