The Week of the Streak

What a week it was!

– Re-create England’s pre-Euro 96 escapades with the ‘dentist-chair’, check.
– Witness one human deliberately eat another human’s sick, check.
– Equal the Oakland A’s 2002 winning streak of 20 games, check.

Like I said, what a week! The two former tales are obviously textbook ‘Football do’ night out behaviour. No rules apply, anything goes, disguised under the pretence of banter. The latter is the result of 18 months of data collation, a scientific formula and an unnatural love for the film ‘Moneyball’. No banter here.

Sunday 11th December, where it all started. I remember it clearly a.) because it was under two weeks ago and b.) it was the first time in my life I actually felt it possible to die from an overdose of calcium (too many Malibu & Milks obviously). I’m sure I don’t need to remind you of the game… Regionalliga West’s Schalke 04 II v Sportfreunde Siegen. If Bayern v Dortmund is categorised as a Heavyweight bout, this could be no more than Super Flyweight. That aside, the game ended 1-0 and The Panel had their first win of the streak.

For the next 6 days the robot would scale the world, from Derby to De Janeiro, Egypt to Eredivisie through Amsterdam and Panionios, twice. Winner after winner after winner. Only Aladdin could’ve scouted more countries in the same time, and I bet he had a monkey on the Ludogorets v Cherno More game.

Ignoring the impossible, if you started with £10, were able to get the optimum price of 3/10 and rolled over your winnings onto the next bet each time, you’d have exactly £1900.50. That’s a lot of Malibu & Milk. Don’t get me wrong, this will not happen every week. But with more analysis and data collation and less Paul Gascoigne and sick eating, we could eventually break the Oakland A’s record of over 20 consecutive wins.

Look at us as your card counter at the blackjack table, your Pete Brand (Paul DePodesta for genuine baseball fans) at the first draft, and fundamentally, turn the bookmakers odds in our favour and make some serious money.

One’s to watch:

Stranraer v Airdrie (Saturday 13:00 – Scottish League 1)

13:00 – Wake Up.
13:15 – First Beer.
13:20 – Second Beer.
14:00 – Receive Tweet.
14:02 – Place bet.
14:05 – Win bet.
14:06 – Third Beer.

Enough said.

Watford v Crystal Palace (Monday 12:30 – Premier League)

Expect the Lynx Africa gift set to be worth it’s weight in gold throughout this one. Boxing Day, early kick off, London derby, seem like the perfect ingredients for a cagey football match. It’s not usually until the second half where Crystal Palace take the rifle and direct it at their own feet, and let’s hope that happens sooner rather than later. Don’t worry if you manage to exhaust all of the deodorant whilst you nervously wait… you’ve got 10 more.

Newcastle v Sheffield Wednesday (Monday 19:45 – Championship)

In true European style, the rest of the continent head off for their mid-season siesta. Here we are, in the gods at St James Park watching Paul Walsh trial his 63rd new hairstyle/glasses combo, spluttering his way through his allotted 30 seconds of match analysis:
“Not much really going on Jeff”.
No shots? No dangerous tackles? No half-chances?
“The Geordies have got their shirts off though Jeff”
Oh piss off Paul.

Anyway, those sweet words uttered by Paul Walsh is exactly what we want to hear throughout the first half of this game. And expect it. The last/only two times these teams have met, the game has ended 1-0 with both times the goals being scored in the second half. Sheffield Wednesday are solid at the back, Newcastle are solid at the front. Lets hope this Yin and Yang lasts at least 45 minutes, but no more than 55.

Jay@gamblingpanel

Weekly Stats

This Week

WeeklyStats
WeeklyTopCountries
WeeklyTopLeagues

This Week’s Wins

Match Full Time Bet Score
Ludogorets vs Cherno More 1 – 1 57.78%
Panelefsiniakos vs PAOK 0 – 7 53.33%
Queens Park Rangers vs Derby County 0 – 1 61.11%
Panserraikos vs Panionios 0 – 3 70.0%
Hajduk Split vs Dinamo Zagreb 0 – 1 60.0%
Latina vs Brescia 1 – 1 82.22%
Gateshead vs Braintree Town 1 – 1 63.33%
Cambridge United vs Crewe Alexandra 2 – 1 54.44%
Walsall vs Bradford City 1 – 1 52.22%
AFC Wimbledon vs Port Vale 4 – 0 111.11%
Austria Wien vs Mattersburg 2 – 0 60.0%
Aris vs Trikala 3 – 2 55.56%
Ajax vs PSV 1 – 1 81.11%
Olympique Marseille vs Lille 2 – 0 51.11%
Hellas Verona vs Virtus Entella 1 – 0 58.89%
Athletic Club vs Celta de Vigo 2 – 1 54.44%

