Manchester United attackers

Manchester United have raced out of the blocks in the 2017-18 Premier League season, logging two wins in two games and scoring eight goals without reply.

Such success has seen Fantasy Premier League managers flocking to United’s assets. Romelu Lukaku started the season with the weight of expectation behind him and it has only grown after three goals in the first two games. The Belgian has already had one price rise and his ownership is now over 54 per cent. However, since the start of the season, even more managers have been buying Henrikh Mkhitaryan, whose four assists have propelled him to second in the FPL midfielder standings. The first two also helped him to a price rise a few days before the Gameweek 2 deadline.

Embed from Getty Images

Following suit was Paul Pogba, formerly the world’s most expensive player, who shot to the top of the midfielder rankings and an instant price rise with a goal and two assists against Swansea City in Gameweek 2. Like Pogba, Anthony Martial has also found the net twice this season, but his appearances in the first two games have come as a substitute and that isn’t an attractive feature for most FPL managers.

Given the clean sheets accompanying Manchester United’s wins have made goalkeeper David de Gea, or the defence in front of him, an attractive addition to many FPL teams, there’s not much space left for the Red Devils’ assets. Therefore, the question for a lot of managers is who to pair with Lukaku in attack. I took a quick look at the early data on actions per 90 minutes and the results only serve to cloud the picture.

Manchester United attackers - actions per 90 minutes

Manchester United attackers – actions per 90 minutes

The sample size is tiny, so we should not put too much stock into this analysis. The results exclude Nemanja Matic because of his defensive posture and Martial because of his limited game time so far.

The first thing to note is that Mkhitaryan has some underlying numbers to back up his emerging role as the creative influence at United this season. Not only has he assisted half of their goals, but he is creating chances roughly twice as often as any of the other attackers. While he has been unleashing a few shots, they’ve mostly been from outside the area so it is no surprise to see him struggling to hit the target so far.

Lukaku’s shot profile is generally positive, with his proportion of shots hitting the target looking very sustainable, particularly given many of them are being hit inside the penalty area. That’s the good news. The bad news is he won’t sustain a 100 per cent Shots on Target conversion rate for long.

Pogba’s shooting profile is a little flat, which doesn’t bode well for sustaining his present rate of return if this pattern continues in future games. However, the French international tended to be a long range shooter in the Premier League last season, so it’s beneficial to his prospects this year that a greater proportion of his shots have been fired from inside the penalty area so far.

Embed from Getty Images

The interesting thing from this small sample is the performance of Marcus Rashford. The young England international has been shooting more frequently than Lukaku and doing so from inside the box nearly as often. All that is missing is the end product. He’s unlikely to be out of the goals for long this season if he can sustain these kind of performances. Priced at just 7.5m – 4.1m less than Lukaku – Rashford is certainly one to keep in mind.


Everything but the goal: GW6

My experiment to see what success, if any, can be had by selecting a player with good underlying statistics from a small number of games had it’s first triumph last week – but only for the control group. The test group, those that haven’t delivered returns in a small number of recent games despite good numbers, are still waiting to get off the mark. I assess players in this experiment as having either found their objective (goal, assist or clean sheet etc.) or missing it.

My test group selection last week was Shane Long. Although he started, the game time curse of being an Everything but the Goal pick struck again and he was substituted for Charlie Austin after 54 minutes having managed two shots, one inside the box and off target and one outside the box and on target. His objective had been to score and he went missing. After getting on the pitch, Austin promptly rattled off four attempts, including three in the box and two on target, and found the net. There was a goal for Romelu Lukaku, my control group pick, and he therefore found his objective. He managed to do it with his only shot of the match.

Embed from Getty Images

My Everything but the Goal pick this week is Son Heung-Min. Against Sunderland in Gameweek 5 the Tottenham Hotspur midfielder fired seven shots. Five of those shots were in the area, including two that were blocked. Two of his shots tested the goalkeeper. He also created five opportunities for team mates without gaining an assist, but his objective against Middlesbrough will be to score.

