Monday, October 20, 2014

Adopt a Team: Arsenal vs Hull review

Despite the suggestions in some section of old media, the so-called "stats" versus "scouting" debate is essentially over. I would imagine most, if not all, teams pay at least some attention to advanced stats with player performance tracked and assessed on an increasingly granular level. The same should be true in fantasy football. The suggestion that you can't scout players purely based on a spreadsheet is completely fair, but is likely something of a straw man argument as I don't know if anyone really believes that to be true. Equally, you are probably missing out if you think you can glean all the information you need by watching games (or more precisely highlights as the majority of games aren't televised in many regions). This game served a nice reminder that both those one sided views are foolish.

The highlights package would give you much of what you need to know in that Arsenal largely dominated the game but were picked off by a couple of Hull attacks and lacked the clinical edge to convert their own chances into goals. The data does a better job at highlighting that dominance:
  • Arsenal: 25 shots, 16 SiB, 9 SoT, 65% possession
  • Hull: 4 shots, 3 SiB, 4 SoT, 35% possession
However, it would also be dangerous to simply open up the box score, see that Hull scored twice on just four shots on target and conclude that this was simply an anomalous result and if you played it 100 times over, Arsenal would win (with a clean sheet), 99 times. Again, the spirit of that argument is true but seeing the game, you really see that the quality of Arsenal's chances was not good and a large portion of that possession was in areas that Hull effectively conceded.

The familiar cliché of Arsenal trying to "walk the ball into the goal" might seem applicable based on the data but from my viewing I might suggest they would have been better advised taking less shots, with several efforts - particularly in the first half after the equaliser - falling into the "hopeful" rather than "expectant" buckets. We'll get to player specifics in a minute but Cazorla and Chamberlain seemed to be the main offenders here.

It's generally dangerous to place too much stock in tired narratives about goals breeding confidence but in this case, the timing of Hull's equaliser did really seem key as the 10 minutes before and the immediate period after the Arsenal opener saw the home team camped in the final third with a second (and third) goal seeming a mere formality. Once they settled down, Hull did an okay job holding possession when they needed to, with Huddlestone and Diame controlling things quite well in the middle.

I won't go through every player in these wrap ups, but a couple of points stood out as being worthy of comment:

Danny Welbeck
The consensus among fans still seems to be that Welbeck simply isn't good enough but he equipped himself well here and impressed both in terms of effort and ability. Welback made dangerous runs all day and of all the Arsenal players it was he - the newest arrival - who seemed most in tune with his teammates, linking up nicely on several occasions. That ability to bring others into play suggests he should be able to contribute both goals and assists this year, something that cannot be said for some of his comparatively priced peers. He finished the day with five shots (all SiB) and a couple of created chances and was overall good value for his goal and single bonus point. Other good options exist in his price range of course but I'd still be more than happy sticking with the England man during Arsenal's upcoming run of games.

Alexis Sanchez 
This might have been the best Sanchez has played for his new team and he was the best player on the pitch for large portions of the game. Five shots (four in the box) and three created chances underline his attacking threat while his nine penalty box touches and eight dribbles highlight his overall involvement. His goal was great but his assist may have been even better and there's really not too much to say about his production or potential, both of which are predictably excellent. The question is whether his price tag is worth (a) the 0.4-0.7 premium over the two Chelsea options and (b) the 1.5m premium over Ramsey - who returned as a late sub here. With Fabregas' ownership up over 50% and a trickier fixture list on deck, the argument is certainly there to be made, but as always it depends on too many factors to try and sum up in a quick blurb, so we'll leave that for another day.

Santi Cazorla
I don't really know what Cazorla is anymore. Ostensibly a creator, he seems to spend long stretches of games with players running past him and thus not contributing a great deal at either end of the pitch. He ended the game with a lot of touches of the ball (138) and perhaps that's his designed role but then I don't know what that mean for Jack Whilshere who seemed to spend a good amount of this game stepping on Cazorla's toes and gesturing where he thought the Spaniard should be. For the first goal, Wilshere was caught upfield and seemed to expect Cazorla to cover but by the time he did he was stuck behind the powerful Diame and never recovered enough to influence the play. On the second goal Wilshere just simply didn't track Huddlestone - Cazorla was along way away so this must have been miscommunication with someone else - who ran free to chip the ball to Hernandez.

