- By Adam Julian
It’s November 1999 and the baggage carousel rumbles at Heathrow Airport. Weary All Blacks collect their luggage after the long haul across hemispheres.
Technical advisor Wayne Smith waits anxiously for a meandering 43kg parcel marked Oversize and Fragile to reach him.
Inside two cases, attached to wheels, is ‘Battlestar,’ a computer that will assist in changing the understanding of rugby forever. Its designer is George Serrallach, a Spanish Biotechnology and Biomechanics professor in Palmerston North with a background in football. He eventually concludes that the All Blacks will not win the 1999 Rugby World Cup, the very dream the computer is designed to help achieve.
Battlestar would soon grew like wildfire – the primal beast a seed in the development of AnalySport Limited then Verusco Technologies, a sporting analysis company that at its peak in the mid-2000s employed 55 staff and enjoyed annual turnover of $3 million analysing multiple codes; most critically New Zealand Rugby, including the All Blacks until the end of 2011 when they won the World Cup.
Battlestar pioneered an accurate visual record of a rugby player’s performance. A click on a name would illustrate all actions in a game, like tackles and carries, accompanied with video of those actions. From this was born the language of Key Performance Indicators (KPIs).
Battlestar made no difference in the 1999 World Cup semi-final against France, the All Blacks beaten 43-31. George’s son Xavier Serralach recalls his father’s premonition:
“Three days before the game against France, George was pacing around the garden area of the hotel, obviously stressed. The manager asked him ‘what’s the matter?’ George said, ‘I’m here with the All Black providing analysis. I think they’ll lose. Should I tell them?’ The hotel manager responded, ‘Is that in your brief?’ ‘No,’ George replied.”
Across tactics, trends and individual performance there was no way the All Blacks would win, despite surface appearances to the contrary. Regression rather than progression had set it. How was a Catalan, only recently introduced to rugby, able to uncover such insight?
George Serrallach was born in Barcelona on February 16, 1936. His father José Antonio Serrallach together with his wife, Monserrat Carulla Soler, founded pharmaceutical company Lainco in 1934. Lainco still exists today, turning over millions annually, manufacturing and distributing pharmaceutical products and chemicals for the agricultural industry in the main.
Despite the wealth generated by his parents George hardly enjoyed a luxurious upbringing. The Spanish Civil War and World War II raged from 1936 to 1945 and José was consigned to jail for three years.
From 1946 to 1948 George suffered from tuberculosis. While bed-ridden George’s Dad would project films of football games onto his ceiling. This is where his whole idea of analysis came from; George realised actions were repeated and could be broken down to the finest detail. From 1957 he began to build a film library of European matches which grew into the thousands by the 1980s.
The 50s was a vintage era for FC Barcelona with the winning of nine domestic trophies. The colossal Camp Nou was opened in 1957 and the brilliance of László Kubala inspired a generation of teenagers. Now recovered from his illness, George was a sharp midfielder/striker offered a contract to join the club. He turned it down to study. He later labelled the decision the biggest regret of his life.
Academic expectation and excellence instead took George to Switzerland where he completed his PhD in inorganic chemistry – the topic of his thesis being “Beitrag zur Chemie des Dreivertigen Antimons” which in English is “Contribution to Chemistry of Trivalent Antimony.” It was supervised by a professor who had been nominated for a Nobel Prize.
Restless George was most at peace with the love of his life Josephine Serrallach. He married the elegant and learned daughter of a doctor in 1963. The couple met at a nearby holiday home in Barcelona and George maintained the long-distance relationship by writing poetry.
The birth of three sons, Edward in 1963, Xavier in 1965, and Oscar in 1970, were joyous moments, but a strained relationship with Dad, the political situation of Spain under Franco’s dictatorship and fear of nuclear Armageddon at the height of the Cold War had George searching elsewhere. Devising a system to measure the safest place in the world from nuclear fallout, he arrived in New Zealand in 1971.
Arrival in New Zealand
He quickly secured a job as a Maths teacher at Pukekohe High School before moving to the Dairy Research Institute for eight months followed by a lecturer position at Massey University in Palmerston North in the Biotechnology Department. He taught himself English by listening to National Radio. Josephine and the three boys arrived in 1972.
George immediately involved himself in football coaching; Xavier Serrallach recalls College Street School winning a seven-a-side primary tournament, beating the favourites in the semi-final. Another early success was winning a National Under-11 tournament with Manawatu in Nelson.
Ruahine Association Football Club remains a leading junior football club in the Manawatu, accommodating fledgling talent between the ages of 4 and 13. George was affiliated to the club for at least a decade. The teams he coached were typically successful and mentored in an innovative fashion. Xavier explained that George was developing video analysis and teaching himself computer programming as early as the late 70s.