This Week’s Loses

Match Full Time Bet Score
Concordia Chiajna vs CS U Craiova 0 – 0 62.22%
Brest vs Bourg en Bresse 0 – 0 68.89%
Raith Rovers vs Dundee United 0 – 0 63.33%
Hannover 96 vs Sandhausen 0 – 0 54.44%

Overall

OverallStats
TopCountries
TopLeagues
TopTeams

How we’re building the perfect model

Ok, we haven’t built the perfect model. Not yet. Take last weeks game between Atlético Paraná and Ferro Carril Oeste. Paraná -18th in Argentina’s Prim B Nacional, Ferro Carril Oeste unbeaten in 8 games. So when the half time stats gave us an 86.67% confidence score, an early red card for Parana’s Diego Reynoso and Ferro looking dominant – it looked a dead cert. 45 mins and 18 shots-on-target later, the full time whistle blew to give The Panel our first loss in 6 tweets. Damn you Yair Iván Bonnin (I am assuming Atlético Paraná’s keeper got man of the match).

Anyway, point being, we will tweet losers. The real trick is minimising those losers. We see tons and tons of twitter tipsters running £10 – £1000 challenge bets with heart on the sleeve bets posted and their followers getting carried away when a run of bets comes in. This inevitably leads to your own confidence gaining, resulting in sloppy betting. Then, it is all gone – and it’s time to start the next challenge.

The panel are different. We stay true our principles and place bets based on our constantly improving betting score.

iRobot

In the technology sector, we are seeing the effect of higher computing processing power, vast volumes of data being captured and a growing range of open tools for data manipulation driving innovation. It is now becoming much easier to deploy automated modelling with increasing sophistication in predictive analytics, machine learning and artificial intelligence. And you will read the headlines:

  • “Robots will eliminate 6% of all US jobs by 2021”,
  • “Boston Consulting Group predicts that by 2025, up to a quarter of jobs will be replaced by either smart software or robots”,
  • “A study from Oxford University has suggested that 35% of existing UK jobs are at risk of automation in the next 20 years.”

A consistent message telling us humans are being replaced by machines and if you have a job that consists of repeatable tasks, a robot one day might do it instead. You only need to look at supermarket’s self-service checkouts, smart meters in the utilities sector, or most impressively, Google’s imminent roll out of driverless cars to show just how far we have come in such a short space of time.

Now, no matter how much I enjoy waking up on a Saturday morning and scouring the internet for 2 hours trying to pull together a range of football stats to make informed betting decisions, my girlfriend certainly does not! Ultimately, a human performing these repeatable tasks is 1) time consuming and 2) leads to errors.

So when the tweet notification comes pops up on my phone:

OVER 0.5 GOALS BET!

#Almagro 0 #AtléticoParaná 0

Shots On: 0-5

Bet Score = 68.89%

#2ndHalfGoals

The factors go through my head…What if Ferro Carril Oeste have drawn 0-0 two of their last four games? Or that Atlético Paraná are better at keeping clean sheets against the top teams? Or that Ferro Carril Oeste struggle against teams outside of Buenos Aires? Or that Ferro Carril Oeste’s ageing stricker Luis Ángel Salmerón hasn’t scored since returning from injury? Or that? Or this? Or that? I don’t have time to Google everything.

This is why our over 0.5 goals betting score is incorporating more and more automated data factors to produce an ever increasing accuracy on our confidence rating. It is driven from multiple stats from multiple sources. We have tracked nearly 2000 games and we are constantly integrating new factors all of which are seamlessly automated into the algorithm. Think of us as your driverless car – we are doing the hard work so you don’t have to. We are currently running at close to an 80% win rate and being on board with The Panel will help you one day swerve the Atlético Paraná vs Ferro Carril Oeste’s on the road.

