I’ve been working on my League prediction. I don’t just want to use my gut. I want to use a method, that can be refined, which uses a set of well defined criteria and gives me a fairly reasonable prediction of League position. It’s a problem which I don’t think can be solved. However, we can get close and I want to keep working on it and getting it closer.
For right now, my model doesn’t work! Or I haven’t crafted a model that works yet. So, instead, here are my gut predictions.
Now, you have to understand that my “gut” is actually my own database of stats which I have compiled over the years, which track things like the number of shots a team has created and faced – which is pretty well refined and includes a reasonably accurate expected goals model. My “gut” also knows which teams have spent money, which haven’t, and which teams have historically done well.
So, it’s a “gut” prediction only because I don’t have a model, yet. But, I need this prediction to use as a baseline. And here it is:
- Man City
- Man U
- Tottenham
- Chelsea
- Arsenal
- Everton
- Liverpool
- Leicester
- West Ham
- Bournemouth
- Southampton
- Watford
- Crystal Palace
- Stoke
- Brighton Hove Albion
- Burnley
- Huddersfield
- West Brom
- Swansea
- Newcastle
What makes predicting the League comically difficult is that 14th and 15th place are basically interchangeable. There is so little difference between these clubs that Stoke, for example, could easily finish 10th just as well as 14th. Instead of predicting each exact position, I think I’m going to have my model try to predict “groups”.
There are basically four groups in the 20. The title challengers – in this case, the two teams from Manchester – who spend money like no other; there are the middling group of Tottenham, Arsenal, Chelsea, Everton, and Liverpool – who spend a lot of money and have huge fan support but these teams all have significant weaknesses which may prevent them from seriously challenging for the title; there are the unwashed masses between Leicester and Stoke – the clubs that the top clubs have to beat in order to win the League; and there are the relegation candidates – clubs which are either small or horribly managed (or both).
With that in mind I have Man U and Man City as title challengers, with Tottenham a strong third candidate. Here I would like to include a graphic which I think illustrates how insane the spending has been for those two clubs and give a reason why I think Tottenham are the third candidate for the Premier League title:
I got this data from transfermarkt. Transfermarkt is often ridiculed but there is no other reliable place to get all of this transfer data along with a crowdsourced market/player valuation system.
Total spend (net spend) among the 20 teams in the Premier League between summer 2015 and summer 2017 is £1.79bn. Man City and Man United have combined to spend £760m of that alone. Meaning that the remaining 18 clubs have spent £1,030bn or just about £20m per club per year.
I also looked at the change in value over the same time period. At first I did this just to see how crazy the League has become but then I noticed that there is a sort of hidden datapoint in that information: spending isn’t the only way to grow the value of your team. That’s where Tottenham come in. They are the only team to have increased their value without spending. That’s all down to the valuation of Eriksen, Alli, and Kane. Those players were found on the cheap and have been made into stars.
Arsenal are also an interesting case here: their valuation has gone up almost exactly by the amount they have spent in the transfer market. Unlike the two Manchester sides, who have combined to spend £587m more than their rise in market value, Arsenal seem to have spent wisely. Though the question there for Arsenal supporters is, why not spend more and make themselves into a title contender?
Chelsea have been a surprise in terms of spending. Once ridiculed for “parking tanks and firing fifty pound notes” Chelsea have stopped wasting money and instead, through special deals selling their players at a highly inflated prices to China, are basically a break-even team in terms of transfer spend.
Arsenal, like every other team here, rely heavily on the value of one or two star players; Alexis, for example, is over 10% of Arsenal’s total team value. So, team value could still change pretty crazily over the next two weeks – if Arsenal sell Alexis and don’t replace him, for example, or if Everton buy a top quality forward (not Gylfi Sigurdsson) these two team’s value could change and along with them my prediction for League finish. Another big question mark is Chelsea, who could lose Hazard and Costa. If Chelsea lose Hazard, Barcelona are reportedly after him as a replacement for Neymar, I would confidently put Chelsea out of the top four and probably even as low as 7th.
This is to say that my “prediction” isn’t much of a prediction yet – there is a lot of volatility in the League… except the top two. I think that Man U and Man City, based on the fact that they have Guardiola and Mourinho as managers, the stats from last season which showed Man City as the very best team in terms of both creating chances and limiting opposition chances, and Man United’s stellar defensive stats along with their acquisition of Lukaku, plus the just astonishing spending these two clubs have done in order to build their squad depth, makes them the presumptive title challengers.