“He was using video before anyone else. Around 81/82 he’d provide players with a page of stats with a paragraph explaining what was done well and what wasn’t. I can remember filming some of Oscar’s games aboard a cherry picker or from the roofs of nearby buildings.
“He always looked for elevation because filming at ground level doesn’t give you the same perspective. When you watch with elevation you get a panoramic view of the whole field and can also view what those off the ball are doing to influence the game.”
In using statistics and video evidence to measure things like pass quality and outcome of moves, George was able to create an empirically measured library of Basic Individual Techniques (BITs) which increased players’ technical understanding of the sport. Key Performance Indicators (KPIs) is language of the same meaning now commonplace in high performance sport.
The unusual methodology incurred the wrath of officialdom and envious competitors, but football under George was fun. Grating shuttle runs and obnoxious shouting didn’t exist. Instead, every player was supplied a ball and encouraged to develop skill and increase lung capacity. Children were taught how to kick the ball properly with their instep, unusual at the time. Every session included varied control and pass practices, as well as exercises to master dribbling and changing direction, which is exhausting. Council road cones often went missing for these purposes. An emphasis was placed on pressuring the last defender and winning by attacking.
He even convinced the most ardent rugby men to convert. Professor Emeritus Graeme Fraser captained the Otago Colts, coached by the inimitable Charlie Saxton, and had an altercation with the legendary All Black prop Kevin Skinner in a Teachers’ College versus Pirates game in the 50s. Later, the Massey University sociologist was against the 1981 Springboks tour and, after hearing his son Grant abused on the sideline by parents, swapped to football.
“George was the coach of the Manawatu Junior Soccer team which Grant was selected for. On Monday evening after each game the team assembled at George’s house to analyse the game. With his guidance they played more skilfully and intelligently, both individually and as a team, Fraser said.
“I used to carry a step ladder and portable video equipment to games. He developed increasingly sophisticated filming technology for systematically analysing each game and the players therein. It was George’s intellect, integrity and unwavering commitment to enhancing successful planning for each game as well as the performance of the players in their respective positions, that led to the ongoing enhancement of the sport.
“George was also a member of the Manawatu Soccer Board which was a somewhat divided body. It initially showed little respect for George’s superior understanding of soccer. George asked me to accompany and support him during endless hours at George Street and cope with the often derogatory and sarcastic comments from members from the United Kingdom about his more sophisticated and insightful understanding of the ‘beautiful game.’”
Coaching a handful of teams at the same time was a demanding schedule but as a long-time fixtures officer for Manawatu football George was at least able to provide pregame instructions for every side under his stewardship. In the interests of efficiency, he developed a spreadsheet system before Microsoft Excel existed to accelerate the allocation of all fixtures.
In fact, George’s advocacy for youth knew few boundaries. Between 1979 and 1981 he had a term as a city councillor. He’d often receive complaints for his tardy dress appearance at council meetings because he’d show up in dirty football attire. He helped establish the youth council, build skate parks, and wanted the city to have a covered 50-metre swimming pool long before most councillors thought it was a good idea. Eventually a less economical, shorter, version was built.
All Whites, Fate, Politics
By 1985 George had captured national attention. He was appointed All Whites technical advisor and assisted their 1986 World Cup qualifying campaign which started with considerable promise, only to fizzle out in a whimper after his analysis startlingly uncovered two leading midfield players, didn’t pass the ball to each other in three consecutive matches. It later transpired they were having “personal issues.”
Personal issues in Manawatu football remained a constant struggle and it reached a zenith in 1986. Despite coaching the Manawatu Under-15’s to their highest ever placing at the 1985 Nationals he lost out on the Under-16 position to retiring Manawatu Junior Football Association President Brian Mollard in a three-way vote. Mollard had never coached a representative team.
After a brief period coaching juniors at Manawatu Red Sox, George was asked by two veterans of the ailing Freyberg club to coach the senior team in 1988. With the support of Brian Knowles, who owned a piece of land in Awahuri, George developed his own academy and training ground with lights, goals, hills to run on and even a football machine thrower made from two motorbike wheels spinning around, firing out balls.
The Football Academy of Technical Excellence (FATE) taught an estimated 1500 children essential football skills during the school holidays. Players of all abilities were welcome to attend and often showed significant improvement. Typically, 70 players would attend ten two-hour sessions which were videoed and reinforced with an hour of homework. There were four levels of achievement with Level 1 teaching 15 basic skills without pressure and the next three levels introducing pressure and new skills, with the end result being the learning of 100 techniques, development of a greater understanding of options, handling greater pressure, and developing “your game according to your own inventiveness and flair.”
