Establishing which sides limit the opposition when out of possession in the SPFL.

Written by; Jamie Kilday @FitbaInScotland

This article is the second part of my series using Per Possession statistics to establish playing styles and Possession Efficiency, the first article is HERE. Per Possession is an idea I’ve been experimenting with for Modern Fitba, an introduction to which can be found HERE

In my first article I looked at which teams make the most of their possession with respect to attempts, goals and xG, now I’m looking at how effective teams are out of possession at limiting the opposition with respect to attempts, goals and xG. I will then look to see if there is any correlation between teams that are effective at limiting the oppositions possession efficiency and points using a new concept of points conceded…I’ll go into that a bit more later on…

First, a few things to bear in mind;

  • All event data is from open play, no point rewarding a team’s possession efficiency for goals scored/attempts from set pieces.

  • With that is mind it's worth noting what validating the metric against total points could be problematic as teams win games with goals and some of those goals will be from penalties, direct free-kicks and set piece routines…obviously.

  • This is only one seasons worth of data so repeatably isn't clear yet. We'll have to see in May if there is anything more to come from this metric. 

  • At the time of writing, we don’t have the data to analyse when a possession began and ended so I will be using averages. Another way this could be analysed if you had the data would be to look at how long it takes a team to get a shot away after first receiving possession.

  • As there is no real average number of minutes of possession across all clubs as it is dependant on the style of play and ability so I am using the ‘Per Possession’ metric to represent one minute of possession.

  • For context here is the average number of minutes of possession for each club in the 17/18 season.

Dashboard 1 (2) (3).png

I’ll cover this over the coming season but for now, let's look at last seasons data. I will say this though, it might surprise people to see that Kilmarnock, a team who did so well last season did so whilst conceding so much possession. Hopefully during the course of this article we will see why they finished in a very respectable 5th last season.

Attempts/Chances/Shots

Dashboard 1 (2) (6).png

There it is. No deep dive necessary, Kilmarnock last season were super effective at limiting the oppositions attempts on goal. This is probably not a surprise to anyone who watch them last season, their style of football is best encapsulated in fellow Modern Fitba co-founder Alex Lawrence’s tactical break down of Kilmarnock’s win over Celtic which can be found on Spielverlagerung.

In my last article Ross County didn’t look too shabby going forward and in isolation one would have thought that they were very unfortunate to be relegated but a look at their defensive statistics raises some serious red flags.

Hearts as ever continue to do their own thing, conceding chances at a similar rate to Hamilton, Dundee and a worse rate than the relegated Partick Thistle is a concern but the Tynecastle side still managed to finish in the top 6.

Dashboard 1 (2) (10).png

Points Conceded? Let me explain. When looking at the attacking side of things I used points earned to see if there was a correlation but when looking at defensive metrics it seemed strange to use points earned (although I take on board that if your defence is crap it is likely that you won’t earn many points), instead I’m using a metric I’ve devised which is Points Conceded. In short, it is how many points the opposition has earned, so a draw is 1 point conceded and a defeat is 3 points conceded.

Here we can see a fairly strong correlation between points conceded and an ability to limit chances. Kilmarnock being so far above the trend line is indicative of the amount of possession they conceded. Early signs this season show that Kilmarnock are not looking at keeping possession more than last season. Based on last seasons fixtures as a like for like comparison to this season Kilmarnock are only conceding 45 seconds less possession per game on average.

Goals

Dashboard 1 (2) (5).png

Ross County are still looking poor and their eventual relegation is looking justified, Hamilton conceded at a similar rate and might count themselves lucky to have stayed up on goal difference last season. Interestingly its becoming clear across the two articles that Ross County and Partick Thistle were relegated due to seemingly different issues. Ross County created chances well but couldn’t defend for love nor money yet Thistle were woeful at creating chances yet seemingly much better defensively (although still worse than average)

Here we can see maybe why Hearts finished in the top 6 despite conceding so many shots. It would appear that a lousy conversion rate meant that they actually conceded very few goals in respect of possession conceded. A look at the xG for last season will give us a better indication as to weather this was by design or if Hearts rode their luck last season.

Dashboard 1 (2) (11).png

The correlation has lost its way a little here which is fine, as a statistic, goals are volatile and not entirely reliable and as previously noted not all points are earned (or conceded) from open play so I’m not overly concerned. If anything it highlights that Rangers and Hibernian rode their luck defensively. Conceding at a high rate won’t matter too much if you are not giving up a lot of possession but if you are looking for reasons why neither managed take second from Aberdeen then it’s not a bad place to start.

Hearts, who I have suggested may have rode their luck, actually still conceded more points than would be expected for a side conceding so few goals. This can be best explained by the fact they rarely lost by more than one goal and had the joint most draws in the league (many finishing 0-0).

xG

Dashboard 1 (2) (4).png

Last season I introduced Expected Points to the SPFL and I will be reprising this for the coming season once we are a few more games into the season but Ross County spent most of last season severely under-performing and look at the state of that xG against…how does a side conceding xG at a lower than average rate finish rank last in the league…

…oh yeah, I remember now.

As I suggested earlier xG would give us a better understanding of whether Hearts low amount of goals conceded was by design or a stroke of good luck. A higher than average xG against would imply that it is the latter which is true.

Its also probably no surprise that Rangers completely overhauled their back 5 over the summer as their Goals and xG conceded were more in line with a side battling relegation rather than European spots.

Of all the aspects I’ve looked at this is probably the furthest from how the table finished up with the exception of Celtic.

Dashboard 1 (2) (7).png

Well…the correlation is…at least its still going in the right direction, right.

So publish this at all? If anything it highlights exactly what it is I like about Scottish football, its so unpredictable. Sides conceding high xG chances frequently should be punished for it but seemingly don’t, and sides who limit their opponent don’t seem to be rewarded.

Conclusion

As I concluded in my first article, a seasons worth of data isn’t sufficient enough to know if there is any value in this metric as a way of evaluating success. Speaking in general terms, sides that sufficiently reduce the oppositions opportunity to score concede fewer/earn more points, that’s hardly groundbreaking but I think that normalising data in this way does give a clearer indication of what a team does with and without the ball as opposed to looking solely at totals or ‘per90’.

This article was written with the aid of StrataData, which is property of Stratagem Technologies. StrataData powers the StrataBet Sports Trading Platform, in addition to StrataBet Premium Recommendations.