DHONI’S VIR­TUAL TRI­UMPH

A ded­i­cated team of 45 tech­no­log­i­cally ac­com­plished back­room boys helps Team In­dia hit trou­ble­some op­po­nents for a six

India Today - - CRICKET - By G. S. Vivek

Right through the sixth edi­tion of the In­dian Pre­mier League ( IPL) this April, Chris Gayle was blur­ring the lines of re­gional ri­valry that the league had worked so hard to cre­ate. As he blasted his way to 175 not out against Pune War­riors on April 23, the cho­rus of ap­pre­ci­a­tion led by his Ban­ga­lore skip­per Vi­rat Kohli was joined by some of the op­po­si­tion play­ers and fans. But even as he was turn­ing out to be the toast of the sea­son, Team In­dia’s an­a­lyt­ics team was con­cen­trat­ing on the prob­lem that lay ahead. In less than two weeks af­ter IPL, Team In­dia would have to face Gayle in the Cham­pi­ons Tro­phy in Eng­land, and they wanted to be ready for him.

With fin­gers fly­ing on their lap­tops, the techies, who work as the national team’s brain trust, put to­gether a plan. Bhu­vnesh­war Kumar was the only bowler who had es­caped Gayle’s wrath in IPL, and Team In­dia’s data sci­en­tists be­gan look­ing for, through pat­terns and vis­ual data mod­ules, why Gayle had been un­able to get him away. Their find­ing was that Gayle had prob­lems hit­ting in the air against bowlers who were quicker than 140 kmph, com­pared to medi­umpac­ers who bowled in the late 120s and mid- 130s. They also found that he had a prob­lem early in his in­nings against pitched- up de­liv­er­ies that swung away. When the teams fi­nally met at the Oval on June 11, Gayle was dis­missed for an 18- ball 21. Bhu­vnesh­war had the left- han­der caught at slip with an away-go­ing ball. Back in the dress­ing room, the team’s video an­a­lyst Dhanan­jaya pumped his fist. Mis­sion ac­com­plished.

There may be sev­eral rea­sons be­hind Team In­dia’s Cham­pi­ons Tro­phy vic­tory, and their ODI World Cham­pi­ons tag. One of them is the in­sight they get from a group of 45 techies work­ing for Sports-Me­chan­ics, which func­tions from a non­de­script sin­gle­storey of­fice in Chen­nai’s Be­sant Na­gar area. Th­ese an­a­lysts have given M. S. Dhoni’s team ex­clu­sive use of a unique pre­dic­tive engine, the Real- time De­ci­sion Sup­port Sys­tem ( RDSS), which al­lows them to fore­cast a win­ning

score in any given sit­u­a­tion, and pro­vides five dif­fer­ent ways to get to that tar­get. In sim­ple terms, it tells the team how they should tackle each bowler by study­ing pat­terns and pre­dict­ing how they are likely to be­have in the rest of the match.

The re­searchers have found, for in­stance, that Pak­istan off­spin­ner Saeed Aj­mal bowls three doos­ras in al­most ev­ery over, and if no bound­ary has been con­ceded in the first five de­liv­er­ies, the last de­liv­ery is al­most cer­tainly the one that goes away. They have also dis­cov­ered that Sri Lankan pace­man La­sith Malinga bowls three full de­liv­er­ies ev­ery over, and have ad­vised the In­dian bats­men that a short back- lift helps while deal­ing with his toe- crush­ing york­ers. Malinga’s ca­reer stats tell the tale. The fast bowler has an ODI ca­reer econ­omy rate of 5.07 and an aver­age of 26.3 runs per wicket, but against In­dia, Malinga’s econ­omy rate shoots up to 6.02 and aver­age to 42.47. Kohli and Dhoni have a strike rate of 111 and 117 runs per 100 balls re­spec­tively against him.

Dhanan­jaya, who took over as an­a­lyst just when Dhoni was ap­pointed skip­per in 2007, says the In­dian cap­tain keenly uses tech­no­log­i­cal in­puts in all his de­ci­sions. “He lis­tens to what­ever I have to tell him. He then takes a call based on that,” Dhanan­jaya says. Subramanian Ra­makr­ish­nan, the founder of Sports-Me­chan­ics— the firm that also works with the In­ter­na­tional Cricket Coun­cil, Asian Cricket Coun­cil, and the Sri Lankan and Bangladesh cricket boards— says: “Videos and data anal­y­sis was ear­lier used in post- mortem meet­ings but now it’s used in re­al­time. We have also started to marry a play­ers’ gut feel­ing with an an­a­lyt­i­cal layer.”

“For ex­am­ple,” he adds, “If Shikhar Dhawan feels that 150 runs is a par score and our engine says it’s 180, we ask the team which op­tion they want to go for. Then we give the team five ways to ap­proach both par scores. Given the bowler and the op­po­si­tion, we tell them what our sys­tem thinks is the best way to ap­proach the tar­get. Video an­a­lyst is an old term. We are per­for­mance fa­cil­i­ta­tors.”

Video anal­y­sis first started in the late 1990s, when footage of the first and last day of a train­ing camp was com­pared to mon­i­tor im­prove­ments. Soon it de­vel­oped into a day- to- day af­fair where live matches were an­a­lysed to find the weak­ness of op­po­si­tion play­ers. It was in 2007 that the an­a­lyt­ics started to be­come more tac­ti­cal than tech­ni­cal. Team In­dia’s back­end staff started com­ing up with com­plex al­go­rithms that could gauge how an op­po­nent would re­act to a given sit­u­a­tion.

But in­ter­pret­ing vi­su­als and data alone is not suf­fi­cient. The most im­por­tant part is de­liv­er­ing it to play­ers in a man­ner that is ac­ces­si­ble. This is done through apps de­liv­ered through a fiercely pro­tected gate­way to which each player is given a lo­gin and pass­word that is changed fre­quently. “Some play­ers like to try out what we tell them in the nets be­fore agree­ing, and some re­view our sug­ges­tions with their per­sonal coaches,” says Ra­makr­ish­nan.

Video anal­y­sis has come a long way since 2003, when Sachin Ten­dulkar had asked, “What is the guy with a lap­top do­ing in the dress­ing room?” Work has al­ready be­gun on the all- im­por­tant tour of South Africa in Novem­ber. The on­go­ing In­dia A se­ries in South Africa is likely to add cru­cial clues per­tain­ing to con­di­tions. Ask Ra­makr­ish­nan about how Team In­dia is plan­ning to tackle pace spear­heads Dale Steyn and Mornie Morkel, and he gives you noth­ing more than a smile: “It’s clas­si­fied.”

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.