A FATE Manual records:
“Targeted training relies on the statistical technical analysis of games to determine which techniques are mostly used in a game and in which sequence they should be learnt.”
Flair, anaerobic non-lactic fitness, speed and technical ability were the core skills in which programs were built upon. The concept of an option circle was introduced and that is an imaginary line around the player where he is encouraged to think abstractly and quickly about all options at his disposal. It’s worth noting the world’s best player at the time, Maradona, only consistently used about 70 techniques – more than half the time using only 15.
Freyberg grew to become Palmerston North City Rangers which boasted five teams, all coached by George. They changed their clothing from green and white hoops to the red, blue and maroon of Barcelona. Their jerseys were imported from Canterbury Sports Depot where ironically Wayne Smith worked.
Rangers seniors were revitalised under George, winning the first division in Manawatu and then a 1990 Central League Playoff, surpassing teams from Taranaki, Wairarapa, Hawke’s Bay and Whanganui to earn a spot in a much tougher regional competition the next season.
“George was a visionary, 20 years ahead of his time researching and developing programmes about the technical aspects and science of all sporting codes. I guess I embraced his vision and provided some much-needed support to him in those early days where by many he was wrongly frowned upon,” Knowles said.
Perhaps George’s greatest coaching achievement was guiding the Manawatu Under-17 girls to four consecutive North Island titles from 1996 to 1999. In this period Manawatu won 34 out of 35 games despite having 300 less players to choose from than Auckland. George was made a life member of Central Football.
While all this was happening, and after 20,000 hours of work, he had devised a computer system that came to be known as SAM (Statistic Analysis Method). The program devised a specific language, breaking down 2000 movements in the game with graphics showing a field divided into squares. A click of the mouse on a square showed how many times the ball went in the square, another click and you could see who touched the ball, click on one of the players’ names and up comes a list of the skills he used and how often. The computer identified mistakes and was coupled with a video editing system, which enabled an analyst to pick out different aspects of a game being videoed live and store those pieces of the game onto a computer disk. A 2-2 draw between Australia and Iran in 1998 showed Australia struck the ball 1037 times, Iran 896.
Arguably the best player he coached was Julyan Falloon, capped by the All Whites. He played professionally in Switzerland and he would have almost certainly enjoyed a longer career had he not been injured. From 1988 to 1992 Falloon was a prolific striker in the legendary Christchurch United dynasty. In 1992 United became the first New Zealand team since 1975 to win the National Soccer League-Chatham Cup double. Three years earlier he played in a record 7-1 Chatham Cup final victory over Rotorua City. For the past seven years he has been the chief executive of Sport Canterbury. Another All White, Jeremy Brown, and New Zealand cricketer Jacob Oram, a promising goal-keeping prodigy, were other high-profile players coached by George.
Despite his credentials, attempts to sell analysis systems to football clients were fruitless. In his native Barcelona he had to turn down bribes for his product. Everton and Manchester United were briefly interested but passed. Xavier recalls in one meeting in which he also attended a prospective coach spent 40 minutes explaining why he didn’t need analysis; Xavier, laughing, considers that “If you spend that much time defending your methods I think you actually need analysis.”
Introduction to Rugby
Flashback to 1996 and Wayne Smith is at a sports conference in Melbourne run by the Australian Institute of Sport. George was giving a presentation on computerised analysis for football and a fascinated Smith decided to attend.
“There were about three of us there and afterwards everyone left. I went up to him, had a chat and asked, ‘Are you selling this to any soccer clubs?’ George said, ‘I’m from Barcelona and even they don’t want it.’ I then asked, ‘Do you know anything about rugby?’ He said no, but that was the start of it,” Smith recalled.
Brendon Radcliffe first met Wayne Smith when the latter was CEO of Hawke’s Bay Rugby. In the mid-90s Radcliffe became the director of player development for New Zealand rugby, helping write coaching manuals. He laughed at how crude statistical analysis was.
“In 1993 I was assisting John Phillips at Hawke’s Bay counting some basic things like tackles and carries with a clipboard and paper. We had a pretty hard-nosed loose forward Gordon Falcon who later became one of Eddie Jones’ favourite players. After one game I said to Gordie look, here’s your tackle stats. I had a tick for a tackle and three marks for missed tackles. He told me where to stick it and without evidence you had nowhere to go.”
David Hadfield, a Massey University lecturer in coaching, sport psychology, and leadership, has widespread contacts. He saw the value of what George had developed and set out to help him. In 1996 AnalySport was founded, renamed Verusco in 2000.
“I`m not really a numbers man but I understand the power of pertinent data and I’d never heard of anything like this. I understood it had to be computerised and I had some connections that could help,” Hadfield said.
