The Analytics Problem In Football
The case for you to follow my stats page on Twitter.
Why does every football conversation around data have to be either too detailed or too vague? Why is every stat either too surface level in what it expresses or too deep that it confuses the reader? Why are casuals excluded from being stat heads?
Why is the football community so against numbers?
A lack of a middle ground is the problem, in Basketball stats exist in 3 levels, on the 1st level stats are traditional stats such as points, rebounds and steals. They center around volume, after that are 2nd level stats which are meant to provide a more detailed analysis on how the 1st level stats came to be e.g Field Goal Percentage, True Shooting Percentage and Rebound Rate, they center around efficiency. the 3rd and final level (for now) is 3rd level stats which look to create a statistical vacuum through which every player can be measured equally against each other no matter the era or level of opposition (lol, what a jump) e.g. PER in basketball but I would rather use WAR as the less popular but more detailed example, Wins Above Replacement is a 3rd level baseball stat that allows baseball writers to compare a player from 1908 with a player from 2019 by comparing the value(total contributions) a player brought to his team over the league average value/contribution for his position, these stats center around determining value by looking beyond volume and providing more detail on these players efficiency and how their efficiency matches up against their peers.
In football the 1st level is the most common with goals, assists, tackles, blocks, interceptions and clean sheets being popular ways to assess a player but for some odd reason we’ve jumped straight to 3rd level stats to provide more detail on a player and that’s what causing mass confusion, that’s where the disconnect is. I can sit down with my nephew and have him understand shot conversion rate to some degree, I think he might start crying when I’m 2 minutes into explaining the intricacies of Post-Shot Expected Goals. Where are the 2nd level stats in football? They exist but no one has taken the time to explore it. Until today. Layman Analytics will be centered around providing detail as to how players get their goals, assists, tackles, interceptions, blocks and clean sheets, it’ll prize efficiency over everything and will cover 2nd level stats mixed with 3rd level stats.
I love Xg but explaining to someone how it works always leaves them looking at you with a confused Phil Jones face, they eventually get it but it is a mission to get there, layman Analytics will be Different. You’ll see it and immediately understand what the data means and will instantly get a deeper look on what the traditional numbers are saying.
Conversion Rate centres around everyone’s favourite stat goals. Goals are the most popular way of measuring an offensive players value but as stats flood football we’ve realized they are not the best way to measure this value. “You can get the conversion rate by dividing the number of goals by the number of shots. Easy-peasy. Then you can multiply by 100 and you get the percentage conversion rate,” this stat helps paint the picture of who is the more elite level finisher, if a player gets 10 shots on target how many does he score?
To demonstrate True efficiency lets use last year’s Premier League Golden Boot Race, the traditional 1st level stat was goals through which Aubameyang, Salah and Mane tied so they shared the honour, equals right? The efficiency charts tell another story, the traditional stat doesn’t tell you Mane is the only one who scored all 22 of his goal from open play, the efficiency charts will tell you Mane and Aubameyang took a Similar amount of shots(88 to 89) while Salah took 135. Of the 135 shots Salah took he only converted 19 into goals(excl pens) for a 14% shooting conversion rate, Aubameyang was 2nd with 18 of his 89 shots being goals for a CR% of 20% while Mane led the crew with a 25% CR but this doesn’t paint the full picture, the true measure of their finishing ability is shown in their Goals Per Shots on Target %. Of the 42 shots Mane had on target he scored 22 of them for a SOT-CR of 52%, Aubameyang only had 34 shots on target but converted 18 of those 34 into goals for a SOT-CR of 53% and Salah converted 19 of his 41 shots into goals for a 32% TCR.
Using Basketball per 100 as a reference let’s simplify all those above numbers by looking at it as per 10 shots how many shots are being converted to goals for the previous Premier League season;
Mane 25% SCR/ 52% TRC – 2.5 goals per 10 shots, 5.2 goals per 10 shots on target.
Aubameyang 20% SCR/ 53%TRC – 2.0 goals per 10 shots, 5.3 goals per 10 shots on target.
Salah 14% SCR/ 32% TRC – 1.4 goals per 10 shots, 3.2 goals per 10 shots on target.
A few things stand out, 1st a separation is clear, 2nd Aubameyang is lagging behind Mane on the Goals Per 10 shots but is the more elite finisher when measuring shots that go on target and is 3rd Mane who is averaging a goal more per 10 shots than Salah, and 2 more goals per 10 shots on target. Might need to look into that 32 goal season to see if Salah is truly Liverpool’s best finisher.
In Basketball efficiency is used as a tie breaker, looking at the above numbers Mane should’ve took the award solo for his true volume and true efficiency, and looking at the numbers this season they are still converting at about the same efficiency showing correlation.
This is an easy way of seeing how many chances one will turn into goals, it’s easy to get and tells the full story immediately. Note, this isn’t to criticize Salah he’s one of my favourite players but I can like him and call him an inefficient shooter even if he’s a great player, he’s the James Harden of football.
The Twitter page will look deeper into second level stats so that any Layman can become a stat head, I’ll be posting stats charts measuring players efficiency rates. The page isn’t to act smarter than anyone or act like a know it all. It’s about creating a fun environment to explore 2nd level stats with other fans, I’ll be very active and engaging with everyone, it’s more a discussion forum really than a page where us casuals can look deeper into the numbers and get a more clear picture of what they are saying.
I’m not doing all this number analysis just for me, I love having basketball convos where things like True Shooting% and Per 100 possession are just common knowledge even if they are detailed 2nd level stats, hopefully one day goals per shot on target and per 10 shots are things people I don’t even know just throw in mid argument with their homies as a way of explaining their point better. Hopefully one day we have Football’s equivalent to Basketball Reference full of such detailed 2nd level stats, more than anything this page and the stats we provide will be about bringing casuals together.
By Mab Sidam.