Just How Super Are Sky's 'Super Sundays'?

‘Super Sundays’ is a phrase that the overwhelming majority of UK-based football fans will be familiar with. Through their bold creation and manipulation of that awareness over many years since acquiring the Premier League rights in the early 1990s, Sky Sports have managed to successfully build a significant hype around their flagship football feature which helps to encourage more fans to tune in on Sundays than on any other day of the week.

At a time when live television audiences for the English top tier have been somewhat wavering, despite the cost of worldwide broadcasting packages shooting upwards, the factor of that Sunday boost has proved particularly important in ensuring that the number of people watching the ‘best league in the world’ hasn’t dropped to a large extent. Games televised by Sky on Sundays in 2014/15 got a 12% higher audience than their average match, while it was 13% higher in each of the last two full seasons. This season it’s even gone to a whole new level compared to those, up at a massive 21% by the turn of the calendar year.

So why is the comparative success of these Sunday slots of such keen interest? Well, by looking into that and the rest of their audiences in recent times, we can examine the health of the Premier League as a product – and evaluate the strategies of broadcasters after seeing how much different factors help to shape those bigger viewing figures. Are Sky and BT getting as much value for their money as they can, or is there a way that they could utilise their positions better?

Assessing the strengths of Sundays

The table below, consisting of data collated from every televised match between the start of the 2015/16 season and the end of 2017* 1, demonstrates this pattern. A comparison between each day and each of the most common timeslots shows that games on Sundays, especially the late afternoon kick-offs, draw a considerably higher average audience than the others do.

Day and Timeslot BT and Sky - All Games BT and Sky - All Games Sky Only - All Games Sky Only - All Games
Days: Games: Audience vs. Average: Games: Audience vs. Average:
Saturday 147 0.84 71 0.95
Sunday 167 1.21 155 1.15
Monday 44 0.89 43 0.80
Other 53 0.88 36 0.85
All 411 1.00 305 1.00
Timeslots: Games: Audience vs. Average: Games: Audience vs. Average:
Saturday - Early afternoon 70 0.80 42 0.81
Saturday - Evening 75 0.87 28 0.96
Sunday - Early afternoon 16 0.64 4 0.78
Sunday - Mid-afternoon 68 1.08 68 0.99
Sunday - Late afternoon 83 1.42 83 1.30
Monday - Night 39 0.87 39 0.79
Other - All 60 0.90 41 0.86
All - All 411 1.00 305 1.00

Thinking about why this is the case is the next logical step to take. And there are a few possible reasons for it. One of them, quite literally, is the time. Games being shown on a Sunday afternoon means they line up nicely with one of the prime slots for television viewership, resulting in them benefitting from the somewhat artificial boost of more people simply being at home and around to tune in**.

Perhaps the most interesting possible factor, though, links into the aforementioned excitement which Sky have built up around these events: the teams that are playing. Arsenal vs. Liverpool, Man Utd vs. Man City; these are just two of the Sunday games that have been shown so far this season, both featuring sides heavily expected to be in the top six at the end of the campaign. Thanks to the broadcaster’s efforts, that’s the kind of fixture that people generally associate with Super Sundays. The best against the best. But is that actually the level of product we’re all getting on a consistent basis? And if it is, or isn’t, how much does the quality of the sides involved end up shaping the audiences?

Just how 'super' are these Sundays really then?

In order to start looking into that, a table detailing how often teams get shown on Sundays by both broadcasters was put together. Alongside that, the percentage of their matches during their seasons in the Premier League that were shown on TV compared to the total number of televised games, on Sundays and just in general, was also included.