The control pick – a player with good underlying statistics who did convert last week – was close between Michail Antonio (eight shots, including five in the box and three on target, for one goal) and Alexis Sánchez. The Arsenal man is the selection because his shots, while fewer in number, were of better quality – three of his four shots in the box were on target (including both of his goals), as was his shot from inside the D. His objective this week is to score.

Season EBTG score: Found 0 – 4 Missing.

Season control score: Found 1 – 3 Missing.

Everything but the goal: GW5

This experiment is in danger of becoming as risky to player game time as my Over the Radar series was to player health last season. Last week my pick Philippe Coutinho was benched on his return from international duty despite Liverpool making special arrangements to fly him home. That’s the second week in a row my Everything but the Goal pick has been riding the pine at kick-off.

Everything but the Goal is an experiment to see what success, if any, can be had by selecting a player who has had good underlying statistics in a small number of recent games, but has failed to produce returns. Each week, I score the player from the previous week on whether they found their objective (goal, assist, clean sheet etc.) or whether they were missing again.

Coutinho managed two blocked long-range shots after replacing Daniel Sturridge in the 76th minute, but he was missing for the third gameweek in a row.

Embed from Getty Images

My pick for Gameweek 5 was close between Sturridge (five shots, three in the box and four on target in Ganeweek 4) and Shane Long, who gets the nod. Both carry some bench risk, but this experiment is about whether underlying performance indicates imminent returns so I feel I should pick a strong performer from the week before if they have a reasonable chance of starting. The Southampton striker had five shots, including four inside the penalty area and three on target, against Arsenal but failed to find the net. His mission against Swansea City is to score.

Last week my control pick – a player with good underlying statistics who did deliver the previous week – was Eden Hazard, but he also went missing. The control pick for Gameweek 5 has to be Romelu Lukaku after his monster performance against Sunderland. Of his nine shots, eight were in the box, six were on target and three rippled the net. I’m looking for at least one goal from Lukaku this week.

Season EBTG score: Found 0 – 3 Missing.

Season control score: Found 0 – 3 Missing.

Agüero replacements: Mid-price forwards

Yesterday I looked at the expensive forward options for Fantasy Premier League managers replacing Manchester City’s injured hitman Sergio Agüero. Today it is the turn of the mid-price forwards.

The choice in this price bracket is deeper than at the premium end of the field. Five options stand out to me as worth considering: Graziano Pellè (8.2m), Romelu Lukaku (8.4m), Bafétimbi Gomis (7.1m), Anthony Martial (8.4m) and Wilfried Bony (8.2m).

With more than a quarter of a million transfers in already during this international break, Pellè, the top scoring FPL forward with 52 points from 708 minutes played, has emerged as a clear favourite. Lukaku (49 points, 694 min) is another popular choice with over 70,000 transfers in. Despite being the prime candidate to replace Agüero in Manchester City’s normally potent attack, Wilfried Bony (11 points, 218 min) has only received just over 20,000 transfers in. After starting the season with four goals in four games, FPL managers have fallen out of love with Gomis (32 points, 628 min) and he has about 75,000 net transfers out this gameweek. It’s a similar story with Martial (25 points, 296 min), who has failed to score in his last two games and is now in 25,000 fewer teams overall than he was at the start of the international break.

Mid-price Agüero replacements

As I did yesterday, I have included Agüero’s statistics to provide a comparison and displayed the data in minutes per action to account for the wide variety of game time among these players.

Pellè’s stats are comparable to those of Agüero when it comes to shots inside the box, but in terms of the other metrics the Southampton striker does not stand head and shoulders above the other options over the season as a whole. Bony is ahead of Pellè in Minutes per Shot, Gomis in minutes per Shot on Target and Lukaku and Martial lead the way in Minutes per Goal.

Since I looked at Martial after Gameweek 6, the regression I was predicting has begun to kick in and the Manchester United forward has only taken two more shots. The Red Devils next two games aren’t easy but after that their schedule opens up though I wouldn’t say those fixtures are easy (eve, MCI, cpl, WBA, wat, lei).