Sorry to the one Arsenal supporter who still reads my blog and who thinks that Arsenal are title challengers. I just don’t see it: specifically because Arsenal’s defense – the number of shots that Arsenal have allowed, the big chances conceded, and the fact that Wenger still hasn’t decided to make a defensive midfielder who can play the ball a priority – puts them firmly in the 4th-7th place group. And even that isn’t an assured prediction: if Arsenal don’t get 100% hunger from Alexis, Özil, and Ox this season I think Arsenal could even finish below Leicester.
I’ll return to this prediction stuff after the transfer window has closed. Hopefully, I’ll have a model in hand by then.
Lucky for you I have a TransferMarkt model ready to go. It uses two variables:
ln_tm_homo = ln(TM) – average[ln(TM)]
ln_tm2_homo = [ln(TM)]^2 – average{[ln(TM)^2]}
TM refers to TransferMarkt team value. Average refers to the league average. And the model itself:
E(PPG) = -.3203079*ln_tm_homo + .0821647*ln_tm2_homo + 1.371233
(n=900, R2=0.614)
E(PPG) is expected points per game.
How did you arrive at those multipliers? Have you tested this model? How closely does it predict league position?
Sorry, just saw your r2 is 0.614. That’s not bad.
It’s the most important component in any multiple regression model for league predictions I have tested. Hands down it beats predictions based for example on expected goals provided by understat.com.
n=900? What are you including in your dataset? Is this for multiple years for a single year, or a single year across multiple leagues? Or a panel of multiple years across multiple leagues?
A question for both of you – how are you accounting for inflation? I’m guessing that if a team keeps the same personnel over the course of a year and performs at exactly the average, their value will increase just based on transfer inflation. Hell, Mbappe has basically doubled in value from June-August (not sure if this is reflected in Transfer Markt)!
Ehhh… I don’t know yet. In this graphic I just took the average increase. Lame, I know.
Not lame at all. Getting into football inflation is a tricky thing. And if you think about it, a lot of that inflation is at the top end of the market. Has Kieran Gibbs’ value inflated at all? I’d argue that you could’ve got 15m for him 2 years ago as well. A team probably sees lots of inflation in values for players who have played and played well – if we sell Alexis for 50m, it’d be more than we paid in his prime – but a lot who don’t play or play badly – Gibbs, for example.
If you assume all teams have this going on, and all therefore inflate/deflate at the same rate, then your approach, which looks at difference, is just fine. It’s not exact, but it is indicative.
It’s a panel, yes. Included:
PL: 2013-2017
La Liga: 2005-2017
Bundesliga: 2007-2017
Serie A: 2007-2017
Ligue 1: 2007-2017
The even simpler model is:
E(PPG) = .429*ln_tm_homo + 1.371
(.012) (.009)
n=900, R2=0.582
The interpretation is that a team with average ln(team value) is expected to get 1.37 points. An additional 10 % team value, is estimated to get you 0.041 extra points. I found the relationship to be independent of league (F8,890=0.89, p=0.53).
I meant for the original model to be homogeneous of degree zero, so that a 1 % increase in the team value of each team, would make the predictions unchanged. I see I have messed that up with the quadratic term, so I will need to figure that out.
Thanks for explaining! Intuitively it makes a lot of sense.
My prediction, based on a very scientific method called the ‘eye test’, was what I predicted on Jonathon’s column just before the season opener:
I think United’s squad has the maturity to win it all this season. Lukaku is primed to have a phenomenal season while Mkhitaryan and Pogba have both had a season to get used to the league and the manager. Matic is a EPL proven player who also has worked under Mourinho. I believe their experience will see them through despite City having outspent them. City also always seems to be one Vincent Kompany injury away from things going south.
As for Arsenal, I said:
Without an upgrade in central MF: 4th to 6th depending on injuries (for both us and others) and how we rotate for Europa league
With an upgrade in central MF: 3rd to 5th depending on those same factors
Tottenham has an excellent group of 12-13 players but playing in Champions League requires more from your core group of players and given their small squad size, I can see them struggling. I still expect them to finish top 4 though.