It's perhaps unfair to hold anyone to this standard, but, if you if look at Fabregas and Yaya Toure, it feels like they basically do the job that both WIlshere and Cazorla did here, which frees their teams up to have Oscar/Silva etc play with essentially no responsibility. I don't know if this has any huge impact on fantasy results but it does feel like this Arsenal side are sometimes lacking an extra body going forward and then at the back, often due to the middle men not seeming to commit to either role.

Mohamed Diame
Four goals in five games could well open the floodgates on Diame who has so far apparently gone unnoticed by the masses (2% ownership and a 0.1m decrease in value since his arrival at Hull). As an actual performance I thought he looked great, offering Hull a strong outlet to try and connect the midfield to Hernandez on the rare occasions they were able to hold onto the ball and, of course, he took his goal very nicely indeed. In fantasy terms though he's the classic regression candidate with nine total shots, four SiB and four SoT barely being enough production to support scoring once every eight games nevermind four times. Considering he seems to be fairly secure in the first team and at least offers something going forward he doesn't look like a terrible investment and wouldn't be a player to target selling but there's not enough here to warrant a transfer either unless you're solely look to capitalise on the likely price rise.

Tom Huddlestone
I've been an admirer of Huddlestone for some time but without wanting to sound like a broken record, while the assist was nice it came on literally his only successful pass in the final third and his average position on the pitch was deeper than Per Mertesacker's. His 13 shots for the season suggest he might get the odd goal but with just one from within the box, they're going to be few and far between (SoB are converted at a rate of about 4% across the league). Again, at 5.4m it's hard to be too critical but I'd suggest either going significantly cheaper or trying to sneak into the next price bracket as there just isn't much potential here other than consistent two-pointers.

Summary
This was a fairly entertaining game from a neutral perspective but with a fantasy football hat on, I'm not sure we learned a great deal more than we already knew. Arsenal are a good but not great side and without Ozil and Giroud the team really looks like it will pick itself for the next couple of months. That is perhaps unfair on Cazorla who isn't without quality but you have to think that Walcott, Ramsey and Sanchez will provide the ammunition for Welbeck more often that not with Wilshere pushed deeper and Chamberlain and Cazorla sliding to the bench (we might even see Chamberlain getting more time in the deeper middle role, where he has impressed in short spells). All four of those attacking options - along with the budget Chambers - look ownable and picking between the midfield trio needs a deeper dive than we're performing here.

As for Hull, this probably isn't the best game to judge them on as they were pinned deep for long stretches of the contest, but in their defense, they showed the moments of quality needed to score against a reasonably good backline and should enjoy success against weaker opponents in the future. Individually, while we see snippets of that quality, no one has the underlying data to really support anything more than a speculative pick, especially without too many promising fixtures in the next couple of weeks. Diame, Ben Arfa and Hernandez showed they have enough talent to offer hope though and so this remains a team worth keeping in mind when filling out the bottom end of your fantasy squad.

This week we're going to look at another of my favourite teams to watch (Everton) face off against newly promoted Burnley.

Friday, October 17, 2014

Adopt a Team: Arsenal stats

Even though we're still just seven weeks into the new season, we've already observed a number of interesting and somewhat strange happenings at the Emirates:
  • in terms of underlying data, the defense has been more impressive than the usually explosive attack
  • in terms of actual results though, the defense has struggled with just a single clean sheet to date
  • Wenger did bring in a relatively big name forward - who's looked great so far - yet there still seems to be a feeling that they need more - probably in part due to the success of Diego Costa on the other side of London
  • the team has quietly assembled the best depth they've enjoyed in recent memory yet they've suffered through so many injuries that the average bench has tended to be more Campbell and Coquelin than Giroud and Cazorla
  • Arsenal out shot Chelsea 10 to five and enjoyed more possesion at Stamford Bridge last time out, yet lost the game (their first of the season) and now find themselves nine points behind the league leaders after just seven games

The below chart attempts to quickly summarise the individual games played to date, using SiB +/- as a proxy for success (or failure).1

The good news is that, at least in terms of giving up shots inside the box, Arsenal's defense have performed very well this year with above average performances in every game except one (at Villa in GW5). Indeed, through seven gameweeks, their dSiB% rate of -33% leads the league, as does their suppression of chances created (-36% dCC%). As we saw in the lineup review, this success has come despite not being able to field their first choice back five once this season. With all but Debuchy now healthy (and Chambers filling in admirably for the Frenchman), it's not impossible that much of this defensive success continues and that would surely give rise to increased fantasy returns in the future. Perhaps their goals per shot on target rate will regress a touch - although at 38% it isn't vastly higher than the surprisingly high league rate of 36%. They will almost certainly benefit from a softer fixture list which sees them face Sunderland, Hull and Burnley in the next three gameweeks and no real marguee opponents until a GW17 trip to Liverpool.2

