Expected Goals The story of how data conq - Rory Smith

• Ashley Flores is a footballer in the Philippines who also works as a “tagger” for a German data analysis company called Impect.

• Impect employs over 100 “taggers” in Manila who watch football matches and input data by clicking keyboard shortcuts every time something happens on the pitch.

• The taggers track metrics like “packing” - how many opposition players are bypassed by a pass or dribble - and the amount of defensive pressure on the player with the ball.

• This granular event data provides insights that traditional stats like shots and passes do not. Impect uses it to analyze matches for European clubs like Bayern Munich and PSG.

• Tagging is not a highly paid job but provides a decent living for Flores. He feels grateful to Impect for staying open and allowing staff to keep working during the pandemic.

• New taggers are trained by analyzing Germany’s 7-1 World Cup semi-final win over Brazil in 2014, which demonstrated the value of Impect’s advanced data over normal stats.

• Impect is a German company that specializes in analyzing football data. It employs over 100 “taggers” in the Philippines who watch footage of matches and tag various metrics like number of defenders, ball pressure, etc.

• The taggers go through extensive training to learn how to accurately tag the metrics. The data is then checked for quality control before being uploaded into Impect’s systems.

• Impect analyzes matches from many leagues around the world based on client requests. The data is used by teams, coaches, executives, scouts, journalists and fans for things like recruitment, tactics, performance analysis, etc.

• The story about Michael Edwards illustrates how football was initially skeptical of data analysis. Edwards was asked if the “computer” predicted a win rather than looking at his actual scouting report.

• Football has historically been resistant to data analysis compared to baseball, seeing itself as too fluid and passionate to be boiled down to numbers.

• But in the late 1990s and 2000s, data analysis began to take off in football due to increased money in the game, professionalization, and influence from academics and other fields. Data now impacts virtually every aspect of the sport.

Here are the key points from the summarized passage:

• Football’s early experiments with data were limited by lack of technology to collect and analyze it. But improvements in equipment enabled the gathering and assessing of data at high speeds.

• The 1990s changed how fans thought about football, with the sport becoming more gentrified and intellectualized. Media coverage expanded and books/video games introduced ideas of statistically analyzing players.

• This receptiveness to data was driven by changing fan demographics and growing up simulating the sport through video games. Numbers helped provide content.

• The data revolution in football has happened quietly, unlike in baseball’s public struggle. Most clubs now use data in recruitment, coaching, and strategy but keep it private.

• Data is now a core part of football but its possibilities are still being explored. While some clubs lead in using data, emotional instincts still override it.

• The speed of football’s data transformation has been remarkable. In 25 years it’s gone from skepticism to having a thriving analytics industry. This book charts that unspoken revolution.

• Ram Mylvaganam did not originally intend to revolutionize football when he started his company Prozone. He was simply trying to sell high-end massage chairs.

• However, in the process of demonstrating the chairs, he realized the technology could be used to track player movement and analyze performance in football matches. This sparked the idea for Prozone.

• Prozone pioneered a new way to digitally capture and analyze football matches. They placed cameras around the stadium focused on the pitch and tracked all player movement, allowing detailed statistical and visual analysis after the fact.

• This digital tracking and analysis of football matches created an entirely new industry and ignited an analytics revolution in the sport. It allowed much more in-depth evaluation of tactics, fitness, and player performance.

• Prozone and the other analytics firms that followed gave clubs new tools to gain an edge. It intensified the arms race as teams tried to leverage data and analysis to improve results and gain a competitive advantage.

• The analytics revolution has fundamentally changed football, from youth academies to the biggest clubs and leagues. It altered how the game is played, coached, managed, and evaluated at the highest levels.

• Ram Mylvaganam and Prozone kicked this off simply from trying to demonstrate massage chairs. Mylvaganam did not set out to revolutionize football, but that’s exactly what he ended up doing through the creation of Prozone.

• Mylvaganam, a Sri Lankan businessman with little interest in football, came up with the idea for a company called ProZone after noticing the inefficiencies in how coaches like Steve McClaren analyzed video footage. ProZone aimed to provide tailored video analysis to coaches.

• Though it seemed innovative, data analysis in football was not new - Charles Reep had been annotating every event in matches since the 1930s to try to understand the game objectively. He came up with many foundational concepts like ‘match performance analysis’ and ‘on-the-ball events’.

• Ideas that seem innovative in football often aren’t - they are repurposed old concepts. Data analysis is an example, with Reep’s work predating ProZone by decades.

• Mylvaganam’s outsider perspective allowed him to identify an opportunity in football analytics that those embedded in the game’s traditions may have missed. Though not wholly new, he brought a business acumen to adapt an old idea to the modern game.

• Charles Reep was an accountant who pioneered the use of data and analytics in football in the 1950s. He collected data on goals scored and passing sequences, finding that most goals came from moves of 3 passes or fewer.

• Reep’s work was decades ahead of its time and showed the potential for data analysis in football tactics and strategy. However, his findings were misinterpreted to advocate a simplistic “long ball” playing style.

• The FA’s Charles Hughes claimed to independently reach similar conclusions as Reep, advocating direct play and “Positions of Maximum Opportunity” to get the ball forward quickly.

• Hughes had a powerful role at the FA and oversaw a generation of English football focused on “kick and rush” tactics rather than skill or possession. This was blamed for holding back the England team.

• In contrast, managers like Valeriy Lobanovskyi at Dynamo Kyiv used data and technology more holistically to great success.

• Reep and analytics were unfairly tarnished by association with Hughes’ flawed application of them. It set back the acceptance of data analysis in English football for decades.

• Valeriy Lobanovskyi, a Ukrainian football manager, was an early pioneer in using data and statistics to improve his teams’ performance. He focused on metrics like sprints completed, shots taken, and pressing.

• In the West, early uses of data were associated with direct, basic styles of play like kick-and-rush. This created a perception that data reduces the artistry of football.

• Opta Index was founded in 1996 to create a player rating system. They assigned points for actions like passes and tackles. The resulting “Game Scores” were aggregated into a single number meant to represent player performance.

• The scores lacked nuance in comparing players across positions. But newspapers still used the underlying Opta data as fantasy football grew popular and fans wanted more statistics.

• When Duncan Alexander joined Opta in 2000, they were still using rudimentary methods like annotating videos and overnight computer processing. But the company was growing rapidly along with public demand for more detailed soccer data.