One’s to watch:

Norwich vs Huddersfield (Friday 7:45 – Championship)

Starting with Friday’s game at Carrow Road. Both teams love a goal. And both teams are starting to gain a bit of form again after dropping off from a strong start. Most recently, both won their last game 1-0 with goals coming in the second half. I’d be surprised if an automated tweet isn’t sent if this is 0-0 at half time – would be a great start to the weekend!

Stoke Vs Leicester (Saturday 3pm – Premier League)

I’m predicting a goal for unpredictable Leicester here. Last weekend’s smashing of City was followed up with a loss at Bournemouth and after a slow start to their Premier League campaign, they need to start picking up points. I’d expect Stoke to keep it tight in a cagey affair but the 7th and 8th best players in the world (Mahrez and Vardy!!!) should have enough quality. But in the second half!

Al-Taawon vs Al Ahli (Saturday 4:45 – Saudi Arabia)

Al Ahli like a goal. 27 scored in 12 games sitting 4th and 5 points off top. However, Al-Taawon shouldn’t be a walkover as they have, in the last two games have kept it tight until 45th and 50th min respectively (winning 1, losing 1 1-0 both times). And with Saudi Arabia at over 85% win rate, if this is 0-0 at half time, I can’t see past an Al Ahli goal. Let’s just hope Asamoah Gyan has still got it (2 goals in 4 starts).

Alex@gamblingpanel

https://www.theguardian.com/technology/2016/sep/13/artificial-intelligence-robots-threat-jobs-forrester-report

http://www.bbc.co.uk/news/technology-33327659

The South American Elephant in the Room

So by now you would have an initial feel for what the panel is about, perhaps even individual members and you will have formed an opinion on how the in-play service is working out for you.

Well I would say the feeling’s mutual, as by now after receiving a twitter alert you must’ve at least once dropped your phone in a panic, loaded up **any betting website**, scroll way down, try to guess the country of the league, mutter to yourself to reduce the stake on the bet slip, then shrug and hit place. Not to mention a loved one / friend / samaritan telling you you’re being rude at meal times.

This week we’ve seen a high success rate in familiar leagues throughout Europe – SuperSaturday comes to mind, Livingston’s mass media coverage victory against Albion in the dying embers, as well ‘that’ goal from Danilo for Udinese. Seems we are creating some memorable moments already out of games that would only get covered by Iain Dowie. But we can’t ignore some rain on the parade from a South American direction through the night. Going back to the Weekly Stats, we have countries like Brazil and Argentina scoring highly in the region compared to other countries but we’ve seen the cliché technicians can behave more like 1980’s Italians and quite content to take a bore draw. So what’s going on? Quite simply the sample of matches tracked continues to grow which will gradually hone our prediction. Remember we are covering an extremely broad span of leagues from the start and slowly we can differentiate our treatment of them.

Keep an eye out for the country’s success rate in the weekly stats and consider the statistics you like to check in priority on receiving a notification – and let us know how you get on. It’s possible to work towards more granular breakdowns within the country and league and based on feedback, weightings of stats could be tailored accordingly. We could one day see those drab Argentinian (30-team league?!) possession drills on the horizon before they’ve even put the cones down.

No surprises that I’m keeping amongst the European leagues for my Ones To Watch this week and are rather close to home:

One’s to watch:

Brighton vs Leeds (Friday 7.45 – Championship)
This near top of the table clash is being billed as a big Championship game on a Friday night. Leeds don’t start the quickest, averaging first match goal at 44 mins and the seagulls are a bit quicker at 28 mins. However, Brighton’s last match was 0-0 and I can see this one starting in a cagey manner in front of the cameras until they manage to find the net. The teams are also a bit close to each other in the table (2nd vs 4th) but both possess the quality to break through.

Swansea vs Sunderland (Saturday 3pm – Premier League)
This bottom half of the prem match-up pits 2 sides who are on the road to recovery against one another. Sunderland are faring better and better with Defoe benefitting from Anichebe’s introduction up top while Swansea, off the back of a 5 nil drubbing, will be winded and without their usual fluency. The Swans have also been level at half-time in four of their previous six home games which bodes well.