Player targets: Defense
As suggested in the lineup post, defensively it seems Chambers is the obvious play here, based on nothing more than his bargain price tag (he is unfortunately suspended for this week after picking up an incredible five yellow cards in six starts). Before we commit to that though, let's take a quick look at the attacking potential of the Arsenal players to date:

The chart shows absolute threat rather than that on a per game basis, so playing time needs to be kept in mind, but in short, no one here is really standing out and in fact, it's the bargain Chambers who's offered the best threat to date, thanks to his solid number of created chances.

Player targets: midfield and forwards
Despite professing an admiration for Arsenal and especially the work of Wenger, I haven't had the chance to see too many of their games this season so as I began this deeper dive into their season to date, I had some expectations that Ramsey would be a player I might need to highlight as not having the supporting stats to support his production. It's not that he isn't a good player, of course, but that centrally deployed role for a midfielder who doesn't take set pieces or penalties can sometimes lead to a lot of touches of the ball but ultimately not a great deal of production. That's not the case for Ramsey though, who has offered the best goal scoring threat in this team - by a distance - while giving as good as anyone in terms of generating assists. Sanchez and Ozil obviously get into those dangerous wide positions more often, racking up 28 and 27 crosses to date to Ramsey's nine, but considering the Welshman has matched them in terms of created chances suggests his potential may even be higher with his passes being more likely to be converted. With Ozil out for several months and Sanchez coming in at a 1.5m premium, Ramsey looks like a perfect candidate to buy back into when he's healthy, especially now that his injury has deflated his once sky high ownership.

Though he's further down the threat ratings to date, the reviews of Welbeck have been complimentary so far and at least per the highlights he seems to have hit the ground running. 13 shots, 10 SiB and five SoT is a reasonable return for four starts and a couple of brief cameos, making Welbeck and intriguing option at just 7.3m. He isn't likely to compete with the Aguero or Diego Costa class of forward, but in his price range he just needs beat out the likes of Pelle, Ulloa, Cisse and Sakho in order to be a relevant fantasy starter. Let's take a look at where he ranks compared to some of his peers to date:

This time we've gone with per game data to allow for better comparability and we see that Welbeck is fairly well placed among both his comparably priced peers and is at least competitive among the more expensive options (scroll the price slider to the right to see Diego Costa, Sergio Aguero et al). Starting this week with Hull, Arsenal have a nice run of games, making Welbeck an intriguing option, although a slight ankle injury makes him something of a risk for this week. With Giroud sidelined, he seems to offer an appealing combination of job security, excellent supporting cast and high upside, which for 7.3m is a package hard to ignore.

Summary
In summary, this Arsenal side hasn't been vintage to date and their shot production is only upper-mid table rather than elite. However, they are creating chances and have looked relatively sharp with Welbeck in the side and there's still reason to think there's more to come. Based on the stats to date, Chambers, Ramsey and Welbeck jump out as the best options at each position but there are plenty of other options to consider, none more so than Sanchez who has looked very good in spells so far this season. I look forward to matching some of the stats with the actual on field performance tomorrow.


1. As a quick reminder, SiB +/- quantifies the percentage of shots inside the box generated or surrendered by a team above or below the opponents' average. So if Chelsea normally concede six SiB and Arsenal manage eight against them, their aSiB +/- would be 33%, representing a 33% premium over the average. Defensively, the idea is the same but a minus number is the sign of a good team as they suppress shot totals compared to the average result.
2. I am half kidding about the exclusion of Man Utd in GW12 there. There's a perception that the team is terrible defensively but can score at will if they want but with an aSiB +/- of -23% on the road, I'll need to see more before I buy into that particular narrative

Tuesday, October 14, 2014

Adopt a Team: Hull Lineup

The second team we're going to adopt this week are Arsenal's opponents - Hull - who have been enjoying a reasonable start to the new campaign, although as we'll get to later in the week, their underlying data suggests trouble might be brewing if they don't start to generate more chances for their talented front pair.

