Analytics might not have fully made its way into Nigerian/African sports but there’s still tons of data being generated and a good portion of it accessible through diligence. This work makes use of historical data and attempts to thoroughly explain the insights uncovered from the data collected on the Nigeria Women’s Football League (NWFL). It should be noted that this is purely analysis and no ML or AI methods were used. The slides are quite many so I will write short, concise text descriptions!

Midseason Summary.

We start at the distribution of FT results and see here how the teams are doing and it is clear how difficult winning on the road has gotten when compared to the data from last season.

As a quick summary, four (4) teams were promoted last season from the Championship and two relegated from the top-flight league. The promoted teams were: Remo Stars Ladies, Dannaz Ladies, Ekiti Queens and Sunshine Queens. This act takes the tally of participating clubs to 16 and the games played at mid season is seven (7) per team across all eight teams in each of the two groups.

The obvious effect of this is that more games will be played and potentially more goals would be recorded. This is evident from the next slide as we can see the goals scored at the midway point is higher this season than the previous season (2023/24).

Interestingly, there has also been an increase in the number of goalless draws recorded and in the average goals scored by the teams at home.

Let’s examine the season in terms of actual match results at FT. As expected, the results tilt towards the home games, and you can confirm the draws by summing the diagonals of the heatmap. You can see that 21.43% of the games—the league highest—have ended with the home teams winning by a lone goal and the next common FT match result is the home team winning 2-1. The visiting teams have mostly had to settle for a point and whenever a team wins on the road, most times it’s 0–1—and that’s a small amount.

I’ve also estimated the FT results as a distribution of probabilities and you can have a look below. Please note that though approximations may affect the figures — quickly seen when you examine the FT draws — but they all say the same thing.

Similarly, when you examine the results in either half, the narrative doesn’t really shunt as there’s been a higher chance that the teams play out a draw in any of these halves or the home team wins the half — this supports the earlier distribution because if the commonest result is a lone goal victory, then a good portion of (either of) these halves should be stalemates as the goal can only come in one half. This per-half dissection also speaks to the goal-scoring prowess of the teams in the halves played this season. The home teams can score up to three goals in either half but the maximum in both halves for the visiting teams is two goals. The chances of scoring multiple goals however keeps decreasing and really, you can see, again, that the trend is the same as it’s mostly a draw in either half or a low margin win.

Next, let’s look briefly into the game weeks and examine the goals and FT results. It immediately comes to attention that from GW4 to GW6, no side secured a win on the road, in fact, in GW6, the home teams recorded a 100% win — this isn’t the first time such would be happening though, in 2022/23 and on GW1, the home teams also had a 100% success.

One can also observe that GW1 and GW3 were the closely contested game weeks where it would seem that teams travelling that week really pulled points off their hosts. One may suspect this is due to the quality of teams playing away that week but team rating is not exactly a feature used within the scope of the views expressed in this piece. The home teams also contribute the most to the goals scored in every gameweek, as shown above. A quick summary of key points from the current season is shown below to give an overview of the season.

Team Performance.

With historical data, there is a limit to how we can analyse the performance of teams but it is possible to get an idea of how the teams have been playing this season and what trends there are. The first place to start would be the league standings after GW7 (mid season) and we can see who the leaders are and by what margin they’re leading. Newly promoted Dannaz Ladies currently sit in second place behind Heartland Queens, and defending Champions, Delta Queens, languish in sixth place. At the moment, the top of Group A is also closely contested with goals being the major decider in these rankings, as all top four teams have garnered 11pts each, thus suggesting that an interesting contest is going to ensue in the second stanza as the teams jostle for a Super Six slot.

When we zoom in and examine the league standings every week, we can see there is a clear stratification in Group A. There’s been some activity in the Top 5, Naija Ratels and Adamawa Queens seem to be shuffling sixth and seventh place while Royal Queens have maintained a flatline all season.

The battle for a Super Six slot is an interesting one to look out for in Group B as Rivers Angels, Bayelsa Queens, Edo Queens and FC Robo Queens jostle for one of three available slots. Edo Queens have been pretty consistent with staying up but it is the ascension of FC Robo Queens that calls attention and we’ll see later on that they’ve been doing it in quite the uneasy way.

If you enjoy numbers, I’ve estimated the first stanza of the season in averages and percentages for both groups. Heartland Queens in first position at mid season is a rise from their position at this stage last season, Confluence Queens were also fourth at this stage last season but for everyone else aside Dannaz Ladies, they’re at least one position worse this season than they were in 2022/23.

Despite winning the relegation battle last season and having +3pts more than they did at this stage last season, Royal Queens sit in eighth place and one can quickly suspect it’s the away games that’s doing the most damage!

The story in Group B is quite different, defending Champions, Delta Queens, are -7pts shy of their tally from this stage last season while record Champions, Rivers Angels, are the biggest earners when compared to the previous season with +5pts. 

