It will be an India-South Africa final

I was planning to write a nice long blog previewing the Champions Trophy semi-finals, and reviewing the tournament so far, but instead became embroiled in a stats-wrestle with Jonathan Trott's ODI career. The stats have not quite finished belly-slamming me yet, so I will post the results in next week's Confectionery Stall, once the destiny of the trophy, and Trott's role in it, have been concluded and sent off to the laboratory for analysis. Seldom has an England ODI player provoked such strongly voiced opinion. Particularly whilst averaging over 50 in a format that has generally been a vehicle for English pain.

However, the numbers he produces in ODIs are intriguing - for example, when England bat first, Trott averages 44 in victories and 75 in defeats - and highlight the shortcomings of traditional career average and strike rate as a means of measuring performance, especially in limited-overs cricket. It was pointed out by several pundits that Trott's career strike rate (76.5) compares favourably with those of Sri Lankan legends Sangakkara (75.9) and Jayawardene (78.4), whilst the England player's average, currently 51.7, is significantly better than Sanga's 39.1 and Mahela's 33.4. But this is measuring a player who has played only since 2009 against two who had played throughout the previous decade, when scoring rates were lower (since Trott's debut, Sangakkara has scored at 79.9, and Mahela at 84.8).

Furthermore, the fact that Trott's average is so astronomical highlights the relative stodginess of his scoring rate; whilst also rendering that provably moderate scoring rate of lesser importance - his function in the team is as an almost supernaturally reliable guarantor of runs. Tune in for the next exciting episode of "Jonathan Trott's ODI Statistics - A Journey Into the Purpose of Humanity", next week. It will probably be even more irrelevant than what usually adorns this blog, but I think the results will be interesting. For those who find such things interesting.

In the meantime, a few brief thoughts on what has been an excellent exhibition of international cricket, blessed with an unusually generous smattering of high drama, cursed by an irritating amount of rain that may become even more irritating over the next few days, and give further fuel to those who criticise the Callaghan government of the 1970s for not building a retractable roof over the whole of Britain when the North Sea oil money was flowing into Britain's now emphysemic coffers.

The belatedly regenerated India have been the only truly convincing team in the group stage, but, such is the nature of one-day cricket, that is no guarantee of victory. Dhawan's batting has made one grateful for the invention of cricket, even more daring and stylish than his facial whiskery, whilst the team's fielding must be making Virender Sehwag think his television has developed a fault and is playing everything at twice the normal speed.

England look the most likely to usurp them, looked imposing in their first game, alarmingly vulnerable in their second, and, for whatever can be read into a 24-over thrash, decent in their third.

(Breaking news in the non-existent ball-tampering row: England have reportedly admitted "psychologically intimidating" cricket balls to terrify the 5½-ounce seamed white orbs into reverse swinging. A source close to the team admitted: "The bowlers growl at the ball like angry bears whilst they're walking back to the end of their run-ups. Sometimes, they even threaten to have the ball sent to be used in net practice at an IPL franchise. The Kookaburra balls used in this tournament are much more susceptible to this kind of mental coercion than, for example, red Dukes balls, which tend to be more at ease with themselves." The source also claimed that Tim Bresnan is the team's leading ball-intimidator. "The balls find Mr Bresnan terrifying. One mild frown of his Yorkshire brow and they will do whatever they're told.")

South Africa's batting, despite thus far firing only on some cylinders, and only some of the time, could still win any match. They were one ball away from elimination against West Indies - or one fewer minute of dry weather - and were only adequate against Pakistan. Adequacy, however, was all any side required against a batting line-up that looked flimsy on paper before the tournament, and convincingly lived down to that lack of promise.

Sri Lanka were majestic in their chase against the hosts last week, but have been otherwise fitful. By the look of the forecast, one or both of Duckworth and Lewis could find themselves trudging off with a couple more Man-of-the-Match awards.

Of the other eliminated teams, New Zealand played a soggy total of just 122 overs of relevant cricket, before leaving the tournament with a paradoxical squirt in the eye from the lemon of irony, when they were ultimately undone in part by the rather batty Net Run Rate calculation system, which recorded their knife-edge one-wicket win over Sri Lanka, in a match more evenly balanced than a tightrope-walking philosopher with identical watermelons for ears, as a thumping one-sided demolition job. As a result, Australia had to chase in their final match with such reckless abandon that a Sri Lankan victory became almost inevitable - thus dumping New Zealand out of the tournament, despite and/or because of the unfair NRR advantage they themselves had been given.

It is hard to pass judgement on Australia. They were pallid against England, then rained out whilst being less pallid against the Kiwis, before that final, statistics-skewed frenzy against Angelo Mathews' side, a rare instance in which two sides were, in effect, playing different matches on the same pitch at the same time. West Indies were West Indies. Could have been better, could have been worse. Simultaneously unlucky and disappointing.

Overall, this final edition of a little-loved competition has made a potent case (a) for 50-over cricket continuing to be a significant part of the international calendar, and (b) for there being less 50-over cricket in the international calendar. The fiddlings, tinkerings and quackery with the ODI format are neither here nor there. The format works when the teams are well-matched, the games have significance, and the pitches produce varied cricket. Personally, I would be happy to see an eight-team tournament with this format, alternating every two years with a 16-team World Cup, which would be played with, essentially, the same structure plus an additional first group phase.

I know there are not currently 16 teams good enough to participate in a World Cup, but (a) no sport has even found a satisfactory format for a number of teams that is not two to the power of something, and (b) international cricket must not restrict its expansion to the T20 format. An extra group stage would add only ten more days to the length of this current contest, making it still three weeks shorter than the unremittingly and wilfully unsatisfying 2007 World Cup. And you could lob in a plate competition for the nations eliminated in the first group stage, so the teams would all be ensured six matches, and the lower-ranked nations would have both the chance to play the world's best, and an opportunity to win a trophy.

● The final tally for my coin-toss predictions in the group stage: nine correct, one wrong, plus a tie and a no-result. However, whilst correctly predicting the four semi-finalists, my one-pence piece has foreseen India v England and Sri Lanka v South Africa as the semi-finals. It has now been forcibly retired in disgrace, and used to buy approximately nine millilitres of milk. The coin's replacement - a spinning empty water bottle - has just predicted a South Africa v India final, to be won by the Proteas.