“George had basically got nowhere turning it into a usable product until we met. Turning a string of figures on a page into usable, accessible charts and graphs that coaches could understand and use was an achievement.”
Slowly, software was built for the analysis of rugby. Smith, who assisted, recalled that it was a laborious process.
“None of us knew much about computers. Our roles were to identify tasks within the game that needed to be coded then weighted to make sense of team performance. Ultimately, we came up with around two thousand tasks as George was obsessed by detail. The more information, the better the analysis. For example, did a pass come from the left or right of a receiver? Was it caught on the chest or fumbled. If fumbled did it go forwards or backwards? You can see how, mathematically, this multiplies!”
“It gave you so many lenses to look at performance and that was the brilliant thing about it, Tackle efficiency, which measured the effectiveness of a tackle on a scale of one to ten, was such a powerful statistic especially for loose forwards,” Hadfield continued.
By 1997 there was an almost workable program but when displayed to leading coaches at the Police College in Porirua it was a disaster.
“I thought I’d show Frank Oliver and Graham Henry, the leading coaches at the time. Frank was distracted and Graham lukewarm, but reluctantly agreed,” Smith recalled.
“George was setting up this massive computer and he’d forgotten a cord, jumped up, and hit his head against a beam on the roof and dropped to the floor. I had Graham Henry coming up and here’s George laid out on the floor in the recovery position. Suddenly he jumped up and shouted, ‘Who’s stupid enough to put beams on a ceiling?”
George recovered from open heart surgery and retired from Massey in 1998, devoting himself fulltime to analysis. AnalySport Limited originally started at his home before moving into the University and, as it grew more, a council building in the Palmerston North Square.
By 1998 Smith was coaching the Crusaders. Despite finishing last in the inaugural Super 12 in 1996, under Smith the team won the competition in 1998, with the computer data providing some valuable information other teams didn’t yet possess.
Three years earlier the Springboks won the World Cup, with Cape Town-based technical adviser Rudy Joubert believed to be the first rugby strategist to make use of the DVC 2000 – a computerised video analysis system from Israeli company Orad Hi-Tech Systems. Worth $81,000 and running on a Silicon Graphics workstation, video recordings of games were digitised onto a computer’s hard drive and the operator could link text and video clips.
At the 1999 World Cup the All Blacks, accompanied by George, had an even more advanced system than the DVC 2000, but Smith admits it was so revolutionary the players took time to make sense of it.
Former All Black Captain Graham Mourie, introduced to the technology by Hadfield while earning his Level 3 coaching credentials at Massey University, quickly understood its benefits. Mourie was officially banned from rugby between 1983 and 1995 for violating the rules of amateurism for accepting royalties from sales of his autobiography. Remarkably, he swiftly rose through the national coaching ranks to become Wellington head coach from 1997 to 1999. He was also instrumental in the establishment of the Wellington Rugby Academy.
Doug Neilson is a Performance Analyst for Taranaki and New Zealand Rugby. In 1997 he was a “nomad” working in a pizza shop, playing 80/80s for Poneke, and studying Video and Television Production at the Wellington Polytechnic. He was offered an apprenticeship opportunity to collect video data for Mourie and Wellington. Soon he was working for George.
“I’d drive up to Palmerston North and do these insane shifts while the World Cup was going on,” Neilson recalled.
“To code a whole game on the 45kg laptop it took 24 hours. What he required at the World Cup though was not only All Black games to be coded but the opposition too. We had three of us working around the clock on two computers, so we had a hotbed where we slept for eight hours and we were on the computer for the other 16. I enjoyed it but Health and Safety would have a few issues with it.”
The process also involved rewinding, pausing, fast forwarding and transferring footage from VHS tapes to a hard drive. Neilson also had issues with the family dog, bitten on the backside during an all-night session.
The Wellington Lions’ health improved with Mourie as coach then technical advisor. In 1997 they finished sixth in the NPC. In 1999 they were runners up and a year later, under new head coach Dave Rennie, champions for the first time in 14 years. With 16 new players including All Blacks Christian Cullen, Jerry Collins and Rodney So’oialo key performance indicators were identified and verified for specific positions for the first time. Notes were individually printed on a page using the metaphor of a dart board. The closer each skill was to hitting bullseye, the better the player had executed the skill.
“Previously we had been providing players a VCR copy of games and it would take a couple of hours to view and review with no built in analysis. This reduced that process to twenty minutes” Mourie said.
Mourie and Nielson later teamed up with the Hurricanes, but it wasn’t always smooth sailing.