Edo Queens managed the same points tally in this campaign as they did in the last but they trail Bayelsa Queens on goal difference and have fallen two spots from where they were at this stage last season. Edo Queens are also one of only two teams never to have dropped a point at home all season.

Next, we attempt to evaluate briefly what the pattern has been during the games for each team by calculating in percentages the minutes spent per instance of goal difference during the games. For this, adjustments were made to the seven goals scored in stoppage time as they were recorded as either a 45th or a 90th minute goal depending on which stoppage time the goal was scored—first half or second half.

Sixth placed Adamawa Queens have spent the most game time being behind in games when compared across Group A but, interestingly, they’ve also spent almost the same proportion of game time as first-placed Heartland Queens being ahead in matches played this season — two contrasting sides.

In Group B, Ekiti Queens, for example, has spent the most time chasing games this season — the league’s highest, but they have also spent a significant amount of time playing a draw in games this season but I’d argue that, for a team who has only scored 3 goals all season while letting in 11, it won’t likely be about how well they are able to comeback to draw games, rather it’s going to be more about how long they have been able to delay conceding the first goal (you’d see more on this in a bit). Rivers Angels on the other hand have spent the most time being ahead in games — the league’s highest, with 50% of total minutes played!

Again, we examine the chances of match results at FT and in both halves. Group B has a high rate of one-all draws — there have been seven 1–1 draws at FT this season and six of them were in Group B!

It is also common to see the home teams winning by a two goal margin in Group B and this possibly gives an idea into the quality of the contests that’s been going down in the NWFL group with a significant number of the league’s heavyweights. One thing to add here is that, in the six games finishing 1-1 at FT, Remo Stars Ladies and Bayelsa Queens have each been involved in three and the Ikenne Ladies have often had to come from behind to force a draw. Conversely, a significant amount of the goalless draws this season have come from Group A and winning 2–1 is just about as common as playing 0–0 in the group as well, however, it is most common that the home teams win by a lone goal.

To round off the performance metrics, we examine the points won or dropped by teams from losing or winning positions respectively. Group B takes the lead on points dropped as second-placed Bayelsa Queens and relegation bound Sunshine Queens have each dropped 5pts from winning positions — the league’s highest, while the defending Champions and first-placed Rivers Angels follow closely with 4pts dropped each.

Adamawa Queens have won 6pts from losing position and only FC Robo Queens (7pts) have won more. Even in the recently concluded GW8 fixtures, Folashade Ijamilusi (FC Robo Queens) had to score a 47th minute equaliser to earn them a point. While this may speak about their recovery powers, it may also be a significant undoing in their race for the Super Six — that’s definitely something to look out for in Group B.

Goalscoring (Attack Vs Defense).

A good portion of football reviews is reviewing the goals and I spent some time checking the data to see amongst other things, which teams have had the most goal scorers out of the 91 different scorers so far this season.

At the team level, I broke the games into eight quarters of 15 minutes each (incl. 45+ and 90+ to denote stoppages) so you could get an idea of when the teams were scoring/conceding. I also checked to see which teams were flourishing in terms of goal differences, in how many game weeks (GW) the teams have scored and which teams were more accustomed to scoring first or conceding first in games. 

Of course the group leaders are high up on goal difference and the bottom placed teams are right underneath the chart but it’s Edo Queens who split the plot in half with 10 goals scored, 10 goals allowed and a goal difference of 0.

Nasarawa Amazons and Dannaz Ladies have put up exactly the same numbers in the first stanza of the season and have exactly the same goal averages. Sunshine Queens have the worst defense in the league as it would seem but Confluence Queens would appear to have the best defense in the league — they’re currently fourth in Group A and tied on 11pts with the top 3 in the group. Right next to Confluence Queens you’d find the top ranked teams from either group and defending Champions, Delta Queens — simply put, the league leaders have been defending quite well. Zooming in on Delta Queens, they’ve only allowed five goals and scored seven, earning just two wins and three draws in the process. They have also spent the most time playing draws during games in Group B and in two of the games where they scored first, they settled for a draw in one and relinquished all three points in the other — for a league where half season means seven games, two is quite a significant number. It would therefore appear that in a season where low margin wins are rampant (mostly 1–0 wins), a team could concede so few yet lose so much.

In a later paragraph, we would use an instance of Ekiti Queens and FC Robo Queens to explain a few things — notice they have identical defenses.

Here we take a look at when the teams are scoring and observe that 63.2% of the goals this season have come in the first one hour (< =60min) of games played!

Further aggregation would reveal that the bulk of those goals have come between the last quarter of the first-half (21 Goals) and the first quarter of the second (25 Goals) with just one goal recorded in the stoppage time in between.

Naija Ratels (4) have allowed the most goals in the last quarter of first halves — joint highest with Ekiti Queens (also 4), while Royal Queens lead the way for most goals allowed in the final quarter of matches played.