• Opta Sports pioneered detailed football statistics and made them widely available to media for free, seeing it as a publicity exercise. This led to a surge in data-driven sports reporting.

• Opta did not focus on selling data to clubs at first. Some managers like Jim Smith used it, but most were not that interested. Opta did not have the credibility yet to convince clubs data was useful.

• ProZone saw the opportunity to sell data to clubs. It started working with Derby County for free, then got a trial with Manchester United after Steve McClaren joined.

• After United won the treble in 1999 using ProZone data, its credibility was established. ProZone then signed up several major clubs as paying customers, establishing itself in the market.

• The key difference was Opta focused on media while ProZone targeted clubs. ProZone overcame club skepticism by getting backing from a top manager in Alex Ferguson, showing data could help teams win.

Here is a summary of the key points in the passage:

• ProZone was founded in 1998 by software entrepreneur John Coulson to provide data and video analysis to football clubs. It grew rapidly as managers like David Moyes saw its potential to track player performance.

• ProZone installed cameras at stadiums and analysts at clubs to collect data and produce reports. Analysts like Gavin Wilson learned on the job how to present data in useful ways to managers.

• The analysis focused initially on physical data like distance covered, which was familiar to managers. Over time it expanded to tactical analysis with the rise of digital technology like DVDs.

• Analysts like Michael Edwards used digital editing to produce video clips for coaches and players illustrating strengths/weaknesses. This allowed more visual presentation of data insights.

• The shift from analog to digital technology allowed more powerful use of ProZone data and accelerated its adoption in football clubs. The data moved from just physical metrics to more tactical and technical analysis.

• Jaap Stam’s autobiography revealed several dressing room secrets at Manchester United, including unflattering nicknames for Gary and Phil Neville used by teammates. This likely contributed to Alex Ferguson selling Stam soon after.

• However, Ferguson and assistant Steve McClaren actually decided to sell Stam primarily because data showed he had lost pace and was making fewer tackles, signs of declining performance. In hindsight, this data was misinterpreted.

• Sam Allardyce at Bolton was an early pioneer of using data analytics in the late 1990s and early 2000s to guide tactics and transfers. He had several analysts interpret data to find key statistical correlations that would give Bolton advantages.

• Allardyce’s “Fantastic Four” insights from data analysis formed the core principles of his tactics: clean sheets, scoring first, high speed running, and set pieces. Allardyce used data to determine what types of set pieces and player movements were most effective.

• Allardyce also used analytics to identify transfer targets like Gary Speed who appeared risky on the surface but data showed were still performing at a high level.

• Chris Anderson was a respected Ivy League political scientist who suddenly became obsessed with soccer. His wife and kids found it amusing that he was blogging about soccer statistics instead of his usual intellectual pursuits.

• Anderson had worked hard to achieve his prestigious academic career after a modest upbringing in Germany as the child of a divorced American GI and German mother. His academic identity and status were very important to him.

• The sudden soccer obsession seemed out of character for the serious academic. His wife was familiar with sociologist Robert Merton’s ideas about unintended consequences - she likely saw Anderson’s new hobby as one of these.

• There seemed no reason for his wife to see Anderson’s hobby as anything more than a benign midlife crisis. But his soccer obsession would lead Anderson down an unexpected path that would transform soccer analytics.

• Chris Anderson was an academic who became interested in football data after reading the book Moneyball about data analytics in baseball. He started researching football data but found very little academic literature on it.

• He started a blog in 2009 to share the football data insights he was finding, joining a small community of football data enthusiasts. However, they were limited by a lack of publicly available data compared to US sports.

• In 2011, Anderson attended the MIT Sloan Sports Analytics Conference where football analytics had a very small presence, showing how far behind it lagged other sports.

• Anderson, a Cornell professor with no sports business experience, was there as an academic outsider curious to learn more. The conference exposed him to the growing field of sports analytics and started him on a path that would lead him to give up academia to pursue a career revolutionizing football analytics.

• Companies like Opta and ProZone had performance data that football enthusiasts wanted, but were hesitant to give it away for free.

• Anderson noticed through his blog that there was a demand for football analytics, especially among younger and North American fans more used to stats in sports.

• At the 2011 Sloan Sports Analytics Conference, Anderson met football analysts who felt stifled and ignored within their clubs, lacking colleagues who understood analytics. The conference confirmed for Anderson the growing interest in football analytics.

• Scott McLachlan was one of the few technical/data scouts in English football at the time, working for Fulham. He attended Sloan to stay current on sports analytics advances.

• At Sloan, Anderson met McLachlan and they quickly bonded over their shared interest in advancing football analytics. McLachlan provided the football expertise while Anderson brought the analytical mindset.

• Their meeting highlighted the gap between football’s traditional evaluation methods and the potential of analytics, presenting an opportunity for those interested in merging the two approaches.

• Chris Anderson was an Ivy League professor who started a football blog as a hobby. Through it, he made connections in the football world, including being invited to visit Fulham’s training ground by tactical analyst James McLachlan.

• Anderson produced an expected goals analysis for Fulham, but they were unsure what to make of it. This showed Anderson he needed to become an ‘insider’ in football to be taken seriously. The analysts interested in his work had no power to implement it.

• Anderson recruited his academic friend David Sally, who had a business background, to help turn his ideas into a viable product for football clubs.

• Anderson felt uneasy leaving his academic career but intellectually couldn’t resist the chance to run a ‘real world experiment’ in football, like Billy Beane had done in baseball. The potential reward of overturning conventional wisdom was too appealing.

• Together Anderson and Sally worked to develop commercial data analytics products to sell to top clubs, hoping to leverage the analytics revolution started by Beane and the Oakland A’s in baseball.

Here is a professional summary of the key points:

Sally and Anderson wrote their book The Numbers Game to showcase how data analytics could revolutionize football. Though it generated interest, the football establishment did not take them seriously as outsiders. The duo realized they needed an internal position in a club to implement their ideas. With no budget for analysis, their only option was buying a team themselves or finding an owner who shared their vision. This led them to pursue ownership despite lacking financial resources. Their motivation was proving data analytics could work in football, not just writing academic theories. Though initially absurd, they grew determined to impose an analytics revolution from the top down if no club would adopt one organically.

