Today, Cricmetric completes one year of its formal launch. This is a good opportunity for us to look back at the accomplishments in the past one year, and also to lay out the agenda for the future.
The best metaphor to characterize the success of Cricmetric so far is that of a glass which is half-full (or half-empty, depending upon your view point). The very fact that the site is still up and running after one year is an achievement in itself, given that none of us work on Cricmetric full time. In this one year time, we have developed new player performance evaluation metrics – both context dependent (win probability added – see for example, the WPA scores of the players in the ongoing season of IPL) as well as context independent (wins), live scoreboard for displaying the win probability of cricket matches in real time, and presented thoughtful articles on the Duckworth-Lewis method for target resetting in rain affected matches and improving the Decision Review System. In addition, we have also compiled an extensive database of limited over international and domestic games, and a player database – an activity that has been going on mostly in the background.
However, we haven’t accomplished as much as we would have liked to. We still don’t have a proper interface on the website to present the player and match statistics that we have gathered. We haven’t been able to write as many articles as we had planned to. Certainly we haven’t got as many eyeballs as we thought we will – but our work did get noticed by people in the Sports (and Cricket) Analytics community.
Going ahead, there are some short term goals, one big goal and one grand vision that we would like to accomplish. The short term goals are related to improving the website interface, and better presentation of the player statistics that we have accumulated as well as of the advanced statistics that we have developed. All these changes will be gradually rolled out in the next few months on the website.
The one big target we have set for the next one year is to present a completely new alternative for the Duckworth-Lewis method. There is no doubt that the DL method is the big elephant in the room, as far as Cricket analytics is concerned. It is the only tool used by the ICC (and many domestic Cricket boards) for target resetting in rain affected limited over games. But there are some clear flaws with the DL method, it is not easily understood by the average Cricket fan and at times teams are able to “game” the method to their advantage. There is a clear need for an alternative to the DL – an alternative that is fair, intuitive and one that cannot be gamed easily. In the next one year, we plan to achieve that.
And finally, our grand vision for Cricmetric. There is a significant difference between watching a passage of play in the match as opposed to just following the events ball by ball with just the basic statistics (bowler, batsman, runs scored, beaten, appeal etc). While the latter tells the current state of the game accurately and even the path to that state (at a superficial level), but an avid follower would know that it does not convey the same information as watching it live. For instance, the body language of the players has a lot of information that is not conveyed in just the events. Therefore, just the knowledge of events is not sufficient for any prediction or performance analysis. Our grand vision at Cricmetric is to reduce this huge information gap between the two alternatives – in other words, distil the events of a Cricket match down to a “quantitative commentary”, which is close to what you might have if you were watching the game live. And with this quantitative commentary, one can perform any kind of advanced statistical analysis of games, teams and players. The possibilities are endless.
We feel that this grand vision is a worthwhile endeavour not just for Cricmetric, but also for the Cricket fans, players, as well as the game of Cricket itself.