Skip to main content
ABC News
Harry Kane, Luckiest Man In Soccer

We’re only two games into the current English Premier League season, so it’s way too early to make any sweeping pronouncements. (Except maybe that Chelsea’s stint in 16th place won’t last long.) But here’s one fairly safe bet: Tottenham striker Harry Kane isn’t going to replicate his 2014-15 performance.

No player in the Premier League last season scored more goals beyond our expectations than Kane.

Soccer’s analytics are fledgling, but one of the emerging metrics is Expected Goals (ExpG), a measure that tries to assign probability that a given shot will result in a goal. A tap-in from inside the six is more likely to go in than a volley from 40 yards. ExpG is an attempt to capture that difference. Throw in any number of other factors — was the shot from a header, did it come off a cross, was the weather cold or rainy? — and one can build a model that calculates probabilities for every shot type at any location. I know it’s possible because I did it.1

Last season, Kane had 19 non-penalty league goals2 on an ExpG total of 11.33. That’s a hard thing to do year after year. Unless he’s a truly elite player — and we don’t have a reason to think he is yet — he’s due for a regression.

For Kane to overachieve again, he’d have to succeed where almost everyone else has failed recently. The top 15 over-performing player seasons, based on ExpG, from the 2011-12 through 2013-14 seasons declined on average 80 percent from their over-performance the year before. These aren’t scrubs, either: Luis Suarez, Yaya Toure, Klass-Jan Huntelaar — most of those are exceptional players. Kane would have to be a once-in-a-generation talent (like Lionel Messi, who has three of the top 10 overachiever seasons of the past four years) to not regress.

L. Messi 2012-13 Barcelona 42 23.8 18.2
C. Ronaldo 2014-15 Real Madrid 38 25.5 12.5
L. Suarez 2013-14 Liverpool 31 18.6 12.4
L. Messi 2014-15 Barcelona 38 27.0 11.0
G. Bale 2012-13 Tottenham 21 10.6 10.5
M. Reus 2011-12 Borussia Monchengladbach 18 8.2 9.8
C. Immobile 2013-14 Torino 22 12.3 9.8
L. Messi 2011-12 Barcelona 40 30.3 9.7
L. Podolski 2011-12 FC Koln 15 5.4 9.6
C. Ronaldo 2011-12 Real Madrid 34 24.8 9.3
G. Higuain 2011-12 Real Madrid 22 13.4 8.6
K. Huntelaar 2011-12 Schalke 04 23 14.4 8.6
M. Harnik 2011-12 Stuttgart 17 8.5 8.5
A. Greizmann 2014-15 Atletico Madrid 22 14.2 7.8
J. Rodriguez 2014-15 Real Madrid 13 5.3 7.7
H. Kane 2014-15 Tottenham 19 11.3 7.7
A. Sanchez 2013-14 Barcelona 19 11.5 7.5
J. Drmic 2013-14 Nuernberg 15 7.8 7.2
Y. Toure 2013-14 Man City 14 6.8 7.2
P. Cisse 2011-12 Freiburg / Newcastle 13 6.0 7.0

Except his short Premier League history doesn’t suggest that he is. In 2012-13, while on loan at Norwich, he took a grand total of four shots. The next season, in 10 appearances with Tottenham, he scored three times against an ExpG of 4.44. It wasn’t until a 1-2 home defeat to Stoke in early November of last year that coach Mauricio Pochettino even made Kane a regular starter.

Kane’s regression might already be happening. Going back to the beginning of April, Kane has scored just twice over nine games. It’s a somewhat arbitrary cutoff, but it encompasses the final stretch of Premier League games after the last international break (England drew with Italy; Kane played all 90 minutes but didn’t score) and the first two games of this season. The two-goal total is below Kane’s ExpG of 2.78.

Kane’s career is made up of 20 games where Kane is almost Messi and then another 21 (albeit broken into two nearly equal sets and separated by a full year) where Kane isn’t quite even Kane.

It’s not all bad news for Spurs fans. Mean reversion doesn’t mean Kane will score eight goals below his ExpG number this season so that his performances average out like magic. He’s just likelier to score something more in line with his calculated ExpG. He’s the team’s main (and close to only) option up top, which should help. (Although there is still time in the transfer window for Spurs to bring in some help.) He should see plenty of chances, enough that his ExpG could be in the neighborhood of 153 — and 15 here is still a big number. If Kane scored 15 league goals it wouldn’t even look like much of a drop-off, even if it represented a complete regression.

CORRECTION (Aug. 21, 1:33 p.m.): An earlier version of this article included a table that misidentified the team Luis Suarez played for in 2013-14. It was Liverpool, not Barcelona.


  1. I built the model using about 200,000 shots from Opta going back to 2011. There are an increasing number of people doing soccer analytics, building their own models and publishing some of their numbers. Inputs and methodologies vary, so there are likely to be variations among leagues, teams or players across the models.

  2. Kane had 31 total goals, but to keep things simple, we’re concerned with only non-penalty league goals.

  3. Because Kane will likely play more minutes this season compared with last (when he wasn’t even a starter for the first three months), I’m granting him an estimate of 700-plus minutes this campaign (assuming he stays upright). That yields an additional 3.5 ExpG worth of shots this season.

Michael Bertin is a writer and data analyst living in Chicago,Illinois.