Team (PL Seasons***) Televised Games - Sundays Televised Games - All % Televised - Sundays % Televised - All
Man Utd 32 69 19.2% 16.8%
Arsenal 31 68 18.6% 16.5%
Liverpool 26 65 15.6% 15.8%
Man City 24 65 14.4% 15.8%
Tottenham 24 58 14.4% 14.1%
Southampton 21 36 12.6% 8.8%
Everton 19 46 11.4% 11.2%
Chelsea 18 61 10.8% 14.8%
West Ham 15 41 9.0% 10.0%
Leicester 13 41 7.8% 10.0%
Crystal Palace 13 30 7.8% 7.3%
Swansea 12 25 7.2% 6.1%
Newcastle (2) 11 27 11.3% 11.1%
West Brom 10 27 6.0% 6.6%
Sunderland (2) 10 21 7.5% 6.5%
Bournemouth 9 26 5.4% 6.3%
Watford 9 26 5.4% 6.3%
Stoke 8 23 4.8% 5.6%
Burnley (2) 8 14 7.7% 5.4%
Middlesbrough (1) 7 13 10.0% 7.7%
Aston Villa (1) 4 11 6.3% 7.1%
Norwich (1) 3 9 4.8% 5.8%
Brighton (1) 3 7 8.8% 7.9%
Hull (1) 2 8 2.9% 4.8%
Huddersfield (1) 2 5 5.9% 5.6%

As shown, Man Utd and Arsenal are the two Premier League teams who’ve featured in televised Sunday games more often than anybody else. Meanwhile, they, Chelsea, Liverpool, Man City and Tottenham are by far the six most shown teams in all slots. And seeing as those are also the six sides with the highest average audiences in the measured period, that’s far from a shock.

Given that games not featuring at least one of those ‘top six’ teams have averaged just 66% of the viewership that a match involving one would, it’s understandable that BT and Sky would favour showing them so much. What perhaps is surprising, though, is that, despite the higher viewing figures and hype around these ‘Super Sundays’, there’s actually no discernible pattern when it comes to bigger teams playing more on Sundays. Half of them are even less likely to be televised on a Sunday than they are on other days.

In fact, their lack of action goes so far as to tip it the other way. 75% of all games in the sample featured at least one top six side, compared to 73% of Sunday games. The below graph gives some more context on that, and shows exactly how Sundays under index in this regard.

Only 46% of the total teams who played (i.e. counting the ‘Home Team’ and ‘Away Team’ for each fixture) in televised Sunday games were top six sides too, compared to 50% on Saturdays and 34% on Mondays. Not a significant drop from that of Saturdays by any means, but it’s enough to dispel the myth that these half-dozen teams make up a greater proportion of Sunday slots than the others. So is it the case that Sundays end up performing better in spite of them playing host to what are actually slightly less favourable matches then?

Based on that, yes. A look at the fixtures between top six sides is necessary to paint a full picture though – because after all, those are the games which Sky are predominantly marketing themselves around. The biggest Premier League event(s) of the week. And, to an extent, that image is correct. 31 of the 53 top six games that Sky have shown in this sample took place on Sundays, the vast majority coming in that high-performing late afternoon (25) slot. On absolute numbers alone that’s a lot higher than the amount televised on Saturdays (13) and Mondays (4).

Yet factor in the distribution of the number of games that they show, which typically is one on Saturday and two on the Sunday, and it’s nothing out of the ordinary. 18% of their 71 Saturday games have been intra top six fixtures, just marginally lower than the 20% of their Sunday games that have been. So maybe it’s true that more of their big games occur on Sundays in a literal sense. But really, that seems to be nothing more than just a by-product of them showing more matches on that day in the first place.

The performance of Sundays when excluding intra top six fixtures

Sky’s Sundays aren’t particularly super after all, it turns out. And linking that back to the audience numbers again, those games between the top six also have limited impact on which television slots get the most – and least – views too. Remove those from the equation totally, as below, and the balance is similar. Sundays still significantly outperform the other days regardless of the big profile games.