“The Hurricanes were on their way to Africa and they’d played a pre-season game a day before they departed. I coded the game and when it got to midnight I decided I’d take a rest. When I went to save what I’d done the entire game had disappeared. If there was a power cut or an anomaly with the software it could cause chaos,” Neilson recalled.
In 2000 Wayne Smith became All Blacks coach and New Zealand Rugby signed a licensing agreement with Verusco, which essentially translates to “truth,” to provide analysis. Some of the early discoveries were compelling. Brilliant fullback Christian Cullen gained greater insight into his defence, telling NZPA, “It’s easy to pick out all your faults. For example, you can see the lines you ran coming in if you miss a tackle.” Prop Carl Hoeft gained a greater understanding of his all-round performance, “This system shows everything you do on the field, you can’t hide anymore.”
George even rubbished the statistics published by Sky Television after the All Blacks beat Scotland 48-14 at Eden Park. Sky claimed Scotland had 70 percent of possession and the All Blacks only 30 percent. With a measuring system accurate to 0.01 percent George contested the All Blacks had 49.4 percent of the ball and Scotland 50.6 percent.
Gordon Slater played 144 games for Taranaki from 1989 to 2002, but except for three midweek matches for the All Blacks on the 1997 tour of the UK wasn’t considered for test selection until 2000. Wayne Smith explained why:
“Gordon Slater wasn’t great in the gym, but the data showed he hit 50 or 60 percent more rucks than any other prop in the country. He was strong in relevant areas.”
ProZone and Verusco Growth
Wayne Smith wasn’t retained as All Blacks coach after the 2001 season so joined Brendan Radcliffe at Northampton Saints, staying for three years. Together they were introduced to technology which was to elevate the Verusco product to the next level.
Clive Woodward was coaching England at the time and explained he had a system called ProZone,” Smith said.
“It was a series of feature-recognition cameras set up around Twickenham. Before each test his scientists would scan photographs of every player, the rugby ball, referee, into a computer and then the cameras recognised everyone on the field.
“You’d receive a package at the end of the game presented like a snooker table. It followed every player so you could get GPS type data. With the cue ball you could highlight certain players, put a tale on them and follow them, join them up to observe their defensive alignment. It was phenomenal.”
Unfortunately, it was beyond the Northampton budget until post-game beers and afternoon tea was established for club members, the required money being raised one gold coin donation at a time.
“I left New Zealand with a mind-set that there’s an absolute truth in the game. There’s no such thing as an absolute truth. The requirement of a coach and a player is to be open-minded enough to receive truths and be intelligent enough to put that picture together yourself,” Radcliffe said.
“When I was writing coaching manuals I was adding layers of complexity for older players when in fact the higher you go the more it becomes about fundamental skills. It’s a time and space constraint, not complexity. Video analysis helped me grasp that.”
In 2011, with Wayne Smith, Graham Henry, Richie McCaw and Daniel Carter as start-up investors, Radcliffe founded The Rugby Site, where the world’s best players and coaches share their ideas in an innovative audio-visual platform. There for example exists footage of Richie McCaw doing what he thinks an openside should do and comparing it with Sam Warburton.
George was equally taken by ProZone but didn’t have the technology to implement a similar system, so he simply employed more staff. Verusco suddenly had more than 30 part-time and full-time staff and by 2005 were coding 190 matches a year, including 60 internationals. Mervyn Dykes captured the typical working environment.
“Eight coders are each given a 10-minute segment of a game and take four to five hours to work their way through it, classifying 86 “tasks” per minute of running rugby to a total of over 4000 tasks per game. A task could be a team function such as a lineout or scrum, or a player’s role like support runs or tackles. Each task is further described by “attributes.” For example, a tackle entry could read “Tackle, Richie McCaw, effective, right shoulder, front on, advantage line as well as describing field position.”
The amount of detail provided might seem excessive but Andrew Sullivan, an analyst with the Crusaders from 2001 to 2008 and the All Blacks from 2002 to 2007, provided an insight with Tracey Nelson on June 23, 2007 at just how convenient it was for the players.
“The players spent surprisingly little time reviewing their performance because they can do it so easily that in a matter of 10 minutes you’ve viewed all of your tackles. Take Richie (McCaw) for example, if he’s made 10 tackles he’ll take a couple of minutes over each one. Probably they spend a half hour to an hour a couple of times a week. Plus, they’ll also go with the specialist coaches for scrums and lineouts.”
Sullivan later worked alongside Robbie Deans (another early student of Verusco technology) with the Wallabies between 2008 and 2014.
Mohi Marsh was a Manawatu Under-19 and Under-21 representative flanker who started working for Verusco in 2002 while studying an Information Systems degree at UCOL. He would later become an IT Administrator assisting in licensing and technical support but started life as a coder.
