The format of the men's ODI World Cup has been in the news again this week, with reports coming out of Dubai that support is strengthening for a 16 team World Cup in 2027, though a 14 team tournament is still also a possibility. Obviously either would represent progress from the retrograde step of the 10 team World Cup, though given the ICC's history one must question whether it will be a true 16 team World Cup, and if it will have a format that will produce an exciting tournament.
The recent history of ICC tournament formats at the highest level is not a good one. Overlong tournaments are the norm with formats designed to protect the big teams as long as possible. Providing content for broadcasters seems to be the main goal of a World Cup, rather than being a celebration of the sport.
There is an assumption that the broadcasters specify the number of teams and/or the number of matches in the World Cup, though according to people who have seen the contract, it specifies neither of those things. Obviously though, the ICC want to maximise their revenue from broadcasters and try and give them what the ICC thinks they want. In theory this revenue would be used to fund development of cricket to produce more competitive teams, leading to bigger and better World Cups though in practice it has been used to line the pockets of the boards that need it the least.
So what makes a good format? Opinions vary, but my take is that the ideal format minimises dead rubbers and maximises "must win" games. Obviously the format that does this the best is a straight knockout tournament which by definition is entirely made up of "must win" games, but I'm ruling this out as it doesn't meet my other criteria of giving all teams a fair number of games and the tournament having a decent - but not unreasonable - length.
I'm also ruling out a T20 World Cup style 16 team format for the obvious reason that it isn't a true 16 team World Cup, instead being a 12 team tournament (10 prior to this year's event) with an extra qualifier tacked onto the front.
I will consider the first group stage separately from the second stage of the tournament and then combine the two to produce what I feel will be the best tournament from an entertainment perspective as well as one that takes into account the concerns of broadcasters.
First round formats
Hopefully it will be obvious to all that there are only three ways to divide 16 teams into equal groups. One group of 16, two groups of eight and four groups of four.
A single 16 team group where all play all would involve 120 matches. Even if two games are played each day and there are no reserve days, that's still a two month first round, with the possibility of the last month being full of dead rubbers. For this reason, this first round format can be ruled out instantly.
Two groups of eight teams would involve 56 matches. A month long if two matches are played, this falls down due to a high probability of dead games. If a group sees four teams each win their first four matches, 12/28 (43%) of games are dead rubbers, assuming the top four go through. If one or two teams are not competitive it can feel like a drag, as happened in the Super 8 stage of the 2007 World Cup, with no disrespect to Ireland or Bangladesh intended. For these reasons, I am also ruling this group structure out.
This brings us to four groups of four, which would feature 24 matches, 48 if a double round robin takes place, though I prefer single round robin. This format has a bad reputation due to the 2007 World Cup, though as I mentioned in the previous paragraph, the Super 8 stage was the bigger culprit that produced an overlong and rather boring World Cup. The first round was the last genuinely exciting first round of a cricket World Cup with plenty of surprises and - most importantly - it only lasted two weeks.
Second stage formats
Now that I've settled on four groups of four for the first round, the first decision to make is how many teams should progress from the first round. The first possibility is to just have the four group winners contest the semi-finals. This is obviously quick, but unsatisfactory as luck in the first round can play too much of a role. So I rule this out straight away and move on to eight team second stage formats.
Here we have a few possibilities. There's the Super 8 stage as with the 2007 World Cup, but as was made clear earlier, I am not a fan of this. Moving to straight knockout with quarter finals then semi-finals is another possibility which, paired with the four groups of four format could mean concluding the World Cup in a little over three weeks.
Another option would be double elimination quarter finals. This format is not unknown in sport as it was used in baseball's World Baseball Classic in 2009 and 2013. It means that the eight progressing teams get at least two games and, as it involves ten games at the last eight stage, prolongs the second stage, thus providing more matches. A variation on this would be to have the group winners play each other with the winners progressing to the semi-finals whilst the losers of these matches play-off against the winners of matches between the group runners up. This is a six game last eight stage and has the plus point of meaning that winning the group counts for something.
The final possibility would be to have two groups of four. Whilst not as cut-throat as straight knockout quarter finals, I'm of the opinion that jeopardy should increase as you progress in the tournament. For this reason, I'm ruling it out.
My favourite of those last eight stages would be the six game one where winning the group makes a difference, though I'm actually going to rule out eight teams progressing. This obviously is no pro-BCCI piece, but winning two of the three games and being eliminated on net run rate is not ideal. There are problems with net run rate in that - unlike goal difference in football - the venue matches are played at is a significant variable.
Varying boundary sizes means that runs can be easier to come by at different venues. This can be partially avoided by playing all matches in a group at the same venue, but this brings with it the possibility of too much wear on the square and also I prefer it when matches at a World Cup are spread around the host nation(s) rather than just four venues for the first round.
So now I move on to twelve team second stage formats. Obviously straight knockout can't work with twelve teams so we have to look at group stages and play-off style formats. Straight away I'm ruling out any possibility of a single group of twelve or two groups of six as this goes against my desire to increase jeopardy as the tournament progresses and also prolongs the World Cup to an even more interminable duration.
The other possibilities are three groups of four and four groups of three. The first of these would mean that there is a situation where a best runner-up has to progress to the semi-finals, and comparing performance across different groups is hardly ideal given the varying strength of opponent each team will have faced. Four groups of three is a more attractive possibility, but if the first two games in a group are won by the same team then the third in the group is a dead rubber, assuming only the group winners progress. Both formats also mean that the position a team finished in during the first round counts for nothing.
This moves us on to play-off formats. The most obvious would be to have the group winners automatically qualify for the quarter finals with cross over matches between second and third place teams to join them. This is the quickest but does mean that teams are only sure of one second stage match, leaving us with play-off formats that, whilst more convoluted, do make it more likely that the best teams go furthest in the tournament and reduce the luck factor.
There are several possibilities here, but I'm going to go straight to my favourite format. This would involve the four group winners playing each other, the four runners-up playing each other and the four third place teams playing each other. The winners of the matches between the group winners go straight to the semi-finals, whilst the losers of the third place team matches are eliminated.
The winners of the third place matches then play the losers of the second place matches with the winners of those then facing the winners of the second place matches. The winners of those matches then play the losers of the first place matches for the remaining two semi-final spots. Whilst I admit this is a complicated play-off system, it has the advantage of meaning that first round performance determines how easy a path a team has to the semi-finals - a first place team only needs one win whilst a third place team needs three. It also means that the first round - and by extension the whole tournament - has no dead rubber matches.
This second stage will involve 12 matches. Combined with 24 from the first round, two semi-finals and a final this gives us a 39 match tournament. Still big enough to satisfy the broadcasters desires for plenty of matches but not so big that it becomes a drag to watch. Only the four last place teams are eliminated after just three matches, whilst 10 of the participating teams get at least five matches. Playing two matches a day during the first round will allow us to get the tournament over and done with in under five weeks, in sharp contrast to recent World Cups that seem to last forever.
Will the ICC use such a format though? History suggests not. Even the greatly reduced jeopardy of early elimination for a big team of this format might be too much for the "big three". Sadly, I feel we are more likely to get a two groups of eight format, followed by knockouts. But still, that a sixteen team World Cup is back on the table at ICC HQ in Dubai is significant progress to undoing the mistakes of the ten team World Cup, even if we do get a rubbish format.