Two weeks ago, I exposed the myth that Romelu Lukaku only scores against top teams and the Everton man kept up his fine form with a strike in his five-shot game against Liverpool. The Belgian has two tough games up next, but after that the fixtures look almost as good as they can get (MUN, ars, SUN, whm, AVL, bou).

Embed from Getty Images

With four goals and five assists in his last five games, Pellè is clearly a man in form. With a nice run of fixtures (LEI, liv, BOU, sun, STO) before a Gameweek 14 match against Manchester City, it is clear why the Italian is such a popular pick this week. The Saints striker’s Shots on Target conversion rate is a touch high. However, given the high percentage of shots he hits from inside the box, there might be a case for arguing it is the Shots on Target rate that is too low and not that the number of goals scored is too high.

Gomis is an interesting case given the common perception that he is out of form. He was hauled off at half-time in Gameweek 7 after failing to record a single shot. That game aside, however, his underlying stats have continued to tick over since I looked at his performance after Gameweek 4, even if the goals have stopped flowing. As Fantasy Football Scout member Balders pointed out, the four fixtures he blanked in were against some of the better defensive teams in the Premier League (wat, EVE, sot, TOT). With four good fixtures in the next six (STO, avl, ARS, nor, BOU, liv), it wouldn’t surprise me if Gomis rediscovered his scoring touch soon.

Embed from Getty Images

Despite some minor speculation to the contrary, Bony’s minutes so far this season suggest he is the first choice back up at Manchester City to Agüero. It is hard to judge a guy who has only played a significant amount of minutes in two games, but the data shows that per minute he is keeping up with the other options considered here in every metric aside from goals. The Citizens may be missing Silva as well as Agüero this week, but they would remain one of the most potent attacking sides in the league even if they suffered a 20 per cent drop in shooting opportunities created. And with three good fixtures in the four games Agüero is likely to miss (BOU, mun, NOR, avl), Bony could soon notch his first goal of the season. However, as Ruth pointed out in the comments on yesterday’s post, switching back to Agüero from Bony could be tricky unless you have two free transfers or a lot of money in the bank.

Romelu Lukaku

Some Fantasy Premier League managers call Romelu Lukaku “a troll” because they feel he lures them into their teams by scoring a couple of goals, then does nothing for several weeks. The week or week after they ship him out, Lukaku will score another goal, leaving the managers frustrated that they had him and sold him.

Embed from Getty Images

This season has been a textbook example so far. In Gameweek 2, Lukaku opened his goal scoring account for the season with a brace against Southampton. His success resulted in more than 300,000 net transfers in (NTI) over the next two weeks and an increase in his price from 8.0m to 8.2m.

Lukaku’s NTI remained positive for the following two weeks too, but after a fourth consecutive blank the tide turned and his NTI exceeded -50,000 in Gameweek 7. He repaid those who kept the faith with his second double of the season against West Bromwich Albion on Monday night, but it must have left those who had transferred him out frustrated.

Romelu Lukaku GW1-7 2015-16

The underlying stats for the Everton forward are solid. His goals from shots on target ratio is a touch high, but he is in the top 10 for all the shooting metrics indicated in the chart so continued returns look likey. He fires off two or three shots in most matches and takes most of them from inside the box.

We should also remember that Lukaku’s opening run of fixtures (WAT, sou, MCI, tot, CHE, swa, wba) has not been easy. Everton have the joint fifth best defence in terms of goals conceded and they have faced three of the four teams better than them in that regard. The Toffees face Liverpool and Manchester United at home and Arsenal away in the next three, then the fixture list opens up nicely through to the end of 2015 (SUN, whu, AVL, bou, CPL, nor, LEI, new, STO).

Lukaku 2014-15On the surface Lukaku looks a good option through that period, but will he frustrate his owners again?

The fixture list (right) shows the Belgian’s goal scoring games last season in blue and the evidence does offer some hope.