“It was the perfect student job really. You had regular shift work at times you weren’t studying. The office was open plan with offices for George, software developer Stacy Waugh and a few others at the heart of the company,” he reflected.
“It took me about six months to become really adept at coding. We were split into pods and typically there would be a chief coder overseeing everyone. It was always a really busy environment because the quicker you could deliver the information the better it was for the coaches.”
By 2006 Verusco was supplying analysis for 11 of the 14 Super Rugby franchises. The TryMaker software had two versions, the Pro Elite which was coded by experts in Palmerston North and the Plus version where users could code themselves. Such was the sensitivity of the information Verusco didn’t provide their physical address but did promote a phone number and email.
Xavier Serrallach, an experienced coder, was in charge of human relations, and Josephine came onto the Board of Directors to help with the company’s financial governance. She had written a PhD investigating the production, distribution, consumption and economics of the wine industry in New Zealand and worked as an Intellectual Property Manager and Commercialisation Manager, first at Massey University and subsequently at Victoria University of Wellington, and was an expert in copyright laws.
Hadfield resigned as a director and offloaded his share in 2006, reflecting that “George was a clever man” but became “increasingly difficult to deal with.” Smith and Mourie described Hadfield’s involvement as “vital,” but the extent of his professional and financial contribution has been challenged by the Serrallach family. *
Today Hadfield runs his own consultancy service, MindPlus, which focuses on mental skills training, business coaching and high-performance coaching. He has worked with most national rugby teams and the Toronto Blue Jays Major League baseball organisation.
Heading into the 2007 World Cup quarter final against France the All Blacks had won 42 of their last 47 tests backed up by the cutting-edge technology of Verusco. Still George was concerned. His analysis showed the All Blacks’ performance had been dipping and the discrepancy between what referee Wayne Barnes was calling and how the All Blacks were playing was too great for them to win.
The result was an infamous 20-18 victory to France. Notwithstanding the controversial forward pass which led to a try for Frédéric Michalak, France, the most penalized team at the 2007 World Cup, received 85 percent of the penalties and free kicks in the game (11-2). The All Blacks had possession for 65 per cent of the match. An irony is that during that World Cup, 69 per cent of all penalties went to the team in possession.
According to Graham Henry, the referee missed 40 penalties. Leading referee Colin Hawke suggested 16. What happens when the analysis is telling you what you don’t want to hear?
“I didn’t give a lot of credence to blanket predictions but I was interested in looking into KPIs which give you clues as to how you’re tracking,” Smith observed.
“The quarter final came down to the wire. We had a couple of decisions go against us, and didn’t take the drop goal which we were instructing the boys to do, but essentially we did what we had always done and the penalty always came. Against France it didn’t.
“At one point we had 18 rucks near the posts and independent referees from around the world observing afterwards said there were seven penalties that could have been called.”
Dr Ken Quarrie is the Senior Scientist for New Zealand Rugby, a member of three World Rugby Advisory Groups, and a Research Associate at Auckland University. In 2000 he was appointed the inaugural Injury Prevention Manager for New Zealand Rugby.
He’s an expert in SAS, Statistical Analysis System, an integrated system of software products which enables programmers to perform information retrieval and data management, report writing and graphics, statistical analysis, econometrics and data mining.
Naturally George and Ken saw eye to eye, with Ken often a mediator in arguments between George and other New Zealand Rugby officials. Despite his main focus being health and safety, Quarrie became a vital cog on the performance side of the ledger in partnership with Verusco.
A 2011 study in the Journal of Science and Medicine for Sport tracked the positional demands of international rugby players. Actions of 763 players were coded from video recordings of 90 international matches played by the All Blacks, providing understanding of the differences in movement patterns, contact loads and activities between positional groups, improving the understanding of the need for different conditioning and recovery programmes for the different forwards and backs. By 2013 this art was perhaps best mastered when the All Blacks won all 14 tests and outscored opponents 183-53 in the last half an hour, illustrating the effectiveness of their bench and optimizing performance time with realistic workloads and recovery schedules.
A similar study was undertaken to elevate the performance of international goal kickers. A sample of 582 international matches played from 2002 to 2011 examined and ranked kickers not on percentage alone but accounted for factors such as the venue, the position on the field from which kicks were taken, the relative scores of the teams at the time at which the kick was taken, and the amount of time remaining in the match. Overall, 72% of the 6769 kick attempts were successful. Forty-five percent of points scored during the matches resulted from goal kicks, and in 5.7% of the matches the result of the match hinged on the outcome of a kick attempt. Daniel Carter was the top-ranked kicker in 2005 and 2006. The All Blacks only lost twice in this period.