I believe Chelsea and Liverpool have the highest chance of dropping out of the top 4. Liverpool, because their defense is actually worse than ours. Chelsea, because they seem totally unprepared for the new season and will spend a chunk of the season trying to catch up while also playing in the CL.
So:
1. Utd
2. City
3. Tottenham
4. Arsenal
5. Chelsea
6. Pool
One last prediction: If we sell Sanchez, we will not make top 4 and if we sell him to City then they will be champions.
Admittedly gut-related, but also being a close observer of the game, I’m going with…
City
United
Arsenal
Chelsea
Spurs
Intriguing about Tottenham. They seemed to weakened rather than strengthened. Is injury (and consequently squad depth a variable worth considering?). Spurs have a strong first 14 and not much outside of it, we don’t know how they’d cope with the prolonged absence of Kane and or Eriksson.
I think their biggest asset is their manager. Of course if Kane gets injured for lengthy period, that changes things but that’s a pretty extreme scenario. Every team has one or two players whose prolonged absence will surely be reflected on the standings (Lukaku for United, Kante for Chelsea, Kompany for City etc.). I just think Pochettino is a very shrewd manager whose cunning goes beyond tactics. He also seems to have an eye for identifying young talent and then developing them well. After years of failure with many managers, Spurs have struck gold.
Kane was hurt a lot of last year and they still coped. Spurs have been the best team in the league over the past two seasons combined. Not sure why that would change just selling off spare parts.
Spurs play their home games at Wembley this season. It will be more difficult to implement their high press than at White Hart Lane, a smaller pitch. They should still be able to beat the weak teams at Wembley, but I expect them to struggle more against the good teams (see their poor record at Wembley in the group stage of the Champions League last season).
Without a doubt squad depth is important! I would argue it is one of the most important factors. That is why I predict Arsenal to do far better than everyone suggests. Just look at the two previous league winners: Leicester and Chelsea. With no european football they were able to field nearly unchanged 11s week in, week out. Whereas teams with very thin squads that are forced to rotate after Europe historically struggle. I personally cannot see tottenham anywhere in the top four this season. They are one or two injuries away from fielding multiple unknowns all across the team.
In theory Arsenal should be able to rest and rotate smartly in the first half of this season in ‘easier’ Europa group games, thus fielding a largely unchanged 11 for the League. Just as we saw this Friday; while not world-beaters, the ability to bring on a ‘Europa squad’ of proven internationals in giroud, walcott, iwobi, welbeck et al off the bench after 60 mins is significantly better than the majority of the competition.
I think the top three will be very tight, maybe seperated by 3 points. While Man city’s attack is phenomenal, I see their midfield and defensive depth as inferior to Arsenal’s. Man u’s 11 is good and they are just stacked with the huge, nasty bastards that win you the league. I predict good things from Arsenal but unfortunately have the feeling that mourinho will grind out Fellaini-based draws against the top 7, walk all over the rest of the league and just do enough to have the last laugh.
Man u
Arsenal
Man city
Chelsea
Tottenham
Liverpool
Everton
We had good squad depth last year, was out of the CL by the second round and could still only manage 5th. I am not convinced we have addressed our defensive/MF issues. Agree that squad depth is important but I think having a quality first team of 13-14 players is even more so. I see Spurs making a run in the league after they drop out of CL. Squad depth may be the reason why they finish 3rd or 4th rather than top 2 but “nowhere near the top 4” might be wishful thinking. They have been the most consistent team in the EPL for two seasons now. I think some fans are being over-optimistic about the Xhaka-Ramsey partnership. I will be very happy to be proved wrong.
I think “actual players available” is a hugely important consideration.
‘Actual players available.’ Available is the key word here. My point is that all of last year’s Champion’s league qualifiers (everyone bar Everton) will not feel able to rotate heavily. Arsenal however, due to their huge squad, can afford to rest and rotate in the Europa.Therefore minimising injuries and mid-week hangovers and maximising the probability that they will have a consistently fit and available first 11 each week.
We have the same squad as last year barring the two new boys.
We are poised to offload some deadwood and probably lose our best player.
Thus our ‘squad depth’ has not improved and nor will our position from last year.
No, I’m afraid not. My point is that our schedule August-January is far easier (in theory) than last year, due to the Europa league. Whilst our competitors for the top 6 all have Champion’s League, and split priorities, Arsenal can actually prioritise the league for the first time in two decades. Combine this with two good signings and a more settled, experienced team. I am very confident that Arsenal will improve on last year.