Here is a summary of the relevant information:

Anderson decided to pursue his football data analytics dream by leaving his academic career behind, moving his family to London, and trying to build a consortium to buy a football club. He took a year sabbatical from Cornell, knowing that if he couldn’t make progress in 12 months, he would return to his previous career path. His colleague Sally helped pitch the idea to Mark Cuban, who was interested but not willing to lead the effort. Anderson realized that to fully pursue the vision, he would need to own a club rather than just advise one.

Meanwhile, Henry Stott, a statistician with no interest in football, created a simple statistical model to predict World Cup results for an office pool. Though basic, it correctly predicted Senegal’s upset of France in the 2002 tournament. A radio producer heard about Stott’s success and had him appear on a show as a novelty guest predictor.

• Billy Beane, general manager of the Oakland A’s baseball team, was intrigued by the frenzied media coverage of Premier League football while on vacation in London in 2003. He saw parallels between baseball and football in terms of inefficiency and opportunity.

• Beane believed the data-driven approach he used in baseball could also work in football. At the time, football clubs had some data but weren’t using it strategically. The only football analytics he found were in a newspaper column by Henry Stott.

• In 2006, Rupert Murdoch held a corporate retreat attended by top News Corp executives including Robert Thomson and Danny Finkelstein of The Times.

• Finkelstein had previously connected with Stott and convinced The Times to publish some of Stott’s analytical predictions, such as Senegal having a 25% chance to beat defending champions France at the 2002 World Cup. This established The Times as innovative and insightful.

• After the success of publishing Stott’s analytics, The Times wanted more regular analytical content. This led Finkelstein to start the weekly “Game Theory” column analyzing data and tactics.

Here is a summary of the key points in the passage:

• In the 1990s, football broadcasting began including more statistics and data onscreen during games. The comedians Baddiel and Skinner parodied obsessive football stats fans with their character ‘Statto’ on Fantasy Football League in 1994.

• In 2002, Finkelstein and Stott launched a stats column called ‘The Fink Tank’ in The Times newspaper. It used data analysis to test conventional wisdom about football.

• Finkelstein hired physicist Ian Graham in 2005 to refine the statistical models behind the column. Their insights were rejected by Sky Sports, who felt stats diminished the emotion of football.

• Fink Tank’s work resembled the ‘Moneyball’ approach described in Michael Lewis’ book. Finkelstein met Billy Beane and was put in touch with a football executive interested in using data - Damien Comolli of Tottenham Hotspur.

• Comolli had used data services like ProZone and Opta in his previous roles. Appointed as Tottenham’s first technical director in 2005, he wanted to use data analytics to help the club compete with richer rivals on a budget.

• Damien Comolli became director of football at Tottenham Hotspur in 2005. He wanted to understand how the top teams in the Premier League won games in order to help Spurs compete.

• By analyzing match data from Opta, Comolli found the top teams had more possession and passes in the opposition half, leading to more shots and goals. This gave him an idea of what Spurs needed to improve.

• Comolli read Moneyball and was inspired by the Oakland A’s use of data analytics to compete against wealthier baseball teams. He contacted Billy Beane and they met at the 2006 World Cup.

• Beane introduced Comolli to Decision Technology, a company focused on sports analytics. Comolli brought them in as consultants and later signed an exclusive deal, integrating their data models into Spurs’ processes.

• Decision Technology analyzed all data they could access, initially focused on team performance. In 2006, Opta introduced new real-time data collection at the World Cup, allowing more detailed player-level analysis.

• Comolli used the data and models from Decision Technology and Opta to guide recruitment and strategy at Spurs as they aimed to break into the top four teams in the Premier League.

• Damien Comolli brought data analytics company Decision Technology to Tottenham in 2007. They provided advanced data and models to help with player recruitment and retention. This was pioneering work at the time.

• Decision Technology’s models used Opta’s event data to assess players’ impact and contribution. They developed concepts like expected goals years before the term became widely known.

• Comolli left Spurs in 2008 but his data-driven approach laid groundwork for future success. Decision Technology stayed until 2018.

• Comolli later had a stint at Liverpool where he championed signing Jordan Henderson among others. Liverpool’s current analytics team has links back to Comolli’s work.

• In 2019, Liverpool and Spurs met in the Champions League final. Comolli’s fingerprints were still visible at both clubs despite having left years earlier. His data focus and signings played a role in their eventual success.

• Though Comolli’s Spurs spell was seen as a failure at the time, Decision Technology believes their decade with the club aligns precisely with its rise into the elite. Comolli helped set Spurs on course to become top 6 challengers and eventual UCL finalists.

• Stott and Comolli believe the work of Decision Technology helped transform Tottenham into a top 4 club despite limited finances, but this revolution went largely unnoticed.

• Decision Technology provided weekly reports on Tottenham’s performance and players for a decade, but received little feedback from the club and felt their input was often ignored.

• Stott believes they were too early - in 2006 football culture wasn’t ready to fully embrace data analytics. They needed buy-in from managers, not just the chairman.

• Lesson is ‘being early is the same as being wrong’ - ideas need the right timing and culture to take off. Football wasn’t ready to break century-old orthodoxies and give up power to data experts.

• For the revolution to happen, data needs to be central and have control, not an optional extra. This was an epiphany Anderson later had with football analytics.

• Timing and cultural acceptance are as key as the substance of the ideas. Football still needed its ‘Moneyball’ moment of complete buy-in to analytics.

• Chris Anderson and David Sally, two Ivy League professors with ideas for using data analytics to revolutionize football, pitched their concept to a roomful of wealthy potential investors arranged by David Blitzer.

• Anderson and Sally felt out of place and nervous in the luxurious boardroom, more comfortable in academia than finance. This meeting was the culmination of years of work networking and pitching their idea.

• London hotels like the Landmark, Dorchester, and Savoy serve as meeting spots for football’s power brokers to conduct business deals and transfers. Mayfair is the heart of football’s financial engine.

• Anderson moved his family to London and spent months pitching the idea to connect with someone who could buy a club and implement his data-driven approach from the top down. He met resistance from the bottom up.

• After many meetings he honed his pitch: football had inefficiencies and outdated thinking that could be revolutionized through data analytics. He wanted to buy a club and make it “smart.”

• English football was amidst a decade of takeovers by foreign billionaires, started by Roman Abramovich buying Chelsea in 2003. Anderson aimed to ride this wave of new owners open to new ideas.

• In the 2000s, there was a gold rush of foreign investors buying up English football clubs, attracted by huge TV revenues and the Premier League’s openness to overseas owners. clubs like Liverpool, Manchester United, and Manchester City were bought by American, Russian, and Middle Eastern owners.

