Here is what I typed from the video in Post 22. "SSAC16: How do you Stop Leicester City? Advanced Tactical Analysis in the English Premier League" Page 1: Overview Where did all this start? Opposition analysis using data Leicester City, the not-so-unlikely leaders How to stop the Foxes ¬– and did anyone? Page 2: Where did this all start? Econometric training – and 15 years of living with numbers Predictions for the 2006 World Cup Soccernomics and The Numbers Game Procuring detailed event data from Opta Serving clients including clubs, investment groups, coaches, and software developers Page 3: Opposition analysis using data Virtually all opposition analysis uses data Recorded data offer completeness and (some) objectivity We want to look at tens or hundreds of similar situations A case study: Leicester City's fast attacks from deep Page 4: Opposition analysis using data What is the team's style in attack and defense? Where on the field is the team strongest in attack? Where on the field is the team weakest in defense? Who are the pivotal players? How can we disrupt the things that work for them? Page 5: Leicester City, the not-so-unlikely leaders Well, maybe they looked unlikely... -Bottom of the Premier league halfway through the 2014-15 season -13 points from 19 matches, 0.7 points per match -Low payroll relative to Premier league leaders Page 6: Leicester City, the not-so-unlikely leaders But... -Bad luck had worsened the team's results -Signing Robert Huth strengthened the defense -Finished 14th with 1.5 points per match in second half Page 7: Leicester City, the not-so-unlikely leaders And... -Top analytics department built in part by Rob Mackenzie -Euro 50 million spent on new players over the summer -Hiring of Claudio Ranieri, former coach of Chelsea (finished 2nd in last season) So, take a 5th-place team, a great coach, and Euro 50 million! Page 8: How to stop the Foxes – and did anyone? What was Leicester's attacking style? A basic distinction: direct play versus build-up play Speed to outpace defenders and create advantages Patient passing to lull defenders and find opening Attacks from deep – about 1/3 to 1/2 of open play goals Page 9 had three soccer field diagrams. Page 10: How to stop the Foxes – and did anyone? Measuring end product Shot 1: Direct free kick, 35 yards from goal, 30-degree angle (arrow) 8% chance of scoring Shot 2: Header from corner, 8 yards from goal, 5-degree angle (arrow) 15% chance of scoring "0.08 expected goals (xG) versus 0.15 expected goals (xG)" Page 11: How to stop the Foxes – and did anyone? For each category of possessions, ask two questions: -Is there a shot? -How good a shot is it? Average expected goals across all the possessions: (no shot, 0.13, no shot, no shot, 0.07) (arrow) 0.04 xG average Pages 12, 13, and 14 were graphs. Page 15: How to stop the Foxes – and did anyone? Slowing Leicester down destroys their chances of scoring -Push them to the flanks and make them pass more Who makes their direct attacks go so quickly? Need to look at all attacks, not just those ending in shots Page 16: How to stop the Foxes – and did anyone? Order matters: Who starts attacks in Leicester's own half? First or second pass or dribble of attack: Drinkwater, central midfielder, 206, 12% Schmeichel, goalkeepers, 174, 10% Kante, defensive midfielder, 161, 10% Morgan, central defender, 129, 8% Fuchs, left back, 121, 7% 15 others, 530, 53% (the ratio of count to percentage for the 15 others doesn't match the ratio of count to percentage for the 5 listed players, and the frequency adds up to 1,321) Page 17 is a graph about Drinkwater and Page 18 is a graph about Kante. Page 19: How to stop the Foxes – and did anyone? The most dangerous attacks start through the midfielders -Disrupt both Drinkwater and Kante The Economist published this online on 19 December 2015 What happened to performance in matches 17-28? (Against just slightly tougher opposition...) Page 20 is a repeat of a graph from earlier and Page 21 is a new graph. The presentation went back and forth between Page 20 and Page 21. Page 22: How to stop the Foxes – and did anyone? In matches 17-28, who started the attacks? First or second pass or dribble of attack: Fuchs, left back, 14%, +7% Schmeichel, goalkeeper, 13%, +3% Drinkwater, central midfielder, 11%, -1% Kante, defensive midfielder, 10%, 0% Simpson, right back, 9%, 4% 16 others, 45%, -8% The first percent is the value and the second percent is the value minus the value from the table on Page 16. Page 23: How to stop the Foxes – and did anyone? Tackles, challenges, blocks, and fouls on the midfielders: Matches 1-16: 3.7% of events for Drinkwater and 3.3% of events for Kante Matches 17-28: 4.6% of events for Drinkwater and 4.8% of events for Kante Page 24: How to stop the Foxes – and did anyone? Later opposition pressed Drinkwater and Kante more Drinkwater and Kante were still prolific but less effective Drinkwater missed two matches, both open play xG losses Shots per game from possessions starting in the defensive half fell by 50%; xG from these possessions fell by 25% Pressing has a cost in defending the goal – but worth it! Looking at the changes in shots per game and xG, the shots were better later in the season, meaning that the average xG per shot went up.