I like your optimism, and tomato soup, as I think / hope Liverpool and Tottenham will have the distraction of CL football, but much will depend on how Wenger deals with our own distraction of Europa League football. Are we going to take the competition seriously? If so, we’ll suffer the same fatigue in the league as our rivals.
More fatigue: there are more games and the matches are further away, plus we are the overwhelming favorites to win so we will probably go way deeper into the competition than we did in the Champions League.
Whoa – big call on the bottom three. I’m not a fan but I will give props to Benitez for sticking with Newcastle and think he can figure out a way to stay up. Clement and Pulis are good proven managers as well. I think this might be the year the wonderkid at Bournemouth gets a rude shock and I’m not sure West Ham have recovered from last year’s debacle. But no modeling going on here.
My entire bottom 6 are really interchangeable. I basically had to make a call so I did.
Hey, you called me out in your blog!
I did?
Agree that Arsenal don’t have the midfield and defense to win the title. I think we look good for fourth place, though, provided we don’t suffer unduly from injury crises. We’ll need a bit of luck, too, in the sense that our rivals will need to struggle a bit, especially Tottenham (new stadium / CL football) and Liverpool.
I would like to see a model that captured the randomness that is a part of every game and the potential for snatching defeat from the jaws of victory. After all which model predicted the 2015/16 season with Chelsea imploding and Leicester winning?
I agree city and united are in the pole position, but literally anything can happen. Including us winning.
I will save the fatalism for March.
Man City do not have a reliable goalkeeper. Man United have a manager who has lost his mind and players who aren’t capable of standing up to the grind of a 38-game season. Tottenham are Tottenham and will always fail. Liverpool are forever rebuilding. Everton are far from ready for any kind of a challenge. Leicester were a one-off, albeit one that threw all statistical models out the window. And Chelsea will miss Captain Terry more than you know.
Arsenal will win the League.
The problem I have with predictions after one game is that there’s a massive gap in the data, called the summer.
So many unknowns at this point.
If Arsenal can get back to gaming Xg (without Cazorla) then we have a chance, if not, meh…
It’s interesting that we’ve had two Italians basically exploiting that technique to win the league while the heavy pressers fell short.
I’m sorry to be off topic, but if you’re a foreigner and have a chance to study in the US, would you? It doesn’t seems as save as other country, looking at the recent incident and the POTUS conference. Although the majority of US citizen is good i don’t think the nazis in the US is a minority either.
Yes would definitely recommend studying in US. Particulalry if your interest is science or engineering.
No, the US is not nearly as safe as other wealthy countries, mostly because we have a hige underclass and a lot of guns. But the real question is would *you* be safe from street crime and from racist crime. That depends on where you are from, where you live and what you do for fun. You probably would be very safe.
America’s shameful secret is that we are extraordinarily generous and kind but we’re too afraid of each other to reveal it in public.
Nazi’s are an infintesimally tiny minority. However, like all humans we are quite tribal at times. This creates ugliness.
Tottenham are very close to the two Manchesters in having the strongest starting lineup in the league and very close to Liverpool in having the worst second eleven. Champions League football is going to take too heavy a toll on them to challenge for the league, despite having the best center forward, best central defenders, and the best young midfielder in the Prem. But this is the last season they’ll keep that squad largely intact. Kane and Alli could double or almost triple the £100,000/wk wage ceiling at Tottenham. And even if by some miracle they keep both of them next season, useful squad members will want to leave for greener pasture. Walker has already left and is undoubtedly making more than Kane and Alli at City. Danny Rose announced his availability. Their fans better enjoy it while it lasts.
I say we will go far this year based on gut feel.
“Stats should be used to help us understand gut feel; if the model does not say so, rebuild model” – anonymous old fashion guy
Great stuff Tim, although I am always wary of any analysis that relies on monetary valuations as a proxy for quality. I’m wondering how a similar player-by-player analysis that substituted whoscored averages for transfermarkt fees would compare. It might kick out the same result, it might be wildly different. Would be interesting though – and I’m sure whoscored player ratings are just as arbitrary as transfermarkt valuations where no actual market price exists for a player.
On inflation, I guess one way to handle would be to apply discount based on the annual inflation rate of the mean transfer fee.