From both posts, you can see Ekiti Queens concede a lot of first-half goals (7 — Most in the league) and FC Robo Queens is right behind them with six conceded in the first-half, but, FC Robo have flourished in the second-half of their games, fighting back with enough goals to win the most points from losing positions and scoring thrice as many goals as Ekiti Queens have managed in that same timeframe! While their defense may be identical, their firepower appears to have been the difference maker.

Again, it is important to restate that since we’re dealing with historical data, contextual information may be lost and the extent of analysis is limited. An example of such is the inability to see from which patterns the goals are coming — open play, error sequences or set pieces for instance; what events are leading to the shots conceded and since there’s only been three penalties scored all season (until GW7) we could also want to know the quality of shot conceded during these instances of goals scored (xG). To accurately collect this data is to transcend the level of data collection used for this work and track event data during games but, the Nigerian sports system is not there yet.

From the 117 goals scored by 91 different scorers this season, Remo Stars Ladies have had the highest distribution of goal scorers with nine, and the group leaders from both groups are closely behind them.

Remo Stars have also maintained the record of scoring every GW — they even carried this on in GW8, scoring in their 2–1 defeat away to Ekiti Queens.

Rivers Angels who have scored four times in the opening 15 minutes of games played this season, are accustomed to scoring first in games played this season, with the Angels scoring first in 6/7 games played by mid-season — they also scored a 21 minute opener in their GW8 fixture at home to FC Robo. As you probably guessed, FC Robo Queens — who concedes a good amount of first-half goals and have won the most points from losing positions, are also accustomed to conceding first in games played this season.

The first goal in home games seems to be tilt in a different direction as holders, Delta Queens — who have scored first in four games this season, have only done so on home turf. On the other hand, four of Rivers Angels six opening goals have come while playing in enemy territory. Heartland Queens have also scored 3/4 opening goals on the road but it is Bayelsa Queens who set the tone being among the leaders in opening goals scored either at home or on the road.

Conversely, we examine those fond of conceding first in games (home or away).

Up or Down: Relegation Trend.

As a recent trend, the teams are split in two groups and this season, there are two groups of eight (8) women’s football teams. In the last season and oftentimes, the last placed teams in either group were relegated except in 2019 where a relegation playoff was held as a one time thing.
The data used for the relegation trend was from 2018 — Date and past seasons whose formats were similar to the current 2023/24 season were prioritised (*Abridged League, Round Robin Group Games). However, since playoffs were a significant factor — as observed by the possible anomaly of the 2019 season — a season where relatively poorer teams (Osun Babes for instance) stayed up on penalties courtesy of a one game contest in the playoffs and at the expense of relatively better relegation counterparts, we focused on seasons where there were no relegation playoffs.

As a note, in the 2019 season, there were four groups of three teams each and six games per team made the entirety of the regular season. So, while the format used for the playoffs format is definitely significant, the groupings, games played and quality of opposition faced by the relegation candidates in the regular season could also be a factor.

We however observed the following about relegated teams from the data used:
🎯 The Playoffs potentially impacts the decision on who gets relegated or wins the league.
🎯 Most times, relegated teams produce identical halves (or worse) in terms of points earned in half seasons, goals averages and a few other team performance metrics.
🎯 Teams who find themselves bottom at mid season end up being bottom at the end of the regular season!

How does this fit in the context of the current season?

The plot of goals averages of relegated teams from similar (older) seasons are somewhat identical to that of Ekiti Queens, Sunshine Queens and Royal Queens from the current season, thus suggesting a possible relegation for these teams. Royal Queens did win a remarkable +6pts in the second stanza of the 2021/22 season to escape relegation but, with a flatline all season, four goals in seven games and 2.33 goals allowed on the road, a repeat of such a feat in this current campaign looks bleak.

Possible Champions.

The data used here was from the last 10 seasons (2013—Date) and the same season format was prioritised as in the relegation trend. The top ranked teams (mostly top 3 but format for selection has varied many times in the last 10 seasons) are often shortlisted for the Championship Playoffs in a bid to decide who wears the crown. Since no ML or AI methods were employed in this work, we can only see what patterns are there in the data and may not be able to make healthy predictions as to whose head the crown would fit this season. Moreover, to predict the winner, the playoffs needs to be factored in as well — since it is expectedly a significant factor, and data concerning the playoffs was not involved in this mid-season analysis.

We can however notice a few things from the last 10 seasons:

🎯 Group A has produced 4/10 Champions from (same as Group B).
🎯 Most times, one of the teams who finish in first place in their group during the regular season ends up winning the league regardless of the Playoff format.
🎯 Teams who finish in second place during the regular season, seem to have a higher chance of winning compared to those who finished below them and entered the Playoffs.

That’s it for our NWFL mid-season review and I hope you enjoy every bit of what I’ve shown you.

Credits: