To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Data Mining Approach to Predict Match Outcome in One-Day International Cricket |
|
Author |
iWaqar Ahmed, Khurrum Nazir Junejo, Tariq Mahmood, and Ghulam Mujtaba |
| Citation |
Vol. 26 No. 2 pp. 187-194 |
|
Abstract |
Accurate prediction of the likelihood of a team to win a game of sports even before it starts is an insight worth millions to coaches, managers, sports analysts, and media persons. In this study, we propose an approach to predict match outcome for the game of Cricket; the second most watched sports in the world. We successfully predict the winner of One-Day International (ODI) cricket match 80% of the time even before the start of the game. To achieve this we consider all ODI matches played between 1971 and 2014 to form a data-set that comprises 11 attributes and 3534 match records. We put comprehensive effort in collection and pre-processing of the raw data to identifying the most decisive attributes for the prediction. We then make use of six well-known machine-learning approaches while experimenting with different intervals, sampling, and attribute selection techniques. Our approach achieves a gain of 25.00% in prediction accuracy with respect to baseline winning ratio of the team. We observe that data of recent matches has a strong influence on the prediction of match outcome. Using our prediction tool, Cricket managers can choose the most appropriate squad for forthcoming ODI match, whereas coach and captain can shape their strategies before the match starts. Furthermore, cricket analysts and media can also use the model for pre-match analysis. |
|
Keywords |
One-Day International, cricket, outcome prediction, classification model, performance evaluation. |
|
URL |
|
|
Title |
Data Mining Approach to Predict Match Outcome in One-Day International Cricket |
|
Author |
iWaqar Ahmed, Khurrum Nazir Junejo, Tariq Mahmood, and Ghulam Mujtaba |
| Citation |
Vol. 26 No. 2 pp. 187-194 |
|
Abstract |
Accurate prediction of the likelihood of a team to win a game of sports even before it starts is an insight worth millions to coaches, managers, sports analysts, and media persons. In this study, we propose an approach to predict match outcome for the game of Cricket; the second most watched sports in the world. We successfully predict the winner of One-Day International (ODI) cricket match 80% of the time even before the start of the game. To achieve this we consider all ODI matches played between 1971 and 2014 to form a data-set that comprises 11 attributes and 3534 match records. We put comprehensive effort in collection and pre-processing of the raw data to identifying the most decisive attributes for the prediction. We then make use of six well-known machine-learning approaches while experimenting with different intervals, sampling, and attribute selection techniques. Our approach achieves a gain of 25.00% in prediction accuracy with respect to baseline winning ratio of the team. We observe that data of recent matches has a strong influence on the prediction of match outcome. Using our prediction tool, Cricket managers can choose the most appropriate squad for forthcoming ODI match, whereas coach and captain can shape their strategies before the match starts. Furthermore, cricket analysts and media can also use the model for pre-match analysis. |
|
Keywords |
One-Day International, cricket, outcome prediction, classification model, performance evaluation. |
|
URL |
|
|
Title |
A Data Mining Approach to Predict Match Outcome in One-Day International Cricket |
|
Author |
iWaqar Ahmed, Khurrum Nazir Junejo, Tariq Mahmood, and Ghulam Mujtaba |
| Citation |
Vol. 26 No. 2 pp. 187-194 |
|
Abstract |
Accurate prediction of the likelihood of a team to win a game of sports even before it starts is an insight worth millions to coaches, managers, sports analysts, and media persons. In this study, we propose an approach to predict match outcome for the game of Cricket; the second most watched sports in the world. We successfully predict the winner of One-Day International (ODI) cricket match 80% of the time even before the start of the game. To achieve this we consider all ODI matches played between 1971 and 2014 to form a data-set that comprises 11 attributes and 3534 match records. We put comprehensive effort in collection and pre-processing of the raw data to identifying the most decisive attributes for the prediction. We then make use of six well-known machine-learning approaches while experimenting with different intervals, sampling, and attribute selection techniques. Our approach achieves a gain of 25.00% in prediction accuracy with respect to baseline winning ratio of the team. We observe that data of recent matches has a strong influence on the prediction of match outcome. Using our prediction tool, Cricket managers can choose the most appropriate squad for forthcoming ODI match, whereas coach and captain can shape their strategies before the match starts. Furthermore, cricket analysts and media can also use the model for pre-match analysis. |
|
Keywords |
One-Day International, cricket, outcome prediction, classification model, performance evaluation. |
|
URL |
