All the English commentators and cricketers were circumspect in their reactions, but we don't have to be: If England isn't in the ICC World Twenty20 semi-final, it is, to a large extent, because of the flaws in the Duckworth-Lewis (D/L) method devised by two Englishmen!
After England scored a respectable 162/6 in 20 overs, West Indies were required to score only 80 in 9 overs with all their 10 wickets available.
Half 162 in 20 overs and we get an equation of 81 to get in 10 overs.
D/L asked England to get those runs in 9 overs. So, by sacrificing a mere six balls, West Indies had the luxury of having all its ten batsmen go hammer and tongs to reach the target. This was a patently unfair target reset by the D/L method.
Actually, the Duckworth-Lewis method now works well for 50-over ODI cricket games for which it was designed. The problem lies in ICC and D/L pretending that the method can work just as well in 20-over games.
We saw that also in the IPL: if captains walked out to toss with the threat of rain, they always chose to chase. This wouldn't have happened if D/L was indeed fair and toss-invariant.
To understand why, just look at the D/L curves below that I plucked off Wikipedia:
Look at the 'spread' of those ten curves with '50 Overs Remaining', and see how much closer the curves come to one another with '20 Overs Remaining'. And notice that by the '10 Overs Remaining' mark, the curves are practically overlapping.
When curves overlap, it means that D/L is no longer compensating for the advantage of having more wickets; the D/L resource is no longer a combined resource featuring both 'wickets in hand' and 'overs remaining' it's now only a matter of runs and overs!
So what's the moral of the story?
For captains on the field it is: even if there's a small chance of rain, always choose to chase and, when you come out to chase, don't worry about losing wickets just worry about hitting big runs off every available ball. That's exactly what Chris Gayle's team did last night.
For ICC the lesson is: stop fooling yourself into believing that a model devised for 50-over games will work for a 20-over game. Get back to the drawing board and create a new model based on real 20-over game data.
And if that's too hard, you might want to consult a certain V Jayadevan from India -- who has already created a model for the 'untouchable' ICL.