July 31, 2015

The bat-first advantage in Tests

Since 2009, teams batting first have a win-loss ratio which is 61% higher than it was between 2000 and 2008

Since 2009, teams batting first have a 25-11 win-loss record in Tests in England; between 2000 and 2008, their record was 23-25 © Getty Images

The Edgbaston Test has been a refreshing change to the template of teams batting first racking up big scores and then dictating the Test match. That template was most prominently on display at Lord's a couple of weeks ago, when Australia amassed 566 after winning the toss and batting first, and never allowed England back in the game. That levelled the series at 1-1, after England had themselves utilised the first-innings advantage in Cardiff to take the lead in the series. The two previous Tests in England this summer - against New Zealand - were also won by the team batting first, though in both those games the third innings of the Test was the decisive one, not the first. When Australia bulldozed England 5-0 at home in 2013-14, they batted first in four out of five Tests, though only once did they top 400. In the last few years, there's a feeling that there isn't as much juice in pitches at the start of Tests as there used to be, which is giving an advantage to teams batting first. Do the stats corroborate that view? Here's what the numbers say:

Since the beginning of 2010, teams batting first have won 91 Tests and lost 80, which is a win-loss ratio of 1.14; in the 2000s it was 0.83 (159 wins, 191 defeats), which means batting first has certainly become a more profitable option in this decade, by about 37%. However, there isn't any clear pattern in the earlier decades; quite the opposite, in fact - the wins and losses for teams batting first were just about even in the 1990s, much worse in the 1980s, and much better in the 1970s. Overall, during this period, the ratio almost even - 0.96.

Since 2000, though, the difference is the most stark when breaking up the entire period into two parts - 2000 to 2008, and 2009 onwards. In the first period, the win-loss ratio for teams batting first is only 0.77 (141 wins, 183 losses); since then, the tables have turned completely, with the ratio going up to 1.24, a whopping increase of 61%.

Decade-wise stats for teams batting first in Tests since 1970
Decade Mat Won Lost Draw W/L
 1970s  198  67  47  84  1.42
 1980s  266  60  83  122  0.72
 1990s  347  113  110  124  1.03
 2000s  464  159  191  114  0.83
 2010s  228  91  80  57  1.14
Stats for teams batting first in Tests since 2000
Period Mat Won Lost Draw W/L
 2000-2008  423  141  183  99  0.77
 2009-2015  269  109  88  72  1.24

Those who've been complaining about England's pitches losing their tendency to seam and swing at the start of Tests in recent years can point to the table below for some sort of validation of their theory - since the beginning of 2009, teams batting first have a 25-11 win-loss record in England; between 2000 and 2008, their record was only 23-25. The change in win percentage for them is very nearly 150%. England isn't the only country where such a change has happened: in New Zealand the win ratio for teams batting first has gone from 0.63 to 2.50; in Australia it's changed from 0.84 to 1.70; even in South Africa there has been an increase, from 0.68 to 1.09. In the period between 2000 and 2008, only in the West Indies did teams batting first win more Tests than they lost (excluding the UAE, which hosted only four Tests); since 2009, this has increased to six countries which have hosted more than 20 Tests, plus Zimbabwe.

The other set of numbers which are surprising are the win-loss ratios in India. You'd expect teams batting first to do well here, but since 2009 they have a miserable 5-15 win-loss record. A closer look at those results reveals that visiting teams were often the ones batting first in those matches, and India were good enough to win in their home conditions despite not batting first. In 25 Tests in India during this period, visiting teams have batted first 17 times, with 13 defeats, three draws, and only one win - by South Africa in Nagpur in 2010. India, on the other hand, have won four and lost two when batting first, with the other two Tests being draws.

England, on the other hand, are 19-3 when batting first at home since 2009, but only 7-5 when fielding first; overseas teams are 6-8 when batting first, and 4-20 when fielding first.