• Many shady brokers and fixers emerged, trying to insert themselves between wealthy buyers and clubs to make commissions. The world was opaque with uncertainty over who was real and who was not.

• Kevin Anderson, a sports lawyer, was pitched by an Arab-American broker named Tareq Hawasli about a Saudi prince interested in buying a club. After months of delays, Anderson finally met Prince Abdullah bin Musaid Al Saud in London to present his idea for running a club sustainably.

• The presentation seemed to go well, though Anderson still had doubts about the legitimacy of it all. In the end, Prince Abdullah did go on to buy Sheffield United, showing it was real after all. But Anderson found the experience disorienting, filled with uncertainty and chaos as he tried to navigate this shadowy football investment world.

• Anderson and Sally posed as tourists to get access to various football clubs in England, including Sheffield United, Aston Villa, and Everton. They wanted to research potential investment opportunities by seeing the stadiums and facilities first-hand.

• Using academic research methods, they analyzed data on the 92 professional clubs to narrow down the options based on location, finances, stadium size, brand value, etc. They focused on London and the North West of England.

• After applying their criteria, they ranked the shortlist of around 20 clubs. Leeds United, Nottingham Forest, and Derby County stood out but were not in the targeted locations.

• Weighting their search towards London, one club stood out - Charlton Athletic. Anderson had already visited and seen potential despite it not being a famous club. It was in a good location and fit many of their ideal criteria.

• They were doing this research to identify a specific club they could recommend buying when they found investors. They wanted to provide a concrete proposal, not just a hypothetical plan. Their aim was to eventually help a club into the lucrative Premier League.

• Anderson and Sally identified Charlton Athletic as an ideal club to purchase and transform using data and analytics. Charlton had fallen from the Premier League but had a good stadium and was located in an attractive area of London.

• They pitched the idea to a group of investors assembled by David Blitzer, who already had sports investments like the Philadelphia 76ers.

• Their presentation focused on using data and analytics to challenge conventional wisdom in football and gain a competitive advantage. They highlighted previous football innovators like Ajax and Barcelona.

• The investors extensively questioned and challenged Anderson and Sally’s ideas during the presentation.

• The core concept was that football is currently making poor decisions that data analytics could dramatically improve. This would then allow the club to get promoted, establish itself in the lucrative Premier League, and be sold at a profit within 7 years.

• Detailed financial projections mapped out the potential outcomes, with the ideal scenario promoting to the Premier League and then selling at a profit after 7 years in 2020.

In summary, Anderson and Sally pitched the investors on using data analytics to transform Charlton Athletic into a Premier League club that could then be sold for a significant profit within 7 years. Their presentation faced many challenges but focused on the financial rewards of revolutionizing decision-making in football.

• On a flight back from Arsenal’s preseason tour, Hendrik Almstadt pitched to Arsene Wenger the idea of acquiring StatDNA, a data analytics company, to help improve the club’s player recruitment. He pointed out poor past signings like Chamakh and Park Chu-Young as examples of mistakes analytics could help avoid.

• Almstadt believed data’s most useful application was in highlighting what not to do - avoiding “stupid stuff” in player transactions. Wenger, renowned for being frugal, agreed to the £2 million expenditure after being convinced it could prevent wasteful spending on subpar players.

• Separately, Arsenal fan and software engineer Sarah Rudd wanted to use data to determine if the common criticism that Arsenal “always try to walk it in” rather than shooting more often held weight. She scraped data from seasons of Arsenal games using Opta files and analyzed the numbers.

• Her analysis found Arsenal did tend to pass more in the final third compared to rivals like Manchester United. However, this was not because they were overelaborate, but due to their style depending on relentless possession and building sustained pressure.

• Rudd shared her findings with the Arsenal analytics team, helping demonstrate the potential value of passionate fans with technical expertise working with data to provide novel insights clubs may lack internally.

• Lucy Ward Rudd worked at Microsoft improving search engine relevance using big data techniques. She was interested in applying these techniques to football analytics but was told there was not enough data available.

• She started collecting what limited data she could find online, like from the Guardian’s chalkboards feature. This convinced her football analytics was worth pursuing. She quit Microsoft to focus on getting a job in football analytics.

• Jaeson Rosenfeld was a consultant who saw potential in providing detailed football data as a business opportunity after reading Moneyball.

• He founded StatDNA in 2009 and outsourced labor-intensive data collection to Laos to create very detailed datasets cost-effectively.

• StatDNA’s process involved recording games, sending them to Laos, having them coded by trained staff spending up to 40 hours per game, then sending the data back to clients. This produced more detailed data than rivals.

• The key points are the origins of Rudd’s and Rosenfeld’s interest in football analytics, Rudd collecting data herself to prove feasibility, and Rosenfeld founding StatDNA using an outsourcing model to provide very detailed data cost-effectively.

• StatDNA was a sports analytics company that provided detailed data and analysis to teams and clubs. They offered time-stamped event data that allowed insights into things like how much pressure was on the ball during a cross.

• In 2011, StatDNA held a research competition to attract more analysts. Sarah Rudd entered and impressed them with her expected possession value analysis. StatDNA hired her, and revealed they were working with Arsenal.

• Arsenal signing Mesut Özil for £43.5 million in 2013 was seen by Billy Beane as an example of a ‘Moneyball’ signing - paying a premium for a world-class player who filled a specific need.

• Hendrik Almstadt, an Arsenal executive, started playing with Opta data as a hobby and realized it could help with recruitment. He advocated exploring it more to Ivan Gazidis.

• Gazidis, with his American sports background, was open to analytics. Arsenal were behind clubs like Chelsea and Manchester City in adopting data analysis, but few clubs were openly discussing it at the time.

• Almstadt was actively opposed to analytics at first, but came around after seeing baseball and basketball teams use data successfully.

• Wenger was open to analytics, having predicted the impact of ProZone years before. He invited Almstadt to present his findings, and they got along well, speaking German together. Wenger supported Almstadt’s work.

• At Arsenal’s 2014 AGM, fans complained about the £2 million spent on StatDNA rather than players. Gazidis defended it as critical for scouting, preparation, and tactics.

• StatDNA’s analysis was much more advanced than Opta/ProZone. Their insights infused Arsenal’s operations in scouting, prep, post-match review, tactics, set pieces, shot selection, squad management, salaries, etc.

