Pushing boundaries

In 2014, Fraser Stewart, Cricket Academy Manager, MCC, asked whether I could do a comprehensive analysis of the boundary-hitting trends in Tests and ODIs. The MCC was extremely concerned with the way bats were becoming thicker and wider and contemplated making changes to the laws to bring some balance to the game. I provided a set of reports. The matter stayed there. The MCC must have looked at the numbers and the other related factors and decided that it was not yet time for the line to be crossed.

Then the 2015 World Cup 2015 was played. Those interested can read my review of the World Cup indicating how batsman-friendly the entire tournament was. And then the trend continued in Tests, especially down under. This April, Fraser again approached me with a request to redo the analysis and provide the recent and current trends. I ensured that the numbers for the 2015 World Cup and the years 2015 and 2016 for both ODIs and Tests were covered in detail.

The MCC considered the numbers provided, in conjunction with other equally important factors like scientific data, anecdotal data, bat sizes, ground sizes etc, and finally came out with a very important status paper. Their key recommendation is outlined below:

"One proposal would be for the maximum thickness of the edge to be between 35mm and 40mm, and the overall depth of the bat to be between 60mm and 65mm (some bats in current use have edges of 55mm and can be up to 80mm deep)".

Let me first explain the ground rules. The numbers on the tables in this article do not relate to all the Test matches during the period. These include only the matches for which data on fours and sixes is available. For the first table covering the 2000s, data is available for all matches. For the earlier decades the data availability varies considerably, going as low as 5% during the 1950s.

In view of the importance of the past 17 years, I have gone into this period in greater detail - by year. Let me first look at the percentage of runs scored in boundaries. This has shown remarkable similarity during these years. The range is between 50% and 55%, with only one year falling below 50%, to 49.8% during 2011. This might be an indirect effect of more singles, twos and threes being scored. There is clearly a reduction of dot balls.

Now we come to the fours frequency. There is a clear change as years went by. The millennium started with a four being scored every 19.8 balls and that frequency has now reached a four every 14.5 balls: a significant change of 27%. Just visualise this: a four almost every two overs. Afterwards, the teams need only a single every over to reach the acceptable runs per over of 3.0.

In the graph I have presented an additional column, which is a summary of the 122 years between 1877 and 1999. For this period the fours frequency was 21.1, not too different to the later figures. Other than the current year, the highest frequency of fours (low frequency values) occurred during 2005. The frequency of sixes has had more topsy-turvy movements during these 17 years. The year 2000 saw a six being hit every 417 balls, the least frequent during this period. The year 2014 saw a six every 206 balls; that is more than two sixes per day or nine to 11 sixes in each Test.

The graph is important because of the first figure, the 19xx one. That value has gone through the roof. The actual value is a six every 614 balls.

Just to get an idea of the way the numbers have shaped up over the past 100 years, not just the past 17 years, I have provided a summary by decade. Unfortunately not much data is available for six decades starting from 1910. Anyhow, virtually no data is available for the first 30 odd years. The best is the period just before World War I. We have data for just over a third of the Tests played. The worst is during the 1950s in which we have data for only one out of 20 Tests. Maybe because new teams came in and record-keeping was indifferent.

So these data are so far out that it is better to ignore those. Four sixes in the 1950s in eight Tests does not tell much. Where data was available we see that a six was hit every few thousand balls. Surprisingly the fours frequency has been quite steady through these 50 years: Around one every four overs or so.

For the 1970s, we have data for about two-thirds of the matches. During these matches a four was scored every 22 balls. That is pretty good. However the sixes frequency was one every 135 overs or so. There was a significant change during the 1980s: to a four every 20 balls and a six every 565 balls. The data availability improved dramatically.

During the 1990s, the frequency of hitting boundaries improved, albeit, by around 5% only. This trend continued during the first two decades of the current millennium. While there was only minimal increase in the fours frequency, the sixes frequency improved a lot more dramatically.

Now we come to a couple of support graphs for Tests. In these I have computed the frequency of fours for the batting teams, bowling teams and grounds. There are no tables for this analysis. The graph would suffice. Since I wanted the analysis to be fair across teams, I have done this only for the period 2000-2016 during which 100% of the data is available for all the teams. If I include the previous years/decades, we would be looking at data covering widely varying percentages of Tests played for different teams.

Australia have the best four-hitting frequency, with a four every 14.8 balls. India are close behind with a frequency of 15.3. This is on expected lines. Then we have no less than seven teams in the narrow band of 16-17 balls per four. Zimbabwe are way below these numbers, with a frequency of 19.5. Amongst the bowling sides, a number of teams are around the 15-16 mark, with New Zealand the worst, in terms of conceding fours quite frequently. The Sri Lankan bowlers are the most difficult to hit fours. West Indies follow close behind.

