T20 Ratings - World Cup Edition (final)
Russell Degnan


Rankings at 16th June 2009form
1.Sri Lanka1332.37+103.40
2.South Africa1320.12+31.17
4.Pakistan1279.52+37.51
6.West Indies1190.47+37.30

Prediction for the final
Pakistan v Sri Lanka 5 runs

Sri Lanka's annihilation of the West Indies (and most everybody else) has put them to the top of the ratings going into the final. Their probability of victory comes to 58%.

A word too on form, and perhaps a question. The main purpose of the form variable is to smooth out the volatility of unexpected results, by "banking" them, such that the effect of that result will not carry over until later games. If a team consistently loses worse than expected their rating will shift more quickly to its new equilibrium. On the other hand, if the loss was unusual, the rating will stay relatively stable.

Which brings up a question, possibly philosophical: is form an indicator of where the rating will go, and therefore should be added to make predictions; or where the rating has been, and therefore should be deducted because that is where it may return? In other words, is form a meaningful number, worth reporting, and in what sense? If results were completely random, then a run of bad form would dip the rating more than necessary, which oddly enough will actually increase volatility (sort of like irrational exuberance in a stock market). Conversely, the relative form of the West Indies and England prior to their Carribean match-up indicated that the series would be closer than the ratings might indicate; similarly Australia's form in test cricket began to dip markedly even before their horrid run against India and South Africa.

The reason for asking this relates to Sri Lanka's current form, currently over 100. Partly, their form indicates only that they are winning. The expected margin against other top-8 sides are so narrow that almost all wins beat the margin by a lot. Similarly, South Africa were equally dominant leading into their semi-final against Pakistan. The numbers are inconclusive to date, but anyone looking for an extra reason to lean one way or the other may like to consider the implications of strong form on their expectations.

Idle Summers 20th June, 2009 13:08:57   [#] 

Comments

T20 Ratings - World Cup Edition (final)
I guess my approach to answering the form question would be to run a regression of results against rating, and then results against rating and form, and see difference it makes, and what the sign of the co-efficient of the form variable is.

I have a question about your figure of 58%. In an earlier post you said that you needed a 400-point rating difference for the probability of victory to approach 65%. The 58% of this post and the 65% from the earlier post don't seem to agree.
David Barry  22nd June, 2009 12:09:16  

T20 Ratings - World Cup Edition (final)
DB, thanks. The regressions don't tell you much. There isn't a lot of data, but adding form to the equation gives slightly worse predictions. This may be because the ratings are tweaked to make the normal margins accurate, however. The difference between them is miniscule.

Regressing form against the change in ratings was more illuminating. There is an insignificant negative correlation (-0.047) between them, which would tend to fit the idea that form is irrelevant as an indicator of future success. I might play with the test figures when I get a chance to see if they differ.

On the percentages, mea culpa. You are right, I was reading an out-dated column that needed updating. The first figure is correct I think, but I haven't studied it in depth. I based the the first figure by calculating the standard deviation of margins (which are normally distributed) against predicted margins.

I actually have another problem now. I am not happy with the volatility of the T20 ratings, so I am going to lower the sensitivity in line with how I do the test ratings. But in the process of tweaking I am noticing a weird problem for the home team: the value of being at home regresses properly through the axis at a measly 20 points (about 2 runs in the margin). Whereas, you cited data earlier that would indicate a very significant home advantage. It may just be a lack of data - most games are neutral. But it is odd.
Russ  22nd June, 2009 14:08:39