• Some at Arsenal resisted the analytics, so Almstadt held workshops to explain how data supported rather than replaced them. He used player profiles to illustrate the benefits.

• Reactions were mixed - some staff were receptive, but Almstadt found resistance from some of Wenger’s technical team. Overall, StatDNA’s data became deeply integrated into Arsenal’s processes.

• Jaeson Rosenfeld’s analytics company StatDNA was acquired by Arsenal in 2012 to help modernize their player recruitment and analysis.

• Several Arsenal staff, including Arsene Wenger and his assistant Steve Bould, embraced the new data-driven approach. But some traditional scouts resisted, seeing it as a threat to their jobs and expertise.

• StatDNA analyst Sarah Rudd gained access to extensive new data at Arsenal, allowing the company to do advanced analytics on tactics, player performances, etc.

• StatDNA was involved in Arsenal’s transfers as early as 2012, though this was not publicized. Analysts believe they prevented some poor signings.

• However, Arsenal did not see the success on the pitch that was hoped for. Reasons include the drawn-out end of Wenger’s tenure, limits on spending, and resistance from some at the club.

• StatDNA had a head start on analytics in football, but that advantage has now closed as data use has become more widespread. Arsenal could not fully capitalize on the opportunity.

In summary, Arsenal made an early investment in advanced analytics via StatDNA, but struggled to implement the insights fully due to internal resistance and the club’s overall direction. The promise of the data revolution went largely unfulfilled.

• David Sally and Chris Anderson, two academics interested in using data analytics to revolutionize soccer, met with Keith Harris, a prominent British banker and soccer financier, to discuss buying a soccer club to test out their theories.

• Harris was the perfect contact - he had been involved in many high-profile soccer club acquisitions, like Chelsea and Manchester City, and was very well connected in the world of soccer finance. His reputation gave him credibility and trustworthiness.

• Though Harris was not totally convinced about the potential of data analytics in soccer, he liked Sally and Anderson enough to offer to help them buy a club using his connections and knowledge of the soccer business.

• Harris could not directly sell them a club since he did not own any, but he could connect them with opportunities and guide them through the process using his extensive network and experience.

• Getting Harris on board gave Sally and Anderson more credibility and opened doors, as he was a “central node” in soccer finance networks. Though data analytics was new to soccer, Harris’s involvement gave their ambitions more feasibility.

• Anderson and Sally had an idea to use data analytics to buy and run a football club, but needed financing and expertise to make it happen.

• They recruited football agent Keith Harris, who was skeptical that outsiders could succeed in the opaque world of football business.

• Buying a football club is very different from a normal business acquisition - it’s an unregulated marketplace driven by rumors and personal connections rather than clear processes.

• Anderson and Sally spent over a year meeting potential financiers and intermediaries, but struggled to find serious backers or clubs clearly for sale.

• The football business world was full of vague promises and opaque motivations from agents and fixers claiming insider access.

• Despite Harris’ warnings about the challenges outsiders face, Anderson and Sally persevered in trying to find financing and a club to purchase.

• Anderson and Sally were part of a group trying to buy a football club to run an experiment using data analytics. They targeted Charlton Athletic but found the process complex and opaque, with uncertainty over who really owned the club. After months of cautious negotiations they seemed close to a deal but it ultimately collapsed due to a “cultural misunderstanding”.

• They received other offers from clubs like Reading and Everton but the most compelling was from Aston Villa’s owner Randy Lerner. He seemed like the ideal owner - wealthy, experienced in US sports, and passionate about Villa’s history and community. But after 5 years his enthusiasm had waned as the club underperformed. He was ready to sell and the group felt they could restore Villa’s fortunes.

• However, the negotiations were slow and secretive. There were rival bidders and suspicion from Villa executives towards analytics. After months of talks and due diligence the deal fell apart over discrepancies in the club’s finances. Again cultural differences and secrecy hindered the process.

• Anderson and Sally were left frustrated, feeling data analytics could transform clubs’ performance. But the world of football was complex and chaotic, with secretive negotiations and cultural misunderstandings hampering their efforts to buy a club.

• Randy Lerner had been a popular owner at Aston Villa, spending heavily on transfers between 2006-2010. But Villa consistently finished 6th, just missing out on Champions League qualification, frustrating Lerner.

• As funds dried up, Villa declined under a series of managers. Lerner wanted to sell the club by 2014.

• Josh Harris and David Blitzer, part of a U.S. investment group, entered talks to buy Villa. A deal seemed close until star striker Christian Benteke suffered an Achilles injury.

• The injury made the U.S. group hesitant about the price. Lerner refused to accept a lower contingency price if Benteke didn’t recover. Talks broke down again.

• After a year of fruitless efforts to buy a club, Chris Anderson was ready to give up and return to the U.S. But his wife Kathleen urged him to keep trying.

• With no deal in sight, Anderson quit his university job to focus on the search. But as more deals collapsed, doubts crept in about whether his dream would ever materialize.

• Expected Goals (xG) appeared fleetingly on Match of the Day, representing a watershed moment in the mainstream acceptance of advanced analytics in football coverage.

• Match of the Day editor Richard Hughes had been considering using xG for a couple of years, seeing it as a way to add an analytical layer to the show’s analysis.

• Opta representatives explained xG to the pundits, though some were skeptical. Hughes knew it had to be introduced delicately given the show’s format and potential for pundit backlash.

• There was precedent for analytics being mocked on TV, such as with ITV’s failed “Tactics Truck” segment in the 2000s. Hughes wanted to avoid a repeat.

• The best approach was introducing xG subtly without much fanfare to allow it to establish itself over time. Match of the Day presenting it matter-of-factly represented a major milestone for analytics penetrating the mainstream.

• Match of the Day began including Expected Goals (xG) metrics in its analysis of Premier League matches in the 2017 season. This represented a breakthrough for analytics into the mainstream football coverage.

• Match of the Day had to cater to a diverse audience, from hardcore fans to casual viewers. So the xG data was presented subtly without fanfare.

• Other shows like Monday Night Football on Sky Sports also began using xG. Pundits like Jamie Carragher welcomed it as a way to avoid seeming ‘old school’.

• Opta developed xG models around 2012 to move beyond just providing data to media, and offer more analytical insights to clubs.

• Opta hired Sam Green, a math and physics graduate, as their first chief statistician to mine insights from their data.

• Green developed xG models by analyzing huge volumes of shot data and correlating shot attributes like distance and angle to scoring probability. This quantified the value of chances created.