After a strong start, Lukaku had something of a slump through the winter months following a busy autumn of Europa League action, but he rediscovered his goal scoring touch toward the end of the year. The former Chelsea striker played regularly last season – he only missed Southampton at home and Swansea away last year, though he also played less than 45 minutes of football against Swansea at home and Newcastle away.

The Belgian failed to score more than one goal in a Premier League game last season, but has managed to do so twice this season already. 

In the final Premier League table from last season there was a six-point gap between Stoke City in ninth and tenth-placed Crystal Palace, who finished one point and one spot above Everton. All of the teams who Lukaku scored against last year finished below that gap in the table.

With no Europa League action and early indications that he could provide more generous returns this year, Lukaku could be worth another look when the fixtures turn in Gameweek 11 to include more teams of the type he had success against last year.

The strikers replaced by Martial

It seems slightly bonkers to be devoting a second article in two days to Anthony Martial given the 19-year-old has only played 116 minutes of Premier League football. However, after more than 150,000 more transfers in since my post yesterday, the Manchester United striker has recorded back to back price rises and is the biggest story so far of this Fantasy Premier League season.

The evidence base for assessing Martial is tiny, but FPL managers are leaping to judgment with their transfers so I feel justified in taking another look at the small amount of data available to us.

Yesterday, I looked at how Martial’s underlying data suggested his goal production was primed for regression. I also discussed how I am struggling to see the merits in swapping Bafétimbi Gomis for Martial, but today I want to take a broader look at the men he is replacing in FPL squads.

So far 11 strikers have been sold by more than 15,000 FPL managers this gameweek, with four – Christian Benteke, Gomis, Harry Kane and Romelu Lukaku – suffering at least triple that number of sales. As you can see from the chart below, they are all in a similar price bracket to Martial (now 8.2m).

Benteke TOR 146579 Gomis TOR 72847 Kane TOR 54290 Lukaku TOR 49460

I not sure comparing the four of them with Martial is very useful because the teenager’s limited game time means his data could easily be the result of a performance blip that hasn’t had time to smooth out. Nevertheless, I thought some FPL managers might be interested to see it anyway.

Given the other forwards have all played more than four times as many minutes as Martial, I thought the fairest way to compare them would be to look at the time in minutes between events rather than raw numbers.

Anthony Martial v Striker Alternatives

With a large sample size, each set of bars in a chart like this would be taller than the set to its left. Shots are the most frequently occurring metric we are looking at here, followed by Shots in the Box (SiB) and Shots on Target (SoT). Goals are the least commonly occurring of these four measures in football. Short bars show something is happening more regularly – for example, Gomis is hitting the target more frequently than Benteke.

The problem with Anthony Martial’s data and the reason I warned about regression yesterday is that his bars are almost flat. If we compare it with the other strikers we see that his minutes per Shot is roughly on a par with the other four strikers, as are his minutes per Shot in the Box. However, by the time we reach minutes per goal we see he is performing significantly better than the others.

Players will have patches where they perform better than average. It could be that Martial “is in form” and, if so, buying now could be make sense, but with just 116 minutes of data it is also possible that this is just a blip.

Players will also have periods where their performance is below average. Poor Harry Kane is an example of a player out of form. Last year he averaged a minutes per goal rate similar to that seen by Gomis now, but he has yet to score this season so we can’t give him a minutes per goal figure. As you can see, he is still shooting frequently and doing so in the box with a similar regularity to these other striking options, but he is failing to hit the target as often. Watching him play against Crystal Palace, Kane seemed to be snatching at some of his shots and that appears to be showing through in his minutes per Shot on Target rate.

Embed from Getty Images

Strikers are not all equal and some will perform better than others. Last season, 328 was the average minutes per goal among strikers who scored five or more goals. Even Benteke and Lukaku, who many FPL managers are selling, are scoring more frequently than that this season.

The best of the best last season was Sergio Agüero, who clocked up a goal just over every 100 minutes on average. In one particularly fruitful match last year he averaged a goal every 23 minutes.

Martial has scored a goal every 39 minutes on average this season. That can be done in one match, as Agüero showed, but even the best striker in the league couldn’t keep that rate of return up all season long.