The four hitting is most frequent in England and South Africa. It is tough to hit fours in the West Indies. However the most difficult place to hit fours is the UAE.

When it comes to the frequency of sixes, New Zealand are the king. Their batsmen do it once every 213 balls. That seems quite easy when the concerned team has a certain Brendon McCullum. Australia are close behind. Pakistan and West Indies follow next. Not surprisingly, England and surprisingly South Africa are around the middle with around 350 balls per six. Sri Lanka score a six only once in 60 overs.

The Bangladesh bowlers have a six hit off their bowling once every 220 balls. New Zealand are also generous when it comes to conceding sixes. They concede one every 226 balls. Australia, England, West Indies (surprise) and Pakistan conceded a six every 300-plus balls, with the England bowlers being the most parsimonious.

The sixes are most frequent on New Zealand and Bangladesh grounds. Sri Lanka are quite parsimonious when it comes to sixes on their grounds: one every 55 overs. In England, it is still more difficult, with a six every 70 overs.

I have given below the boundary-hitting frequency values in Test matches for a few top attacking batsmen. The table is ordered on the frequency of boundaries.

The table is self-explanatory. It is not a surprise that Virender Sehwag leads this table, closely followed by Adam Gilchrist and David Warner.

Moving on to the ODI matches, a much shorter period of history but replete with many changes in these 45 years. The first table is the one by period.

The percentage of runs in boundaries has registered a significant increase from a third to nearly half across the 45 years. We are yet to reach the 50% mark.

As it happened in Tests, the fours frequency has not changed much: from 20 balls per four to 13.4 during the past few years. This is possibly because a four represents the essence of good batting and all the batsmen are capable of playing strokes resulting in fours. Similarly the number of fours per match moved by about 30%: from 28 to 39 during all these years.

Sixes are a different thing altogether. Not all batsmen are adept at hitting the ball over the ropes. This distinction is clearly seen when we peruse the changes across the years. The frequency of sixes has increased dramatically from one every 33 overs to one every 12 overs. Consequently, while an average of 2.7 sixes were hit in an ODI match during the initial years, that figure is hovering at over seven now, despite the fact that the average number of overs per match is now 45 and during the early years it was around 53.

The frequency of boundaries has changed from one every 18 balls to one every 11 balls. The number of boundaries has increased by about 50%, from 30 to 46.

The World Cup is an important reference point to analyse ODI scoring patterns. Hence I have done an analysis by World Cups.

One every 57 overs in the 1975 World Cup to one every nine overs during the 2015 World Cup. The changes in four-hitting numbers are not that drastic.

As I have done for Tests, I have done a more detailed analysis for the current millennium. The fours frequency has not even changed by 20%. However the six frequency has changed by over 60%.

The number of fours per match has increased from 34 to 42 and now back to 38. That seems to have plateaued. For the 39 years leading to the millennium, this figure was 29.2. So no great change.

There has been a dramatic increase in sixes per match: from 4.1 to 8.8, over two times. The previous 39 years, 3.3 sixes were hit per match. So the real move up is later.

Below are the boundary-hitting frequency values in ODI matches for a few top attacking batsmen. The table is ordered on the frequency of boundaries.

The table is self-explanatory. It's not surprising that Sehwag leads this table also, closely followed by Shahid Afridi and Gilchrist. Let us tip our hat for Paul Stirling of Ireland, who occupies the fourth position. It is indeed a very creditable achievement.

To view/download the MCC document on "State of the game", please click here

Masterpieces by Mendis and Chase
I am glad to inform the readers that the Kusal Mendis classic has vaulted into the top-20 in the Test Innings Ratings. This is a completely revised ratings exercise and, as I have mentioned in my response to Jasprit's comment in the last article, will be revealed only in my book. I will not say anything other than that the "Top Test innings" chapter has to be revised considerably. A truly great innings, and totally unexpected.

As for the extraordinary last day at Sabina Park and the innings of Roston Chase, even the most die-hard supporter of India has to agree that this was great for world cricket. Roston Chase: a name that will be talked about for years to come. Four West Indian batsmen, all in their early-20s, did the impossible, scoring 340 for 2 wickets on a last-day pitch, starting at 48 for 4. Chase's innings was not ranked that high but inarguably as important as that of Mendis'. An important point was that the rating points were shared by the four batsmen. I could not stay up the whole night, as I did in 1999, when I (and everyone else, for that matter) was 17 years younger, but watched four hours of play. I was certain at that time that West Indies would hold on. So assured was their approach.