Michael Lauren is a Senior Scientist and Operations Analyst for the New Zealand Defence Force. The fusion of scientific minds even led to the creation of ALFRED – Agent Laboratory for Rugby Experimentation and Demonstration.
Agent based modelling is for simulating the actions and interactions of autonomous agents in order to understand the behaviour of a system and what governs its outcomes. It’s commonly used in the military and was used to explore moves which occur in a single phase, usually following a set piece such as a scrum or lineout.
An example of a move that was simulated using ALFRED came during the tour of New Zealand by England in 2008. Experiments with ALFRED suggested that a hole would appear in the defence if the defending team’s loose forwards were either decoyed by the half-back’s run, or slow to come off the back of the scrum. In such cases, the left-wing running on an angle that changes the direction of the attack would find a hole, and only needs to link up with the first five-eighth (fly-half) for a try to be scored.
It transpired exactly to plan with halfback Jimmy Cowan passing to winger Sitiveni Sivivatu who distributed to first-five Daniel Carter to score under the sticks.
2011 World Cup
Once again the All Blacks were favourites to win the World Cup in 2011 but the pressure of hosting the tournament at home would amplify any shortcomings. France in the final proved to be a stubborn opponent, but the All Blacks had extracted a simple but exhaustive clue on how to improve their chances for success.
“We decided by 2011 that getting off the ground was one of our most important KPIs. Guys who’d made a tackle, been a ball carrier, needed to be quickly off the ground and back into the game. When we won the World Cup in 2011 we were typically 40% quick off the ground that is back to our feet in under three seconds. In the final you can hear Andy Ellis yelling the whole time ‘get back to your feet, get back to your feet.’ We were 64% off the ground which was the best we’d ever seen and miles better than anyone else. Now you’d use 80% as a yardstick. That’s how data can change the game.”
In defence for most of the last quarter the All Blacks won 8-7.
“By 2011 we realised we needed to match the patriotic intensity of other countries whilst maintaining clarity of thought so that we could execute better. That led to us taking our Red head to Blue head psychology out of the classroom and onto the training field. We put the players under intense pressure at times to test their ability to stay ice cold and calculate under the most extreme pressure.
“Then, in 2015, given our pool experiences in 2007 we decided to play full contact games amongst ourselves on Tuesdays leading into games on the weekend. This hardened us for the playoffs but often meant we weren’t at our best in pool games against teams like Namibia and Georgia. They would play like their lives depended on it and did their countries proud whilst we were slightly below our best, often due to a bit of training fatigue and much to the consternation of our supporters back home. The plan was a brilliant one as we were able to fly from the quarter final onwards.”
George himself was fond of the saying, “When you plan for something, you plan for the worst possible scenario, so you never get a surprise – you’re prepared for everything.”
The All Blacks won 123 of 148 (83%) test matches from 2000 to 2011 about 20 per cent better than their major rivals during their time with Verusco.
Unfortunately, New Zealand Rugby cut the Verusco contract in 2012 due to financial constraints. Its true analysis was a labour-intensive exercise with many matches costing between $1500 to $2000 to analyse. In 2007 the company coded 286 games. Another factor was that Apple had superior camera technology, though Verusco was catching up. However, many in the know claimed the quality of analysis suffered when Verusco departed.
Without the New Zealand Rugby work Verusco was unable to retain most of their staff. Josephine suffered from cancer in 2012 and sold the intellectual property to Sport Universal Process (S.U.P.) which was a group comprising ProZone Sports Ltd, Amisco and TS Media, based in Nice (France) and Leeds (UK).
Daniel Carter was George’s favourite rugby player. In the 2005 Lions test where Carter scored 33 points he performed 197 actions and only two of them were mistakes.
Football remained his greater passion. His favourite players were Barcelona’s Iniesta and Xavi who worked well with Messi giving him an ideal platform to take his game to another level. He wrote a novel, The Merry Go Round, about the life of a football player in the English league. In 2010 he developed Football Skills 3D, an early training software for children on Apple iPad and iPhones.
What was George like to live with?
“George could never relax. He’d take a computer to the beach and he would complain when there was sand on it,” Josephine laughed.
“He was interested in the environment and a member of the Values Party, a precursor to the Greens. He was interested in travel, science fiction, stars, and energy. He was generous, smart, passionate, determined and trusting. We were lucky to travel to Easter Island, the Galapagos Islands and Antarctica together. We were also following lunar and solar eclipses wherever the phenomena appeared in the world.”
Father, teacher, coach, entrepreneur and inventor George passed away aged 76 on Tuesday August 14, 2012. His funeral at Beauchamp Crematorium Chapel was attended by hundreds.