Newcastle vs Birmingham (Saturday 3pm – Championship)
What a shock to the system up at St James’s. Untouchable at the start of the season but 3 losses consecutively shows they are no longer feeling invincible. I suspect this to be a tactical display from Rafa knowing he can’t afford to concede early against a decent Brum to get the league leaders back on track. The fact that both red cards the magpies incurred from the last match have been overturned will pack them full of motivation to get the result here as well. Birmingham would’ve thought they were on a decent run being unbeaten in 4 but were turned over convincingly against Barnsley last match so expect this to sharpen focus and hold up Newcastle.

Mike@gamblingpanel

The Legend of David Htan…

What a week it’s been… London took Black Friday too literally by having a black out, Eric Bristow threw his full support behind the rainbow laces campaign, and David Htan became a household name, (well in The Gambling Panel’s house at least).

What better way to spend your Saturday lunchtime than nervously praying for a goal in the Myanmar vs Malaysia AFF Suzuki Cup grudge match. In the 89th minute of this tense game David Htan secured himself a place in The Gambling Panel’s hall of fame by smashing home a late winner. To think, he was almost denied this acclaim in a first half that contained a missed penalty and the woodwork being struck twice. Fortunately for Htan and The Gambling Panel the half finished 0-0 and the rest is history.

It was a week of late winners with 8 goals being scored from the 84th minute onwards. This goes to show that if you’re willing to hold your nerve and chase the bigger odds, then the higher returns are out there. But how do The Gambling Panel pick their games to bet on?

Of course it all starts with the tweet, it could be tempting to rush straight to our favourite bookies and lump on as soon as possible, but as you’re starting to learn, that is not how The Gambling Panel work. I’m not saying there are any problems with this method but this is football and it’s full of uncertainty so we want to give ourselves the best chance to pick out those winners and make that all important profit. Once the tweet is received The Gambling Panel start looking further into the game, the betting score has already given us a starting point and takes into account a wide variety of in-play stats, but what about the team’s involved. We look into league position, goals scored and conceded, head to head record, and current form. When we can we also like to look into the importance of the game; will the result have a big impact on the title race or relegation battle, are the teams rivals, has there been a lot of media hype. But the most important thing we look out for is how often do the two team’s draw games 0-0!!! The ideal scenario would be a good top half team vs a weak bottom half team where both score and concede shed loads and contested a 5-4, 3-2, 3-3 in previous games. How that game would be 0-0 at half time I don’t know, but this equation would be Gambling Panel heaven. Obviously we’re not going to wait until only this criteria is met before placing a bet, but you see what we’re getting at.

In my ones to watch this week I’ve tried to build some of this criteria in, so let’s hope we get some half time 0-0’s and then once the tweet is received its over to you; dive straight in if you choose, carry out your own research, or play chicken with the odds. Which ever way, good luck and let’s beat those bookies.

One’s to watch:

Nottingham Forest v Newcastle (Friday 19:45 – Championship)

A game that looks like a both teams to score certainty, this is definitely one we’ll be hoping will be 0-0 at half time. Forest score and concede for fun (32 of each in 18 games) and prior to this week Newcastle were flying and have been scoring 2 goals a game on average. Neither team have had a 0-0 this season and in the pub after work on Friday we’ll be spending our winnings as quickly as we made them.

Stevenage v Doncaster (Saturday 15:00 – League One)

This game ticks a few of the extra boxes we look for, 14th vs 3rd, averaging 2.9 and 3.5 goals in games played this season, and only one 0-0 between them. Both teams are on a good run of form so I expect them to both have a go. It looks to be a certainty for goals.

Carlisle v Rochdale (Saturday 15:00 – FA Cup)

Cup games are always unpredictable and can at times can be cagey and I’m expecting something similar here. Neither team will want to fall behind early with the potential of a money spinning 3rd round tie within touching distance, but with both teams looking for promotion in their respective divisions neither will want a replay. It is because of this that I feel this game is a prime selection. There have been an incredible 60 goals in Carlisle’s 19 league games and that doesn’t include their 5-3 win in the last round. Rochdale aren’t far behind with 52 goals in their 20 league games and none of these have ended 0-0. Look out for a tight first half before both teams open up looking for that dream ticket to one of the Premier League big boys.