• The simplicity of xG made it intuitively appealing and helped it gain traction after years of analytics being dismissed in football. Seeing it on Match of the Day signaled it breaking into the mainstream.

• Sam Green at Opta developed the concept of Expected Goals (xG), a way to measure the quality of chances in football. xG assigns a value between 0 and 1 to chances based on how likely they are to be scored.

• xG provided a more objective way to assess whether a team created good chances, beyond just looking at shots. Green’s version of xG gained popularity and acceptance, helping bring analytics more into the mainstream.

• At the same time, figures like Simon Wilson at Manchester City recognized football was lagging behind other sports in analytics. Wilson helped City invest heavily in analytics, but still felt outside help was needed.

• Wilson devised a plan to publicly release City’s data from the 2011-12 season to spur innovation. This ‘mcfcanalytics’ initiative failed to spark a revolution of open data, but did help identify new talent and ideas.

• Releasing the data allowed amateurs and academics to access it, accelerating football analytics and leading to new innovations like expected goals.

• Manchester City’s release of its data in 2012 was a ‘course-changing move’ that allowed the analytics community to demonstrate what could be done with football data. It sparked interest across the sport in analytics.

• In the years after, many major clubs hired data specialists, either from within football or from outside fields like economics and statistics. Consultancies also emerged to provide data services.

• Analytics also began taking root on the continent, especially in Germany and the Netherlands. Clubs like Bayer Leverkusen brought in outside experts and built internal analytics teams.

• Key figures included Jonas Boldt, who embraced data-driven recruitment at Leverkusen, and Sven Mislintat, who developed his own analytics platform at Borussia Dortmund.

• Red Bull’s clubs were early adopters, with Leipzig’s Johannes Spors recruiting Bastian Quentmeier to build their analytics department in 2016.

• The growth of analytics across Europe’s major leagues showed the transformative impact of Manchester City’s data release, bringing the ‘Moneyball’ approach to football and sparking a new era in the sport.

• Christian Quentmeier was working at Opta before being recruited by RB Leipzig to build their data analysis system from scratch. Ralf Rangnick and others at Leipzig were impressed with his work and saw him as a pioneering data scout.

• Sam Green worked at Opta before being hired by Aston Villa as their first Head of Research in 2015. But the manager who hired him was fired right before Green started, and replaced by Tim Sherwood who did not value analytics.

• Some analytics principles like avoiding low percentage long shots were gaining traction in football, but the acceptance was mostly silent rather than public endorsement.

• At Villa, Green suggested good transfer targets based on data like Hakim Ziyech and Joe Gomez, but some of his actual signings did not work out as well, in part due to Villa’s struggles as a club.

• Green had a strained relationship with Sherwood, who did not believe in analytics. Another Villa analytics hire, Hendrik Almstadt, also clashed with Sherwood and said he told them to “fuck off” when they tried to help signings acclimate.

• Both Green and Almstadt left Villa within about a year, showing the challenges data analysts faced gaining acceptance and authority at clubs in those early Premier League days.

• Chris Anderson became chief executive of Coventry City with big ideas but little experience actually running a football club or business. The cramped quarters of the club’s modest training ground confronted him with the vivid realities of running a real football club.

• Despite his intelligence and self-assurance, Anderson was aware he had much to learn about the practical day-to-day operations of a club.

• He was quickly inundated with requests from club staff seeking his approval and sign-off for various decisions and expenses, highlighting how hands-on and demanding the job would be.

• Anderson realized he would need to learn on the job under immense pressure, implementing his ideas for change while managing the club’s ongoing challenges.

• His theories about using data and analytics to transform football would now be tested in the real world, away from abstract models and spreadsheets. Coventry would be the proving ground for his unconventional ideas.

The summary highlights Anderson’s transition from theorist to football club chief executive, and the eye-opening experience of moving from conceptual ideas to hands-on management. It emphasizes how he was confronted with the vivid realities and demands of running a real club, and would need to rapidly learn and adapt while trying to implement transformational changes.

• Mark Anderson started as CEO of Coventry City Football Club with limited funds available from the owners SISU. Many staff came to him requesting small amounts of money for supplies, but he had no discretionary budget.

• Relations between the club and SISU were strained, with fans and city officials disliking the owners. Anderson tried advocating for more funds but SISU refused.

• Manager Tony Mowbray and technical director Mark Venus were initially suspicious of Anderson as an outsider. But he worked to gain their trust by not interfering and bonding over shared interests like football tactics.

• Venus had assembled a talented squad on a tight budget, including future Premier League players. Anderson let him continue recruitment unimpeded.

• Similarly, Anderson did not try to change Mowbray’s coaching methods. They built rapport discussing aspects of the game itself, not just club operations.

• Gradually Mowbray opened up, seeing Anderson’s passion for football. Their relationship moved beyond mutual wariness to trust and cooperation.

After a few months at Coventry City, Anderson realized that the football world was not as alien to him as he initially thought. Though empowering Mowbray and Venus was partly out of necessity due to his lack of time, it did not represent a major change in how he approached his job. Most of Anderson’s time was spent on administrative tasks and negotiations rather than directly involved in coaching or transfers. The club’s tight finances meant transfers were opportunistic, like signing veteran Joe Cole on a free transfer.

Selling players like James Maddison was also driven by financial necessity rather than inclination. Anderson spent most of his time on menial tasks like finding sponsors, meeting with fans, negotiating stadium leases, attending away games, and providing content to the local newspaper. With little budget, he could only tinker with incorporating data, hiring a part-time student analyst and another enthusiast. They produced weekly reports analyzing performances, but Anderson could not devote much time to his passion for data.

At the club’s first Christmas party in years that Anderson paid for, he felt part of something but knew the feeling wouldn’t last. Despite limited time and resources, he still pursued his goal of incorporating data into the club’s decisions, viewing it as a way to improve regardless of the level. The opposition was not as entrenched as he expected; the challenge was fitting it into their constrained reality. Though not cutting edge, Anderson tried to subtly help Coventry get the most out of its resources by making data part of existing processes.

• Anderson realized that empirical truths from data analytics posed a threat to the power of football’s traditional “hereditary in-group” which passed down ideas from generation to generation. Data democratized knowledge and threatened their “protected knowledge.”

• At Coventry, Anderson did not see overt hostility to using data, just a lack of time - the coaches were focused on immediate needs rather than theoretical long-term improvements.

