What's in a rematch?

You could write a book listing and explaining the reasons why the format of the World Cup is wrong. It might not sell particularly well, and your publisher might ask you to tone down some of the language if you commission someone from an Associate member nation to write an appendix about the ten-team format for the 2019 tournament, but it would nonetheless be a chunky tome to keep in your lavatory, use as a doorstop, or burn on a cold winter's night.

The format for 2015 - 42 matches, dragged out over a month, to painstakingly reduce the 14 teams down to the best eight, followed by a breakneck seven-game knockout to reduce those eight down to one, ensuring that the eventual world champions might well be decided in significant part by luck - is not without its benefits.

There are a large amount of matches, which is good if, for example, you really enjoy watching cricket matches, a social category that, I assume, contains most of the readers of this column. It is an even better format if you enjoy watching matches that are not particularly important in the grander cricketing scheme of things, or one-sided drubbings.

Best of all, the knockout stage will be gripping. One flare-up of a strong team's Achilles heel could be enough to send them home; one surge of slugging brilliance could take a less-fancied team through; a few high-octane new-ball overs could negate a potent batting line-up; a three-game form-spike is all that is needed, which might even be within the compass of West Indies. Or England. Or the UAE. But probably not the UAE.. Definitely not the UAE. The tournament will end well, and dramatically. It will also have a significant lull in the middle weeks.

In the last World Cup, after 30 days, we were left with the eight traditional Test nations in the quarter-finals. Bangladesh almost sneaked through, England almost sneaked out, and Ireland upset some applecarts, but, essentially, not only would you probably have correctly predicted the quarter-finalists before the tournament began, but you would also probably have correctly predicted them 20 years previously. Or 30 years previously, if you had correctly read the runes that South African politics was going to change a bit. A very big bit.

The bookmakers' odds suggest that there is approximately a 65% chance that all of the top eight teams will make it to the quarter-finals (and, conversely, that there is only a one-in-three likelihood of any of Zimbabwe, Bangladesh and the four Associate teams making it out of their group), which is not a massive amount of dramatic jeopardy with which to fuel a month of sporting tournament. Clearly, it must be difficult, if not impossible, to optimise both the commercial and sporting ideals of a tournament such as this in a sport with a limited number of top-level teams, but, whilst the 2011 and 2015 format is by no means the worst possible World Cup option (as proved by, for example, 2007), it is a significant distance away from being the best, or even good.

Nevertheless, a World Cup brings its own weight and drama, especially one in which the host nations have realistic aspirations of victory, and the tournament begins with a flurry of heavyweight showdowns in which momentum, whatever that is worth, and form can be built up.

More importantly, psychological points can be scored that might impact the latter stages. Might. Not will. But might. There is a reasonable chance that one or more of the final three matches at the end of March will feature of rematch of a group-stage game.

There was only one in the 2011 edition - Sri Lanka's semi-final win against New Zealand, in which they repeated a group-stage victory - but if such rematches do transpire in the semis or the final, the statistics suggest that the team that won the first game is highly likely to win the second.

Seventeen World Cup knockout matches have been played between teams that had previously met earlier in the same tournament. Only three of those 17 have been won by a team that lost the first encounter against the same opponents - the 1999 final, when Australia thrashed Pakistan, after losing a tightly fought group match to Wasim Akram's team; the 1996 semi-final, in which Australia avenged a group-stage defeat to West Indies; and the 1983 final, when Kapil Dev's India beat the West Indians, although this is something of a rogue entry in this list, as the teams had played twice in the group phase, winning one and losing one each. As India had lost the second of the group matches, however, we will keep it in the stat. No arguments. This stat is now legally binding.

Pakistan won the 1992 final against England, after having been saved by a rain-induced no-result while being comprehensively obliterated in a group match; and the 1999 semi-final between Australia and South Africa, following Australia's dramatic Super Six win days earlier, was technically a tie, but since Australia went through on net run rate, it basically functioned as another win for the team that won the earlier contest.

The other 12 knockout rematches have been won by the team that triumphed in the earlier match (13, if you include the 1983 Indians repeating their group win against West Indies) (which starts to get confusing) (that is excluded from the stat) (but replaced by the 1999 Australia-South Africa tied rematch, which I am now counting as a win). Furthermore, all of the seven rematches that have been played in the last three World Cups have been won by the team that won the initial game.

Due to the changed format used in 2011 and 2015, such meetings are less likely to happen - just that one in 2011, as opposed to an inevitable three rematches in each of the previous three tournaments, when the Super Six / Super Eight format meant that the semi-finalists had played each other in one or other of the two group phases.

Nevertheless the evidence suggests that although the month-long group stage somewhat inexplicably retained for this World Cup (while being simultaneously jettisoned for the next two tournaments) may serve little function other than to all-but-guarantee the presence of most or all of the more established cricketing nations in the knockout stages, a victory over a potential future semi-final or final opponent could be of considerable advantage. Even if that win occurs weeks and weeks before the climactic final few days, and no one can quite remember when, why or how it happened.

1975
West Indies beat Australia in group stage (by seven wickets) and in the final (by 17 runs).

1979:
No rematches.

1983
India beat West Indies (by 34 runs), then lost to them (by 66 runs), in group stage; then beat them again in the final (by 43 runs).

1987:
No rematches.

1992
Pakistan beat New Zealand in the round-robin (by seven wickets) and in the semi-final (by 4 wickets).
England beat South Africa in round-robin (by three wickets) and in the semi-final (by 19 rain-assisted runs).
[Also: Pakistan had a deluge-aided no-result with England in the round robin, then beat them in the final (by 22 runs).]

1996
Sri Lanka beat India in the group stage (by six wickets) and in the semi-final (by fire-aggravated forfeiture, whilst winning comfortably).
Australia lost to West Indies in group stage (by four wickets), but beat them in the semi-final (by five runs).
[Also: Sri Lanka gained a technical walkover victory against Australia in the group stage when Australia forfeited the match due to safety concerns, then beat them in the final (by 7 wickets)]

1999
Pakistan beat New Zealand in the group stage (by 62 runs) and in the semi-final (by nine wickets).
Australia beat South Africa in Super Six (by five wickets after Gibbs foolishly dropped the World Cup), then tied with them in the semi-final, knocking them out on net run rate.
Australia lost to Pakistan in the group stage (by 10 runs), then powerclobbered them in the final (by eight wickets).

2003
Australia beat Sri Lanka in the Super Six (by 96 runs) and in the semi-final (by 48 runs (D/L).
India beat Kenya in the Super Six (by six wickets) and in the semi-final (by 91 runs).
Australia beat India in the group stage (by nine wickets) and in the final (by 125 runs).

2007
Sri Lanka beat New Zealand in the Super Eight (by 6 wickets) and in the semi-final (by 81 runs).
Australia beat South Africa in the group stage (by 83 runs) and in the semi-final (by seven wickets).
Australia beat Sri Lanka in the Super Eight (by seven wickets) and in the final (by 53 runs (D/L).

2011
Sri Lanka beat New Zealand in the group stage (by 112 runs) and in the semi-final (by five wickets).