The current analysis provider for New Zealand Rugby is Hudl, founded in 2006 by University of Nebraska trio David Graff, John Wirtz and Brian Kaiser. The product started as an analysis tool for the football team and is now used by more than 160,000 teams around the world. The company employs 2,200 staff. The All Blacks have two touring analysts in addition to support staff often utilised to study multiple angles. Two drones frequently capture footage of Crusaders training at Rugby Park, Christchurch, while unions like Canterbury, Wellington and North Harbour use Hudl to share footage of senior club and representative fixtures which in turn provides ideas for coaching.
Stacy Waugh and Doug Neilson remain leaders in the field of analysis with the latter observing.
“A lot of the coding can be done by artificial intelligence now with football leading the way. Rugby is a bit more complex because you have the merging of groups in scrums and lineouts so the computer has to figure out who’s who. You can measure how far apart players’ feet are, stick figures exist with data running down the screen measuring where they are in relation to the ball and each other. You can see the technique of the halfback from any angle in 3D when he passes the ball. Coaches I’ve worked with recently are using Dictaphones with the GPS to see what players are saying in the heat of the moment.
“AI will reduce people’s workload but I think you’ll always need people. Right now I can sit at home with access to three of four cameras and code a game from my laptop. I’ve always joked with the players that eventually they won’t be needed and it’ll be the analysts at home playing our war games against each other.”
Has the analysis of the game made it less exciting?
“KPIs provide clues as to how you’re tracking. Computer analysis backs up your intuition and helps you see something different in a player every day. Every coach in the world uses digitised footage and stats. The secret to being a great coach is extracting those gold nuggets and translating them into performance,” Smith responded.
Doug Neilson believes analysis is responsible for removing much overt violence from the game and making it safer. **
George’s oldest son Edward is a bicycle mechanic/engineer and runs Planet Bike in Rotorua, a cycling tourism business. Oscar is a medical doctor in Byron Bay who has published ground-breaking research on postnatal depression. Xavier assists his brother Oscar and remains very interested in sport. Josephine has worked on causes advancing human rights and climate change and keeps busy cooking, exercising and reading.
Wayne ‘“Techno” Smith pays tribute to George:
“A Spanish professor who loved Barcelona football and knew nothing about our game revolutionised rugby. Computerised analysis is the single biggest advance in rugby during my time in the game, and that is the legacy of George Serrallach.”
*In investigating the history of George Serrallach’s contribution to analysis in sport, I discovered that people had very different opinions about who contributed to the development of AnalySport and Verusco and the extent to which they were involved over what periods. I interviewed members of George’s family, along with David Hadfield, Graham Mourie, Wayne Smith, Doug Neilson, Ken Quarrie and Mohi Marsh. There is no single, unambiguous version of “the truth” of George’s story that I can present that will satisfy all parties. I can state that all those whom I interviewed were generous with their time, and I have no reason to doubt that what they told me is what they genuinely hold to be what happened. As a journalist who wasn’t a first-hand witness to the events, I don’t, and can’t, know what actually transpired. All I can do is follow the well-beaten path of historians; to present the various versions of the events as they were told to me, and to refer to documentary evidence where I have been able to source it. David Hadfield said he owned half of the company. This is disputed by George’s family. It is clear that Hadfield had a long involvement and that should be acknowledged as well as his long and successful track record in high performance sport. Smith and Radcliffe conceded George could be “exhausting” and “frustrating” to work with and his passion and determination for his work often came at the expense of social subtleties. What is clear is that George was a unique and exceptional intellectual, software developer and visionary whose industry and entrepreneurship changed New Zealand sport for the better. There were many people who contributed to further developments of an original vision that was George’s. Illustrating how those ideas changed sport and rugby particularly is the key focus of this work.
**RugbySmart, the brainchild of Ken Quarrie, is a programme to reduce the incidence of permanently disabling injuries resulting from rugby. Since 2000, there has been an 89% reduction in scrum-related spinal injuries, and a reduction of 56% in all injuries resulting in permanent disablement. The success of RugbySmart has seen it copied in a number of countries around the world, including South Africa, Australia and Japan. A study undertaken with the support of Verusco analysed 140,249 tackles in 434 professional matches coded from video recordings for height and direction of tackle on the ball carrier, speed of tackler, and speed of ball carrier; injuries were coded for various characteristics, including whether the tackler or ball carrier required replacement or only on-field assessment. Injuries were most frequently the result of high or middle tackles from the front or side, but rate of injury per tackle was higher for tackles from behind than from the front or side. Ball carriers were at highest risk from tackles to the head-neck region, whereas tacklers were most at risk when making low tackles. The impact of the tackle was the most common cause of injury, and the head was the most common site.