Day and Timeslot BT and Sky - No Intra Top Six BT and Sky - No Intra Top Six Sky Only - No Intra Top Six Sky Only - No Intra Top Six
Days: Audience vs. Average: Difference: Audience vs. Average: Difference:
Saturday 0.85 +0.01 0.89 -0.06
Sunday 1.20 -0.01 1.15 +0.00
Monday 0.87 -0.02 0.80 +0.00
Other 0.89 +0.01 0.87 +0.02
All 1.00 +0.00 1.00 +0.00
Timeslots: Audience vs. Average: Difference: Audience vs. Average: Difference:
Saturday - Early afternoon 0.81 +0.01 0.81 +0.01
Saturday - Evening 0.88 +0.01 0.98 +0.03
Sunday - Early afternoon 0.71 +0.07 0.87 +0.09
Sunday - Mid-afternoon 1.14 +0.06 1.05 +0.06
Sunday - Late afternoon 1.38 -0.04 1.27 -0.03
Monday - Night 0.85 -0.02 0.78 -0.01
Other - All 0.92 +0.02 0.90 +0.03
All - All 1.00 +0.00 1.00 +0.00

That’s not to say that those games don’t drive a huge upturn in viewership on a match-by-match basis; compared to an average game, a meeting between two top six sides can be expected to boost audiences by roughly 60%. What it does show, though, is that the theory of bigger games on Sundays being the main driver of that day having larger audiences is false. Seems like the timing has been the key all along.

So should matches between the top six be distributed in a different way?

The debate that this consequently brings up, then, surrounds the distribution of games, and whether Sky should pick when to show their matches in a different manner. We now know that a typical game is going to get a higher viewership if it’s on a Sunday anyway – so could it be more beneficial for them to split up the two big factors of a) games between the top six, and b) that late afternoon slot on Sunday, on a more regular basis?

Creating some ‘Example Weekends’ does suggest it might well be. These ‘weekends’ were put together by following a couple of steps, the first of which being to collect the average audiences for non-top six games in each of Sky’s four most common TV slots. Following that, in each example one of those four would then be changed to a top six fixture, and the viewership number updated to reflect the average of such a game in that individual slot. Then a total of those audiences would be calculated, allowing us to see what the best combination is. The below is a graphical representation of the results, with ‘Example 3’ being how Sky most commonly arrange things at the moment****.

As shown, two of the three other proposed structures would result in Sky getting just as good, if not better overall viewership. None of the individual slots would get quite as high an audience as the late Sunday afternoon ones do right now, however across the weekend it would lead to a more balanced and, most crucially, more watched Premier League product in the UK.

Admittedly this concept is far from perfect. One of the issues is down to the amount of cases there are to work with, seeing as there have only been four top six match-ups on Monday nights in these two and a half seasons. That means the above average multiplier (2.27) for such games in that slot is, at best, rather unlikely to continue (as seen earlier Monday nights have hosted a very low proportion of those six sides in the past though, so it may well be the ‘normal game’, 0.78 figure which is slightly unrepresentative). And while the boosts elsewhere seem far more reasonable, only the late afternoon timeslot on Sundays has hosted a double-digit amount of these big fixtures.

Another possible flaw is how a change in this general structure might, for better or worse, affect consumption habits. For instance, moving the biggest game to regularly be in the middle of the day on Sunday instead might mean that less people leave their TVs on for the later, less popular match of the afternoon; alternatively, there could be an overspill effect which leads to the audience going even higher. Given the limited examples of scenarios like this one, it’s hard to make a definitive proposal for change.

Finding some middle ground, then, instead of coming away with a bold hot take in either direction, is the right way to look at this. No, there’s not a clear, fool proof method that Sky can adopt to make their audiences skyrocket. But informed assumptions and predictions can be made based on general viewing patterns both for the English top flight and other events: and following those there certainly seems to be scope for the broadcaster to make improvements on what they’re currently doing.

But how much choice do broadcasters have over when matches are shown?

