Monday, January 07, 2013

ICC, BCCI Do Not Understand Draws

Teams play to win. They play to get the better of their opponents and end up ahead. They play to defeat their opponents. In Cricket, there is a subtlety to this endeavor which is not found in many other sports. This is the notion of a Draw.

Unlike in other sports, in Cricket, a Draw does not mean parity. It merely means that a definite conclusion was not reached. This is a distinction that is lost even on the ICC's official rankings method. Not only is the distinction lost, its insignificance is reflected by the fact that the basic rule about how a Draw is measured as opposed to a Win is not even available on the ICC's website. According to wikipedia, 1 point is awarded for a Test win, and 1/2 point for a Draw. An additional point is available for a series win. A recent presentation by David Kendix, the ICC's official statistician does not even mention what happens in case of a draw in Tests. The FAQ is silent on this point as well.

To assume parity, and divide 1 point equally between two teams in the event of a Draw in a Test Match, is plainly wrong.

The Draw is misunderstood in other ways as well. The Ranji Trophy tournaments has three knock out matches - Quarter Finals, Semi Finals and a Final. This, in a tournament of 4 day games. It is no surprise that teams like Mumbai approach the knock out match as they do. In the ongoing Ranji quarter final against Baroda at the Wankhede Stadium, Mumbai have now batted for 2 full days and reached 6/524. They are playing, first for the first innings lead, with the outright win being an after thought. Why wouldn't they? The first innings lead will mean that they reach the semi finals of the Ranji Trophy.

If the ICC does not understand the Draw because they conflate inconclusiveness with parity, then the BCCI does not understand the Draw, because they draw conclusions from a result which is designed to avoid them. The whole point of an inconclusive result is that there is no winner and no loser. Yes, one team almost certainly was closer to victory or defeat than the other. But this is not reason enough to effectively declare one team the winner on the first innings lead, even if you impose a requirement that both innings have to be completed in order for a result to be possible. After all, the Ranji Trophy is not a single innings tournament is it?

Instead of wasting the whole of January on playing 4 quarter finals, 2 semi finals and 1 final, it would be better to reduce the number of Ranji Trophy groups to 2 - one with 14 teams and the other with 13, and play 4 extra rounds of league matches. This would give players more first class experience, it would provide more games for India players to return to if they need some match play to test their fitness and form after an injury or illness, and, it would get rid of this ridiculous obsession with batting and put the onus on the bowlers for once. The way to win Test Matches is by having high quality bowlers, just as the way to produce high quality batsmen is have quality bowlers bowling on good pitches.

If further incentives have to be offered for teams to play for a result - to provide incentives for a team to declare at 450/5 instead of plodding along to 550 all out, perhaps the Draw ought to be penalized marginally more than a Loss as I have proposed earlier. There are several ways to do this.

1. The Bonus Points Approach (BPA): In this approach, apart from a baseline of points for specific results (say 5 for a win, 2 for a draw and 0 for a loss), 10 points (to be divided proportionally between teams) are available for in a loss, while 5 points (to be divided proportionally between teams) are available for a draw.

2. The Raw Points Approach (RPA): In this, the Raw Points accumulated by a team, and those conceded by a team would be calculated, and these would be simple difference between the two at the end of the season would determine the ranking. The larger the difference, the higher the ranking.

Here is an illustration of these two approaches using 5 match scenarios. The formulas in parenthesis in the "Category" column refer to the Sr. No of the row mentioned in the first column.

If the Ranji Trophy league were to be expanding, with the final and the other knock out rounds being eliminated, either of these systems would be useful for determining a winner of the Ranji Trophy. In Test Cricket, either of these systems would work as well, using the ICC's system of weighting matches - matches in the 2 most recent years are given highest weightage, while matches in the 2 years before these, are weighted at 50 percent.

The examples in the above table are designed to compare different approaches to a 4 innings match

Match 1 is an example of a bold approach by Team A - a declaration at 450/5 - a lead of 150, risking sizable 4th innings chase, but making a result more likely by moving the match along. Team A wins in this scenario chasing 176 in the 4th innings.

Match 2 is a more conservative approach from Team A. A longer, bigger first innings (2nd match innings), followed by a follow on and shorter 4th innings run chase. The team uses up all its resources in the first innings and builds a lead of 225. As you will see, even the best case scenario - a 10 wicket win for Team A chasing 101, does not improve its BPA or RPA scores (line 25 and 27 respectively).

The approach in Match 1 makes the outcome of the type seen in Match 3 more likely. In this instance, Team A bats 2nd, declares 150 ahead, and then gets set a target of 175 in the 4th innings. You can imagine a team doing this if it is willing to improve its chances of scoring an outright win by risking defeat. Team A survives and avoids defeats by 2 wickets in the 4th innings, finishing 50 short of the target.

The approach in Match 2 makes the outcome of the type seein in Match 4 more likely. Team A runs out of time because it played the 1st innings more conservatively. Set 101 to win in perhaps an hour on a 5th evening pitch, it has to settle for a draw. If you compare the Draws in Match 3 and 4, bolder approach in Match 3 (even in the worst case scenario), is not penalized that much more than the more conservative approach in match Match 4.

Match 5 is the worst case scenario for Team A's bold approach. They end up losing by 25 runs.

But consider the arithmetic here. This arithmetic is why I prefer the Raw Points Approach (RPA) slightly more to the Bonus Points Approach (BPA, which is a compromise between the current system and one that only measures performance). Both these approaches mean that there is something to play for on every single ball - there is no incentive for the game to meander along, simple because the benefits of winning are so much higher than the benefits of a draw. In BPA, a win is worth about 3 Draws (wins will typically accrue somewhere between 11 and 13 points, while draws will accrue somewhere between 4 and 5 points). In the RPA, it is worth about 7-8 Draws if you turn an advantageous match situation into a win, and about 4-5 Draws if you lose trying to turn an advantageous match situation into a win.

The RPA incentivizes chasing outright wins better than the BPA. But the BPA would still be a massive improvement, both in the Ranji Trophy and in the ICC Test Championship.

The current system, by allowing teams to "win" based on a first innings lead, tamper with the contest between bat and ball, by turning cricket into a contest between the bat and the clock, and the ball and the clock. This, as I have shown in this post, is because both BCCI and ICC (in this instance) misunderstand that nature of a draw. Contrary to the conventional criticism of the Ranji Trophy, which is that teams don't play to win, they do play to win. The problem is that they are allowed to "win" without actually winning.

If you think this system is complicated, think about the fact that it is entirely contained, from raw match scores to final outcomes, in that single table. Anybody who can add, subtract, multiply and divide, will be able to use this system. It has one added virtue - it is not easy to game, as it is difficult to predict exactly how many points a team will end up with based on that team's performance alone. So it is not a case of - "If you score 400 runs you get so many points". It is a case of "if you score 400/5 you will get so many points if some other things also happen, but if you score 400/9 you will get a significantly different number of points if some other things also happen". The only given is that at any given point in a game, chasing an outright victory guarantees the most beneficial points distribution for any team.


  1. ICC is good for nothing, its BCCI who controls the proceedings.

  2. I disagree with your view that points should be awarded based on a different system, but this isn't because I enjoy 1st innings batting marathons. Rather, in the case of a draw in a knockout match the advancement should be awarded to the team who scored at a faster run rate over the course of the match. The implication of this would be that this team had forced the pace of the match, and played for a result, wheareas the other side had played for a draw and therefore voided their chances of winning. Plus it would be a much simpler system for determining the winner.