Erzgebirge Aue v VfB Stuttgart (Sunday 12:30 – Bundesliga 2)

Sunday lunchtime takes us to Germany where we’ll be looking to shake off our hangovers with an early winner. Erzgebirge are struggling at the bottom of the table and Stuttgart are riding high. Both are averaging close to 3 goals a game at the moment and neither have drawn 0-0. The one concern is Stuttgart, they have been flying out of the blocks in recent weeks and scoring lots of early goals but there seems to be a trend forming. In games where Stuttgart have not scored in the first 20 minutes, it has then taken them till early in the second half before making the breakthrough. Let’s hope Erzgebirge hold firm early on and this trend may continue.

Andy@gamblingpanel

The Weekend Preview

First rule of the over 0.5 goals market… Do not bet on women’s international under 20’s football.

Second rule of the over 0.5 goals market… DO NOT BET ON WOMEN’S INTERNATIONAL UNDER 20’S FOOTBALL.

Third rule of the over 0.5 goals market……. You get the point.

Wait!! What!! Goal!!!!!!!!! Scrap that. Bet on everything!! ESPECIALLY women’s international under 20’s football (do not bet on everything. When the fun stops and all that…)

Another winner for The Panel. Admittedly, The Panel’s algorithm and extortionately high betting score could not of predicted Ye-Ji Namgung of South Korea Under 20’s would have the composure of an anesthetized Andrea Pirlo to dispatch the penalty so confidently. However – it could highlight to us there is an extremely high probability there will be a goal in this game.

Since the surge of the ever increasingly popular in-play markets, I’ve blindly followed multiple online ‘tipsters’ yelling at me from behind their keyboards “BET NOW”, “CHALLENGE BET 478”, “OVER 10.5 GOALS IN THIS GAME, GUARANTEED”, BOOOOOOOOOM”. Of course Kakamega Homeboyz (genuine Kenyan team…) vs AFC Leopards (again a genuine Kenyan team) finished 0-0. The stats obviously suggested it at half time, the league position proves neither team are dominant in the goal scoring department, so why did I assume Big Barry’s Bets (not a genuine tipster, nor Kenyan football team) knew something else that we/the live in-play stats or the league table didn’t! They’re playing in Nairobi on a Tuesday morning for god sake! Who the hell is motivated on a Tuesday morning to do anything!!

Luckily, The Panel has a resolution. We use stats, lots and lots of stats. We no longer need to rely on Barry from Kettering and his dodgy named Kenyan teams (unless they’re betting score is greater than 50% obviously). The algorithm will produce a betting score based on stats for any game that is 0-0 at half time and publish onto our Twitter feed (@gamblingpanel) any where the score is greater than 50%. Obviously the higher the betting score, the greater chance there is of a goal in that game being scored. This still does not ‘guarantee’ winners, but with over a years worth of stats, the vast range of live data now available and the 80% win rate we’re currently running at – it’s a better option than relying on the wi-fi signal from Barry’s parents basement.

(No Barry’s were harmed in the writing of this post).

 

Ones to watch this weekend:

Unfortunately we can’t guarantee these games will appear on the feed this weekend (probably because they will all be 4-4 at half time) but if they do, based on the stats we have… Lump on!

Note – these are value selections where the half time price in the over 0.5 goals market is likely to be between 1/4 – 3/10.

Derby County v Norwich (England)

I know what you’re thinking ‘Derby F***ing County’. And rightly so, they’ve been woeful in the scoring department and would’ve been nowhere near this list a month ago. How things change. Derby’s last three games have averaged 3.33 goals per game, Norwich’s 4.33. If my maths if correct – you can expect a 2.2 – 1.7 Norwich win. If this is 0-0 at half time and does not meet the minimum betting score criteria, I’ll eat Steve McClaren’s brolly.

Dnipro v Shakthar Donetsk (Ukraine)

A bit Michael Schumacher here… Off piste. Too soon…. Sorry. The formula for this one is simple, Dnipro + Shakthar / arctic conditions x previous meetings = GOALS. These two have averaged 4 goals per game in their last 4 meetings, had only one 0-0 between them this season and success rate in the over 0.5 goals market for Ukraine is 92%. Chicken dinner is on the menu after this one.

Hearts v Motherwell (Scotland)

The Panel will have their very own Chris Kamara for this fixture. Whilst three quarters are sitting in their pants eating Wotsits from their belly buttons, I’ll be their bumbling, red-card missing, unbelievable Jeffing, Lionel Ritchie lookalike live feed from Tynecastle. Scotland has been a kind mistress for The Panel in recent weeks and I’m hoping this fixture is no different. With a win ratio of 85% and Hearts’ previous three fixtures ending 3-3, 2-2, 3-3 – I’m expecting goals, and hopefully all in the second half….

