The Canadian Premier League launched Centre Circle DATA!on Tuesday, giving analysts and fans alike unprecedented access to statistical information on a wide (seriously, wide) variety of data from the 2019 CPL campaign.
And, while this writer fancies himself as somewhat well-read on this game we call beautiful, this kind of information reveals storylines, anomalies, and insights that you just don’t get from watching matches alone.
So, in that spirit, here are three statistical talking points that emerged after a coffee-fuelled foray into the world of Centre Circle DATA!
3 insights from Centre Circle DATA!
1. Mo Kourouma, take a second
There’s an interesting phenomenon out in Halifax, surrounding attacking dynamo Mohamed Kourouma.
The 24-year-old certainly passes the eye test; he’s a constant threat in, around, and definitely outside the box, right? With 17 shots on goal, he’s got that part down. But, with 32 shots off target, and just one single goal scored in 2019, something is amiss here.
The problem, as indicated through Centre Circle DATA!, seems to be in Kourouma’s shot decision making. He’s attempted a lot of shots from well outside the box, to little effect. As such, Kourouma’s average shot has about a 5 per cent chance of actually finding the back of the net, whereas an average shot across the world ranks in at about 12 per cent.
Here’s where Kourouma took his shots from in 2019:
What can we extrapolate from this? Well, when comparing his shot accuracy to the location of his shots, one thing is clear; most of those 17 shots on target (11, in fact) came from inside the box. And, most of those 32 shots off target came from outside the box.
So, this yellow line right here? Anything above it, go ahead and shoot. Anything below it? Stop, re-asses, and try something else, because …
Goals aren’t the only contribution a player makes in a match. In that regard, Kourouma, with three assists in 2019, was a valuable asset to HFX, but still below from his expected assists of 6.82. That number, right there? It’s the highest of any player in the league. Expected assists is the total of the expected goals a player creates for their teammates. So, the best advice for Kourouma? Stop shooting from well outside of the box, and try playing provider instead, because this is where he has the potential to shine, better than any other player in the entire league.
“Given Kourouma’s impressive expected assist numbers, he’s definitely a player with tremendous potential,” Oliver Gage, Head of On-Field Performance and Recruitment at the Canadian Premier League, told CanPL.ca. “With some purposeful, structured learning about things like shot selections, players can make huge strides in their development and improvement in a short space of time.”
2. Who’s the xGolden Boot Winner?
The Canadian Premier League’s inaugural Golden Boot recipient was Forge FC ace Tristan Borges, who notched 13 goals and five assists after 1,973 minutes of play in 2019. But, it wasn’t always a sure thing for the 21-year-old Canadian, who defied his xG and xA numbers (9.35 and 4.08) to pip his peers Terran Campbell of Pacific FC, Easton Ongaro of FC Edmonton, and Dominique Malonga of Cavalry FC for the honour.
Speaking of Malonga, those same numbers, mentioned above? They had him pegged for an unprecedented, unmatched lead at the top of the scoring charts. See, while Borges won the Golden Boot in reality, the data indicates that Malonga really ought to have hit 15.56 goals – that’s his xG after a season’s worth of work, where he also generated 27 big chances, almost two times as many as Borges, at 15.
Malonga: G (11) – xG (15.56) = -4.56 Borges: G (13) – xG (9.35) – +3.65
Basically, this boils down to simple fact that Malonga took better-quality shots, more often, than Borges, but Borges finished more of his opportunities in actuality. Here’s how that breaks down, by the excellent Alex Sheppard:
Woooeeyy having fun with #ccdata. Great work by @canpldata to put this data in the hands of nerds of all kinds
“For what it’s worth, expected goals are a better predictor of future goals totals than actual goals,” Gage offered. “In this regard, Borges has defied expectations, while Malonga created more high-quality chances, more often. If we played this season 100 times, I’d expect Malonga to outscore Borges in the majority of them.”
3. Valour should have lost by five
Remember Cavalry’s 8-0 drubbing of Valour FC? It’s an outlier result in the league, as no team came close to matching such a one-sided result, apart from York9’s 6-2 win over HFX Wanderers FC. As far as blowouts go, Cavalry’s 8-0 win was unprecedented in the 2019 season. Here it is again, for your viewing pleasure:
But, was this the expected result for Valour FC?
See, Valour was expected to score 0.39 goals against Cavalry, and, predictably, didn’t find the back of the net. Meanwhile, Cavalry was expected to score 5.32 goals against Valour, coming into this fateful night at IG Field. If the data had its way on average, this match ends 5-0.
But that’s not what happened. The reason why is nuanced. You can look at player morale, fatigue, refereeing, a whole slew of factors, and still not get a full picture as to why Valour conceded eight. But, from a purely technical standpoint, this match really is an outlier, and Rob Gale can rally his troops, as such.
“Expected goals for and against at the team level is simply an extension from the player level – a culmination of every shot you take or concede in an individual game,” Gage explained. “Form isn’t necessarily included in this metric, though the quality of shots – where they’re from, and the type of assist – does play a role.”