Winning Metrics in the PSL – part 2
By Alex Rathke
**This post originally appeared on City Press’s Sports pages on June 11th, 2018. If you missed the issue, we have added the blog piece below. **
In the first part of this two-part series, we looked at which metrics gave the highest win percentage in the PSL. The top 5 were:
- Shots on Target 2. Key Passes 3. Counter-Attacks 4. Box Shots 5. Shots
In this part, we wanted to look if the PSL teams had a preference between these 5 and if so which ones. Would a higher score in one or more metrics give away their style of play?
Before we proceed and show the fabulous table, we had to tweak the dataset accordingly. The number of games that each team played in the season was 30 games. The dataset we have access too has minor discrepancies between the teams. This can be explained by the lack of televised games (or so we think) as the data provider relies on the match footage. The number of games that each team played in our dataset can be seen below.
The only clubs that had all their games televised were the two most popular clubs (Kaizer Chiefs & Orlando Pirates). The two Soweto giants. Closely followed by Bloemfontein Celtic, Free State Stars and Mamelodi Sundowns. This discrepancy between the number of games played would unfairly value the metrics at face value.
In order to rectify this and make the values fairly comparable for all teams, we divided the total number of times each team won the “battle of the metric” ie: more Shots on Target than their opponent / the number of games played.
ie: For example, Pirates were superior in Shots on Target 16 times than their opponents across their 30 games played. 16 divided by 30 = 0.53. For simplicity, we multiplied each of the numbers by 100 to get the total as a percentage; 53%.
This number allows us to compare how many times Pirates (in this case 53% of the time) prioritised shooting on target over other specific metrics that we previously examined. We have shown that Shots on Target are highly important in the PSL when trying to win games. The equation we used is as follows:
No. of times team was dominant in “specific metric” / No. of games * 100 = % figure
The teams are sorted in order of league finish (Champions – top; Relegation – bottom). Each team’s focused “area” is explained by the given equation above. The bars are coloured in order from red (low – 0-20%) to blue (high – 80-100%). Any teams stand out from the crowd?
Champions Sundowns, for example, did not emphasise out-shooting their opposition but rather out-shooting them in the box (generally higher quality shots). Furthermore, they also had a higher percentage of shots hit the target than other clubs. On the flip side, Bidvest Wits had quite the odd season from a numbers perspective. At one stage they were bottom of the league but we at Laduma Analytics never feared them for relegation contenders (we will look at them in a separate piece). As we can see, Wits majorly emphasised shooting more than their opponents, nearly 15% higher than their nearest competitors! 62% and 54% of the time, their shots were from good locations and on target but there was something else and thus their season was “unlucky”.
Another team such as Free State Stars – who in 2017/18 won the Nedbank Cup relied heavily on counter-attacks (57%) to beat teams. A clear game plan from the cup winners and something that may need adjustment for next season. Only time will tell.
Tell us if any of these statistics were surprising. We’d love to hear your thoughts.