While I think the stats are quite useful, you omitted the need for a winning mentality ..thats what i saw against Leicester city.. Wenger, for the first time in his life, was able to grind out a result on opening day after falling short.. I think Jens will infuse that mentality in this squad ..
i believe arsenal have the talent to win the league but they need to be managed well. the most efficient move is to change the formation yet again. last season, arsenal went to 3 central defenders and that change helped all of the defenders look better. now, arsenal need to help the midfielders. there’s been so much talk of an “elegant beast” (can’t remember which of you came up with that term) but the acquisition of this mythical, unicorn-like creature is not going to happen. the way forward, go from two central midfielders to 3. it’s the most pragmatic approach and, like with the defense, it helps all of the midfielders look better, too. what i’m saying, yet again, is arsenal should switch from a 3-4-3 to a 3-5-2.
arsenal used to play with two center mids but went away from that for good reason when patrick vieira left. no one really plays with two center mids anymore so if you’re going to do it, your center mids better be amazing two-way players. arsenal’s best two-way players are alexis and coquelin. the obvious problem is one of them is not even a center mid and the other one is no where near the class of vieira. the practical solution is put a third body in there.
not only does the formation change help the midfield look more solid, but you can help the strikers look good as well. playing with two strikers eliminates the need for a center forward. all of a sudden, there are chances for players like perez and walcott and those players would likely shine as their speed would stretch defenses. like i said in preface, it all comes down to how well arsenal are managed. would wenger be so bold? unlikely but who would have ever thought wenger would go to a back 3? we’ll see.
“i believe arsenal have the talent to win the league but they need to be managed well.”
– Joshuad
I am in violent agreement with this but I have been so since about 2004. This has been 75% of our problem the last decade or so. That, and and leading the injury table for several of those of years.
Interesting idea but are you going to drop Ozil? He doesn’t work as part of a forward pairing because he simply doesn’t score enough goals. Maybe he bangs shots in while practicing all the time but in an actual game, how often does he scuff the shot? He certainly has the skill to score 15 goals, I think he simply lacks the mentality or confidence to be a big time scorer, despite Wenger’s encouragement. And while he can make up the numbers in a midfield three, his actual defensive contribution would be close to 0%, functionally making your trio a duo. Man City made a very attack minded 3-5-2/4-1-4-1 work because Silva and DeBruyne actually do contribute defensively. I’ve never been enamored with Mesut’s defending. When he played centrally in a 4-2-3-1, Ozil didn’t drop into the midfield but was part of the forward pairing defensively.
bro, i’d be willing to drop anyone to win the league.
i’d give ozil every chance to keep his place in the team and even start him. however, if i felt the need to go to the bench and ramsey or iwobi prove more effective, you bet your tail that i would drop ozil in a heartbeat.
the beauty of it all is that the truth would be there for everyone to see, which puts all the pressure on ozil. at that point, he either performs or he doesn’t. however, i’m not the one responsible for managing the team. if i were, i do believe i have the moral courage to make that call. no player is worth more than a bpl trophy.
Get what you’re saying. I’m just not sure that we have the necessary guile and creativity in the side without Özil. We already lack technicality in the team and Özil is certainly a technician. Does midfield trio of say Ramsey, Iwobi, and Xhaka have the necessary creativity and balance to compensate for Mesut’s absence. It’s a tough question. I would say I’d rather start our three best attackers-Özil, Alexis, Lacazette-and provide a solid platform, both in terms of players and formation, to support them. I can see your point though
I would not put Iwobi in MF. He is defensively not switched on enough and as good as he is on the ball, we need someone with a bit more speed next to Xhaka. I do think the side can be creative without Ozil though. A front 3 of Alexis, Giroud & Laca with support from Ramsey, Xhaka and a seriously upgraded box to box midfielder should create more than enough chances. As things stand however, I think we need to play Ozil.
Which is my point. Weaknesses notwithstanding, Özil’s ability to take out an entire defense with one outstanding pass is second to none in world football. You don’t take that kind of creativity out of a side without due consideration.
look, arsenal haven’t won a championship with his guile while other teams have won championships without it. i just believe it would be prudent if arsenal considered all of their options, including dropping him. i’m not saying they should do it but they should consider it. at the end of the day, it’s all about winning the championship.
