Wednesday, 8 June 2011

England v Sri Lanka, 2nd Test at Lord's: Day 5

So, somewhat inevitably, the second Test rolled to a draw. The fifth day at Lord's was rather inconsequential to be honest, so we won't linger on it (or the window incident). Here is the tale of the five days, in terms of how much each player contributed to his team's chances of winning/losing/drawing the match:

Eng win
SL win
Tremlett 14.9% -7.3% -7.7%
Morgan 7.7% -9.6% 1.8%
Strauss -8.2% 21.5% -13.4%
Finn 0.3% 2.4% -2.7%
Trott -0.5% 0.0% 0.5%
Cook 1.8% -26.7% 25.0%
Bell -5.5% -1.3% 6.7%
Pietersen -11.8% 7.9% 3.9%
Broad -18.6% 10.5% 8.1%
Prior 17.7% -21.7% 4.0%
Swann -26.1% 11.6% 14.5%

SL win
Eng win
Maharoof -19.2% 16.8% 2.4%
Weleg'ara 6.0% 6.2% -12.2%
P.Jay'dene 2.3% -4.5% 2.2%
Sang'ara -3.1% 5.1% -2.0%
Parav'ana 0.0% -9.2% 9.2%
Sam'eera -3.0% 4.0% -1.0%
M.Jay'dene 4.2% -8.7% 4.5%
Herath -6.7% -8.1% 14.8%
Lakmal -4.2% -0.2% 4.4%
Fernando -13.4% 4.7% 8.7%
Dilshan 19.4% -29.3% 9.8%

And here is the story of the match in graphical form; note the extent to which England's decision to keep batting on hindered their chances of winning:


  1. So in England's 2nd innings, their win probability peaked at about the 300th overs in the match - what was their score then? Using cricinfo's scorecard I've only narrowed it down to 'before KP got out', so about 200 ahead?

    And it's interesting to see that SL's chances of a win went to zero at about the same point - or earlier.

    I can't imagine any captain, on a pitch that has yielded so many runs, declaring with such a small lead...

  2. Hey Neil, well spotted. Clearly SL's win prob shouldn't go to zero quite so quickly: this is one of the flaws (as in the SA v India 3rd Test). Basically at the moment the program looks at all the potential scenarios that can occur; to assess draws, a fixed run rate is assumed and the scenario is classed as a draw if it exceeds the number of overs remaining. This works well for early game scenarios but does cause problems late in the game when teams have the option of upping the run rate. We are working towards using probability distribution functions for run rates, or being able to change the run rate based on the game situation, to get around this issue...

  3. A bit late but... I like the way that the contributions highlight that Cook's 3rd innings go-slow helped force the draw.

    Aggressive batsmen (Tresco, KP, Prior etc) have long been given credit for helping win games. Doughty accumulators have their downside, imho.