Win-loss stats for teams batting first, by host country
  2000 to 2008 2009 onwards
Host country Tests W/ L Ratio Tests W/ L Ratio
Zimbabwe 22 6/ 12 0.50 8 6/ 2 3.00
New Zealand 36 10/ 16 0.63 25 10/ 4 2.50
England 63 23/ 25 0.92 46 25/ 11 2.27
Australia 54 21/ 25 0.84 33 17/ 10 1.70
Bangladesh 28 9/ 15 0.60 22 8/ 7 1.14
Sri Lanka 49 17/ 22 0.77 30 10/ 9 1.11
South Africa 49 17/ 25 0.68 28 12/ 11 1.09
West Indies 44 15/ 13 1.15 32 11/ 12 0.92
UAE 4 3/ 1 3.00 18 5/ 7 0.71
India 44 11/ 16 0.69 25 5/ 15 0.33
Pakistan 30 9/ 13 0.69 2 0/ 2 0.00

Combining the stats for the four countries where pace, seam and swing have traditionally played the most prominent role - Australia, England, South Africa and New Zealand - it's clear that batting first has paid more dividends here too, over the last six-and-a-half years. The win-loss ratio for teams batting first has gone up from 0.78 (71-91) to 1.78 (64-36), an increase of almost 130%. Winning the toss and batting first has produced similar numbers, but the table below also shows that captains have recently got it very wrong when they've opted to field first: the win-loss ratio has reduced to 0.60, from 1.50 in the period between 2000 and 2008. The percentage of instances of captains batting first has remained the same over these two periods - 57% since 2009, 59% between 2000 and 2008 - but the outcomes have changed considerably.

The runs-per-wicket stat in each innings over the two periods don't look very different, but it's obviously been enough to make a difference to the overall result. Add up the averages of the first and third innings (which is usually when the team batting first bats), and the total is 65.93 in the period between 2000 and 2008; the sum of the second and fourth innings is 68.55. The corresponding totals since 2009 are 71.38 and 62.88.

Since 2000, in Tests in Eng, Aus, NZ, SA
Parameter 2000 to 2008 2009 onwards
 W/ L ratio (team bat 1st)  0.78 (71-91)  1.78 (64-36)
 W/ L ratio (won toss and bat)  0.87 (45-52)  1.86 (39-21)
 W/ L ratio (won toss and field)  1.50 (39-26)  0.60 (15-25)
 1st inngs - Runs per wkt  36.57  37.73
 2nd inngs - Runs per wkt  35.57  34.17
 3rd inngs - Runs per wkt  29.36  33.65
 4th inngs - Runs per wkt  32.98  28.71

Since 2009, three teams have won at least twice as many Tests as they've lost when batting first - South Africa, England and Australia - but only South Africa have managed it when batting second. England's win-loss has dropped from 2.5 to 1.14 when fielding first, while Australia's has dropped from 2.17 to 1.0. New Zealand are the other team which have struggled when fielding first in a Test - their win-loss record drops from 13-11 to 2-11.

Teams batting and fielding first since Jan 2009
  Batting first Fielding first
Team Tests W/ L Ratio Tests W/ L Ratio
South Africa 27 11/ 4 2.75 24 14/ 7 2.00
England 40 20/ 8 2.50 40 16/ 14 1.14
Australia 46 26/ 12 2.17 27 11/ 11 1.00
New Zealand 32 13/ 11 1.18 21 2/ 11 0.18
Pakistan 21 9/ 8 1.12 36 12/ 14 0.86
Sri Lanka 33 10/ 11 0.91 24 7/ 8 0.87
India 25 8/ 10 0.80 36 16/ 10 1.60
Zimbabwe 8 3/ 5 0.60 6 0/ 6 0.00
Bangladesh 14 4/ 7 0.57 20 2/ 13 0.15
West Indies 23 5/ 12 0.42 35 8/ 15 0.53

S Rajesh is stats editor of ESPNcricinfo. Follow him on Twitter

Comments