• Midseason data showed Coventry were overperforming and would regress to midtable. Anderson struggled with how to communicate this without dampening morale. He realized the value of intangibles like belief and hope.

• Coventry did regress as the data predicted. Though vindicated, Anderson took no solace in being right. Data is only useful if communicated properly to benefit those on the ground.

• Anderson cherished glimpses of the privileged insider view but knew Coventry could never become a truly data-driven club on its budget. To properly test his ideas required full control and commitment to analytics from top to bottom.

• Though unable to transform Coventry, Anderson’s time there taught him football is not just a numbers game but also about emotion, communication and interface between data and people.

• John W. Henry bought Liverpool FC in 2010 after a court battle with the previous owners. As a new owner unfamiliar with football, he wanted to transform Liverpool into a data-driven club like he had done with the Boston Red Sox baseball team.

• Henry had made his fortune trading futures and was a billionaire. He was a partner with Tom Werner, who had produced major TV shows. Together they owned the Red Sox, which they turned into a successful, analytics-focused team, ending an 85-year World Series title drought.

• Henry tried to hire Billy Beane from the data-driven Oakland A’s baseball team for the Red Sox. When that failed, he hired a young Yale graduate as GM and used analytics to help the Red Sox win World Series titles.

• Henry planned to bring a similar data-driven approach to Liverpool FC. He believed data could help identify talents and gain an edge over rivals. In November 2010 he met with two analytics firms at Dilke House to pitch working with Liverpool.

• The two firms were Prozone, which collected physical performance data, and StatDNA, a statistical analysis firm started by Ian Graham. Henry was impressed and wanted to hire them.

• The move showed Henry’s commitment to using data and analytics at Liverpool, following the successful model he had implemented with the Red Sox baseball team.

• John Henry, owner of the Boston Red Sox, bought Liverpool FC in 2010 and wanted to apply “Moneyball” principles of using data analytics to gain an edge.

• He hired Damien Comolli, who believed in the power of data, as Director of Football Strategy. This was a relatively new role in English football at the time.

• Comolli arranged a meeting between Henry and Henry Stott of Decision Technology, a pioneer in football analytics. Henry wanted to acquire Decision Technology for Liverpool but they were under contract with Tottenham.

• Unable to acquire Decision Technology, Liverpool still wanted their expertise and knowledge in football analytics.

• FC Midtjylland in Denmark is considered a “Moneyball” club that relies heavily on data analytics. Their chairman Rasmus Ankersen devised a “Table of Justice” league table based on analytics rather than actual results.

• Midtjylland shares an owner with Brentford, another data-driven club. The use of analytics by clubs like these can be traced to owners who believe in the power of data.

• Benham and Bloom built successful gambling consultancies, SmartOdds and Starlizard, by using data analysis to find inefficiencies in the football betting market. Their firms assess many factors to generate probabilities and advise high-rolling clients on where odds are mispriced.

• Though described as professional gamblers, their businesses operate on a large scale to minimize risk. Both worked with the statisticians behind the Dixon-Coles model that inspired their approach.

• Bloom also owns Brighton FC, which shares principles with Starlizard. The club often signs players overlooked by others but whose data profiles suggest quality.

• Benham rejects “Moneyball” as a label, arguing it misrepresents using statistics scientifically to predict outcomes. His clubs use data but also seek other edges like nutrition, injury prevention, and sleep science.

• Brentford scrapped its academy to avoid developing players for richer clubs to poach. Its B team focuses on late bloomers cut from other academies.

• At Midtjylland, Ankersen sees a “laboratory” to test ideas to play more efficiently, meaning effectively rather than negatively. As with Moneyball, data helps rethink tactics and strategy.

• Rasmus Ankersen, co-owner of Danish club FC Midtjylland, was determined to make football more analytical and data-driven. Midtjylland focused on improving set pieces, realizing they accounted for many goals. They designed a detailed set-piece playbook and became the most effective set-piece team in Europe.

• Midtjylland also analyzed that crossing and throw-ins were inefficient, so they changed tactics to improve these areas. They brought in specialists to help implement the changes. The club used data to inform all decisions - from player recruitment to half-time tactical changes.

• For analytics to go mainstream in football, an undisputedly successful big club had to adopt it wholeheartedly. Liverpool, under new ownership of data-driven Fenway Sports Group, became that club.

• Michael Edwards, Liverpool’s technical director, advocated using data analytics but manager Brendan Rodgers preferred traditional scouting for transfers. A dispute over signing Roberto Firmino (backed by data) versus Christian Benteke (preferred by Rodgers) highlighted the analytics divide.

• Behind the scenes, Liverpool was transforming into a data-centric club under Edwards. But initially there was no clear strategy, just a loose commitment to using data. For analytics to take hold, Liverpool needed a defined vision and firm implementation.

• When Fenway Sports Group took over Liverpool in 2010, they hired Damien Comolli to overhaul the club’s approach to player recruitment and analytics. But after mixed results, Comolli was fired in 2012.

• FSG doubled down on analytics by hiring Ian Graham from Decision Technology to build a data science team from scratch. But Graham said it would take 6 months just to acquire the necessary data and build models.

• In the meantime, Liverpool struggled on the pitch under Kenny Dalglish and then Brendan Rodgers as manager. The “transfer committee” approach to signings was criticized as overly data-driven.

• A clash emerged between Rodgers, who favored traditional scouting, and FSG, who wanted a more analytics-based approach. Rodgers wanted Christian Benteke while FSG preferred Roberto Firmino.

• They compromised by signing both, but the truce didn’t last. Rodgers barely played Firmino while the team struggled. Rodgers was fired in 2015, leaving FSG at a crossroads over whether to persist with analytics or change course.

• After Brendan Rodgers was fired as Liverpool manager in 2015, the club’s owners realized the problem was not their transfer committee model but rather that Rodgers had not fully bought into it. They resolved that the next manager must be willing to work within the club’s structure.

• The club was confident in its use of data and analytics to identify transfer targets, as evidenced by their faith in signing Roberto Firmino despite his struggles under Rodgers. Liverpool believed their models provided a complete picture of a player’s ability.

• When hiring Rodgers’ replacement, Liverpool quickly chose Jurgen Klopp over Carlo Ancelotti. Despite Klopp’s struggles in his final season at Dortmund, Liverpool knew from their data analysis that Dortmund’s decline was due to bad luck, not a drop in performance.