I think we definitely need a third MF player on the pitch to give us more possession especially when we play against the top sides. IMHO, the best move would be to get another runner next to Xhaka who is more defensive minded than Ramsey. For sh*ts and giggles, let’s call this player Matuidi. That would give us a MF triumvirate of Xhaka-Matuidi-Ramsey, which I believe will bring out the best in Xhaka in a way Juventus brought out the best in an older & slower Pirlo in 2012 by using Pirlo-Vidal-Marchisio in MF. They used a 3-5-2 formation with Asamoah and Lichtsteiner as wing backs. However, I don’t think Wenger will ever go for this. The best we can hope for is a 4-3-3 in Wenger’s attack-first philosophy. Our full backs will end up playing as wing backs anyway and if Xhaka drops deeper to collect, we will be a team that’s 3-4-3 in attack and 4-3-3 in defense. I would be okay with that. At the end of the day, formations mean very little if you can’t defend as a team. That seems to be our biggest weakness.
I have no qualms about dropping Ozil. He is great to watch at times but he’s a luxury and to win the league we need a more practical approach. Still, if he can perform as one of the wide players, I don’t see a need to drop him.
What’s a two-way player?
Ha! Just read that Juventus signed Matuidi for less than £20 mil. Think we missed a boat there.
said player could also be coquelin or elneny. he doesn’t have to be world class if he’s intelligent and kinetic enough.
a two-way player is a player who contributes well to both the attack and defense. while coquelin doesn’t contribute directly to goals, he links play well and can go by defenders. arsenal tends to play well with him in the side.
Great piece Tim.
The bit that stood out for me was how you have Chelsea, who you call a ‘team with significant weaknesses’ in the same group with Everton.
What do you think has changed since last year, when they won the title, that now puts them in the same group with Everton?
I can understand the point you’re making about United/City spending overtaking Chelsea but I’m surprised to see Everton be placed in the same group as them in your model.
I agree that the Big Four have become the Big Two and that the five teams in the old big four are now six.
Other than that, I think all y’all’s love of maths has caused you to forget the faces of your fathers.
1. Arsenal
2. Some one else
3. Etc.
Thorough enjoyed The Dark Tower series but I hear the movie blows…
“The greatest mystery the universe offers is not life but size. Size encompasses life, and the Tower encompasses size. The child, who is most at home with wonder, says: Daddy, what is above the sky? And the father says: The darkness of space. The child: What is beyond space? The father: The galaxy. The child: Beyond the galaxy? The father: Another galaxy. The child: Beyond the other galaxies? The father: No one knows.
You see? Size defeats us. For the fish, the lake in which he lives is the universe. What does the fish think when he is jerked up by the mouth through the silver limits of existence and into a new universe where the air drowns him and the light is blue madness? Where huge bipeds with no gills stuff it into a suffocating box and cover it with wet weeds to die?
Or one might take the tip of the pencil and magnify it. One reaches the point where a stunning realization strikes home: The pencil tip is not solid; it is composed of atoms which whirl and revolve like a trillion demon planets. What seems solid to us is actually only a loose net held together by gravity. Viewed at their actual size, the distances between these atoms might become league, gulfs, aeons. The atoms themselves are composed of nuclei and revolving protons and electrons. One may step down further to subatomic particles. And then to what? Tachyons? Nothing? Of course not. Everything in the universe denies nothing; to suggest an ending is the one absurdity.
If you fell outward to the limit of the universe, would you find a board fence and signs reading DEAD END? No. You might find something hard and rounded, as the chick must see the egg from the inside. And if you should peck through the shell (or find a door), what great and torrential light might shine through your opening at the end of space? Might you look through and discover our entire universe is but part of one atom on a blade of grass? Might you be forced to think that by burning a twig you incinerate an eternity of eternities? That existence rises not to one infinite but to an infinity of them?”
― Stephen King, The Gunslinger
stephen king is weird; an acquired taste, indeed. i read “the stand” back in ’95. ironically, i still don’t know if i enjoyed it but i did read all 1100+ pages. what’s certain is that it’s the longest book i’ve ever read.
I agree with Arsenal requiring a defensive midfielder. Although, I am still hopeful of the defense picking itself up and getting more used to the 3-4-3 system.
All players need to play through pain and niggles and even though Kos has a more severe problem, I back him to give a good season playing once a week.
Holding had an under par game against Leicester but, we know how good he can be on his day. I back him to come good as well.