If Sky were to pursue an active change in this manner to try and increase the number of viewers they get, it has to be said that it’s not quite as simple as just putting any game whenever they want, though. Certain things are out of their – and BT’s – control. UEFA’s fixture calendar is one of those, with European competitions having a say in when those better teams can and can’t play. Throw in other sporadic things such as domestic cup ties and policing reasons, and the boundaries which the broadcasters have to work within are tightened further.

Biggest of all however is the manner in which Premier League rights are sold, and the structure of the ‘first pick system’ that Sky and BT abide by. There are places to find more specific details on exactly how the distribution worked for 2013-16 and 2016-19, but a very brief summary is that broadcasters bid for groups of games at specific times before then selecting which matches they want to show when during each gameweek. This is a breakdown of how things went for the latest deal:

Package Broadcaster Games Included Games First Picks Second Picks
A Sky Saturday early afternoons 28 6 14
B BT Saturday evenings 28 9 0
C Sky Sunday mid-afternoons 28 0 15
D Sky Sunday late afternoons 28 18 0
E Sky Monday (min. 18) and Friday (max. 10) nights 28 0 0
F BT Midweek evenings (6) and Saturdays (8) 14 3 7
G Sky Bank Holidays (8) and Sundays (6) 14 2 2

Compared to the 2013-16 deal, Sky once again got a 75% market share of the games, as well as gaining another six first picks to take their total up to 26. This essentially means that in all but 12 weeks of the relevant seasons, they get to guarantee the selection and airing of the best match being played.

The problem they’d encounter when trying to shuffle around how they schedule these games is that the majority of those first picks have to come in the late Sunday afternoon slot. Only eight don’t. And on the 18 occasions that those will be used up, Sky are presumably going to snap up what they see as the week’s standout game that the most people will watch. Otherwise there’s a decent chance they’ll lose it to BT.

Information regarding who chose which match during the picking procedure is kept confidential. All we know about it for sure is that it’s unquestionably a long, complicated process, with a load of constraints to consider: hardly a surprise given that this is one of the biggest sporting competitions around and that the overall package sold for an enormous £5.1 billion back in 2015.

Maybe Sky have thought about adjusting their picking method but decided that the way they’re doing it right now is best. Perhaps they’ve tried to change things and found that it’s just not logistically reasonable within the Premier League’s pick system. But having a highly significant number of second picks secured within packages A and C (29 in total) as well does make it seem like there should be room for flexibility when they’re making their choices.

And if there is, then there’s plenty to back up the idea that trying something new may be more beneficial for them – it’s not as if their Sundays are actually all that super at the moment, anyway.


*Period of games used ranges from 08/08/15 through to 31/12/17, with viewing figures taken from the UK audience bureau, BARB. Averages for each day and timeslot through the whole article are all indexed, with 1.00 representing the average audience in each of the specified scenarios.

**Figures for the 'day-by-day audiences throughout the week' are made up of the average of that time and day (as specified on the X-axis) throughout 2017. Audiences are indexed, with 100 being the highest point.

***Number of seasons in the Premier League for each team during the data period used is 3 (i.e. 2015/16, 2016/17, and 2017/18 so far) unless otherwise specified in brackets.

****The intra top six game during each of the 'Example Weekends' is indicated by the one outlying, highest average for each timeslot (e.g. for Monday night games, 1.77 represents it, compared to the 0.78 for the rest of the matches in that slot).

  1. Period of games used ranges from 08/08/15 through to 31/12/17, with viewing figures taken from the UK audience bureau, BARB. Averages for each day and timeslot through the whole article are all indexed, with 1.00 representing the average audience in each of the specified scenarios.

  2. This is a loose guide and doesn’t hold in 2017, where England played 4 Tests vs. a strong South Africa side in July / August, and 3 against a weak West Indies team in August / September.

  3. Viewing Hours = Average Audience x Duration in hours. One person watching for 1 hour = 1 VH

  4. A linear regression model.