• Ian Graham met with Klopp early on to prove the analytics department’s insight. Though the details may be somewhat exaggerated, Klopp was open-minded about analytics and collaborative decision-making. This fit Liverpool’s model of having experts in different areas.

• Klopp surrounded himself with specialized staff like fitness coaches and nutritionists. He delegated to those with more expertise, rather than pretending to know everything himself. This matched Liverpool’s philosophy of finding any possible edge.

• The book came out in 2021 and details how Liverpool FC has integrated data and analytics into their operations under manager Jurgen Klopp.

• Ian Graham, head of Liverpool’s Research Department, does not work out of the training facility but him and his team of physicists and academics embrace being part of the club culture, including having breakfast in the player’s canteen.

• Graham’s department focuses on models like Expected Possession Value to quantify player contributions and guide recruitment and strategy. Many of Liverpool’s star signings under Klopp, like Mohamed Salah, were driven by Graham’s data models.

• Liverpool blends data with traditional scouting and expertise but the club’s belief in analytics has been key to their recent success. Graham and the Research Department are fully embedded in Liverpool’s processes but they don’t publicize details to maintain a competitive advantage.

• There is no inherent tension between a player being expensive yet undervalued. Liverpool’s record fees for Van Dijk and Alisson proved to be wise investments. The ‘Moneyball’ concept is not just about finding cheap players, but accurately evaluating a player’s worth to a specific team.

• Under Klopp, Liverpool embraced analytics and data, led by Edwards and his team. This data-driven approach, along with Klopp’s coaching, has brought tremendous success - Premier League title, Champions League, other trophies.

• Liverpool is now seen as a model for how to effectively incorporate data analytics into a club’s operations. Other clubs look to emulate their approach. Liverpool’s success has convinced English football to take analytics more seriously.

• The impending departure of Edwards was seen not as a relief but as a crisis, showing how pivotal his analytical work had been. Liverpool is now a standard-bearer for the use of data in football.

• The rise of outsider analysts like teenager Ashwin Raman shows how clubs are now tapping into the public analytics community to find talent, valuing their skills over traditional credentials. Liverpool’s success has sparked a analytics revolution across football.

• Raman Singh started a football analytics blog as a teenager in India, gaining attention for his insightful analysis. He was hired part-time by Scottish club Dundee United to do data analysis, helping with recruitment and scouting. After a couple of years, he took a break to focus on college.

• The rise of data analytics has opened doors in football to people like Raman - scientists, statisticians, and hobbyists who previously had limited opportunities. Clubs are now hiring physicists, mathematicians, and other specialists for data science roles.

• Monchi, sporting director at Sevilla, has a reputation as a transfer market guru, unearthing talent and selling players on at a profit. Despite his traditional approach, Monchi has embraced analytics, with Sevilla investing in AI and data to stay competitive. He sees it as complementary to his experienced judgment.

• Data analytics is making football smarter and more global, with clubs able to identify and recruit talented analysts from anywhere. The game is opened up to bright minds who previously were locked out for not having an elite playing career. Analytics provides another lens to identify quality and value in the transfer market.

• Data and analytics have become increasingly prevalent in football over recent years, with the vast majority of clubs now incorporating them in some way into their player recruitment strategies. They use data platforms like Wyscout and InStat to identify potential transfer targets.

• Clubs are also using data to determine which clubs and systems would best suit the style of players they are looking to sign. Kai Havertz had his agents commission a data study to assess which club he would fit best at before signing for Chelsea.

• However, traditional scouting and relationship building remain crucial in transfers. Data has made the recruitment landscape more competitive by eroding informational advantages clubs used to have.

• Early data analysis suggested corners were inefficient chances, but teams like Liverpool and Brentford have used data to make their set-pieces more effective, seeing them as an opportunity for easy improvements.

• Data does not necessarily lead to simplistic long-ball football, as evidenced by stylistically-focused possession teams like Liverpool, Brighton and Brentford being data leaders. The game is influenced by many factors beyond just data analytics.

• Football analytics and data usage has increased dramatically in the last 20 years, influencing how teams play, recruit, and make decisions. However, some feel progress has been slow and football still has a long way to go compared to sports like baseball.

• Tactics like counter-pressing and playing wingers on the ‘wrong’ side have become normalized through data showing their effectiveness. Coaches like Guardiola have promoted these tactics philosophically, but data has also pragmatically proven they work.

• Many analysts feel football is still in the very early stages of effectively using data. Football’s “data sucks” compared to other sports, with few teams doing anything truly useful with it. Secrecy between clubs also hinders progress.

• Reasons for slow progress include resistance from managers and clubs just paying lip service to analytics. Analysts have felt dismissed, struggling for recognition, which colors their perspective.

• Better data collection by companies like StatsBomb shows promise for future progress. Demand from top clubs is growing rapidly. However, football likely still only scratched the surface of what analytics can do.

• Event data from games is valuable for teams to analyze player and team performance, but ‘broadcast tracking’ data from TV broadcasts would be even more useful. It would allow tracking of off-ball movement and provide more widespread access to spatial tracking data.

• With more data available, the key differentiator between teams will be who can best analyze and model the data. Teams are hiring mathematicians, physicists, and engineers to gain an edge.

• Analytics and data science are now much more accepted in football compared to 10-15 years ago. Conceptually, the value of analytics has been proven. Football analytics features prominently now in conferences like Sloan.

• Many influential figures in football such as club executives and owners are now ‘outsiders’ rather than traditional football people. This openness has allowed the rise of analytics.

• The author, Chris Anderson, has also gone from an outsider to becoming an advisor on football analytics, showing the changes in attitudes. Football is now more open to different perspectives and expertise.

Here is a summary of the key points from that passage:

• The author thanks the many people who shared their stories and insights that made this book possible, naming several individuals specifically.

• Particular thanks are given to Lukas Keppler and Leo Lachmuth for answering many emails and introducing the author to Ashley Flores.

• There is an acknowledgment that this telling of the game’s recent history is not comprehensive and apologies to those not given proper prominence.

• Gratitude is expressed to various journalists, authors, and sources who provided background information.

• Special thanks are given to agent David Luxton for encouraging the project and introducing the author to Chris Anderson and David Sally, authors of The Numbers Game.

• Sincere appreciation is offered to Chris Anderson for sharing his experiences and story over the years.

• Finally, heartfelt thanks are given to family and friends for their patience and support during the writing process.