FBC Melgar v Sporting Cristal (Peru)

A real Panel favourite this one. FBC Melgar make it into our Champions League positions as third best performing team. Love a 0-0 at half time, love a 1-0 win, perfect for The Panel. Usually steer clear of 1st v 2nd but confident both of these teams cant paper over the Grand Canyon sized cracks of these leaky defences. Expect 45 minutes of frustration, expect red cards, expect late winners, expect money in your account.

Jay @gamblingpanel

Football’s Statistical Revolution

A shift from a ‘gut instinct’ approach, to a statistical approach

The well documented fairytale that saw Leicester City crowned Premiership Champions last season showed the world just what is possible in a David vs Golliath fight against wealth of the modern ‘superclub’. Stats.

English clubs have seen over the last few years how European clubs are taking advantage of the unbelievable Premier League TV riches. An innovation was required. Clubs like Leicester City needed to avoid paying the likes of £30m here or there for rookies, washed-up internationals and mid-table passengers. “There is now one price for English Premier League clubs and another price for everybody else,” one agent said. You only need to look down the road at Aston Villa to see how badly wrong you can get it. 

So how do you compete with the turbulent market? Rob Mackenzie, Head of Technical Scouting at Leicester from December 2011 – February 2015 is credited for a change of approach in Leicester’s scouting framework. “An approach that encapsulates statistical analysis combined with live observations in my opinion has to be the most cost effective, efficient and accurate way to assess a player” stated Mackenzie. 

With this methodology, Mackenzie, alongside Ben Wrigglesworth and Steve Walsh as Leicester’s statistical-focused scouting team have pulled in players such as Jamie Vardy from Non-League, Riyad Mahrez and N’Golo Kante from Ligue 2, and a host of Premier League rejects. They were all analysed and brought in to fit a system, a style of play, based on the statistics gathered. 

The Leicester story encouraged The Panel that a similar strategy could be adopted to aide punters in the betting market… 

The Panel History 

The Panel started over a year ago as a pub conversation and subsequent WhatsApp group in which us four guys were convinced that with a meticulous, cautious mindset, we could exploit certain betting markets. Our aim was to ‘beat the bookies’, in our own Leicester fairytale, where a statistical approach is pivotal to its success… 

We are living in a world in which so much data is generated, yet it is hard to pull together and digest. This is no different in the football environment with tons of hard stats, opinions and supporting data being generated and published on the web. So we began by analysing different markets and testing what factors (data sources) influenced different markets. Whether it was purely league position, league form, home form, away form or the more complex data factors such as player injuries, weather conditions or media sentiment – we used anything we could get our hands on to make informed decisions on a bet. 

We have had our success stories, and our failures, but all was used to build knowledge of what works and what doesn’t. 

With full time jobs to contend with, pulling this kind of information in a manual way was difficult to fit in. So we took one of our most successful markets, and used that as a test case for winning against the betting sites… 

Goal “Guaranteed”

In-play betting became a favourite way of using statistics to make a betting decisions. It is such a fast changing situation that you can exploit generous odds on particular markets on particular games. 

We started to use a variety of statistics to create a betting score against every 0-0 at half time as to how likely there is to be a goal. The key was automation. Our automated script pulls together all the sources of information we need to calculate the score and post an email notification when a game to bet on is available. Based on that algorithm, we tracked every game that finished 0-0 at half time and published games where the betting score was greater than 50%, the results were fascinating. 

Across 1800+ games, from 40+ countries, we were experiencing a win rate of just over 78%. Now, when you can get odds of 1/2 – 1/3 (If betting between 45mins-60mins and avoiding teams like Barca!), with a 78% chance of winning – with discipline, you can beat the bookies over a sustained period. 

We want to give you the opportunity to follow our bets on this market so have created a publication to Twitter than you can follow and, as with Rob MacKenzie at Leicester, use the statistics as a base to make an informed decision. 

Alex @ ThePanel 

http://www.optasportspro.com/about/optapro-blog/posts/2013/blog-rob-mackenzie,-head-of-technical-scouting-at-leicester-city/ 

http://www.mirror.co.uk/sport/football/news/special-report-amateur-premier-league-8535701