Hull have switched between a 4-4-2 and 3-5-2 base formation thus far this year, making their lineup a little harder to forecast on a weekly basis. We don't really have enough of a sample of draw any concrete conclusions about the team scoring/conceding more/less goals when in either formation so I'm not going to focus too much attention there in this post. McGregor, Davies, Dawson, Robertson and Elmohamady all appear to be close to automatic choices here with Chester the first man in if the back three is favoured. Elmohamady is classified as a midfielder in the PL.com game so we'll touch on him in a second, leaving three options to consider. Davies and particularly Dawson come with the stronger reputations, while Bruce and Chester have the higher ownership numbers,1 yet it's Robertson who looks like the most interesting option here. We will get to individual data in the next post, suffice to say that Robertson leads his defensive teammates in created chances by a distance and offers as good a goal threat as any of them. Add in his lower price tag and he becomes the clear player to target here. Between the posts this is still very much McGregor's job, but with him sidelined until towards the end of Novemeber, Harper makes a reasonable rotation option for those looking to save a few pounds.

Whatever the formation, Steve Bruce had tended to deploy the Huddlestone and Livermore combination in the middle of the pitch, yet while that decision may be wise in terms of securing Hull's Premier League status, the pair do little to excite fantasy managers. In one sense, Huddlestone has shown flashes of excitement with 13 shots ranking up among some of the league's more promising midfield talents yet with just one of those coming from within the box, it's tough to forecast anything more than the two goals he chipped in with last time season.

The remaining positions - those which are most important to fantasy managers - are the trickiest to forecast and it's that uncertainty which may temper any enthusiasm surrounding this talented group. Whether in an advanced role or playing alongside Huddlestone and Livermore, Diame seems to have the inside track on one of the remaining midfield spots and he currently leads Hull midfielders in shots inside the box and penalty box touches. However, it's reasonable to assume that Ben Arfa has been brought to the KC Stadium to contribute in the first team sooner rather than later, which puts both these players into the risky category. To date Bruce seems to favour deploying both Jelavic and Hernandez, though it wouldn't be too surprising to see him roll with Ben Arfa, Diame or Ince behind a lone forward in away games, as was the case before Hernanez arrived. All this essentially means we have five players to fit into three spots.2 His pair goals in just four games suggests that Hernandez is the automatic choice here but his nine total shots (five SiB) are not spectacular and a local media review suggests there is still room for improvement in his overall game. He's a player who I will be watching closely at the weekend given his reasonable price tag and upside but I couldn't suggest investing in him just yet until we see how Ben Arfa is going to impact this team.

There is without doubt talent in this team yet with unremarkable team totals - despite playing a softer schedule to date - and some uncertainty around which of the attacking will play and how they will be deployed it's hard to immediately focus on any one player to monitor. Hernandez, Ben Arfa and Jelavic all carry solid price tags though and may yet offer some value in the future. At the back, it's the promising Robertson who stands out and he'll be another key focus for the weekend as he's tested against some of the league's finest. Next up is a focus on individual and team stats to date to potentially highlight some areas to capitalise on or avoid.
1. I was never a fan of essentially punting a spot on your bench just to save, at best, 0.5m so would not condone the 11% ownership of Bruce here. We all get stuck with dead weight from time to time as we concentrate transfers on other parts of our teams but Bruce is buried on this depth chart and will contribute essentially nothing for his price tag.
2. I'm assuming that Elmohamady will be preferred either as a wing back or as a more traditional winger in a four man midfield given that he offers the only natural width in this midfield pack.

Monday, October 13, 2014

Adopt a Team: Arsenal Lineup

The first team we're "adopting" this week is Arsenal, and so we'll start with the basics and look at who's played so far and who's likely to keep playing in the future. At the back there isn't too much to talk about with a fairly clear first choice back four of Gibbs, Mertesacker, Koscielny and Debuchy with Chambers likely providing the first bench option in both the middle and on the right with Monreal serving as Gibbs' understudy. Injuries have prevented that lineup from playing together all season, and that trend will continue with Debuchy (three months, ankle) sidelined for a sustained period. That will allow Chambers to continue his impressive start at the Emirates (after his GW8 suspension) and so at 4.6m he would be the clear short-medium term option to use here.

Despite the constant criticism of not spending enough money or being able to compete with the league's big spenders, Wenger has assembled a deep and talented midfield unit; one that is the core strength of this side (at least in terms of the attacking options). The easier portion to forecast is the piece that is of less interest from a fantasy perspective, with Flamini, Wilshere and Arteta essentially competing for two spots in the middle of the park.1 With Arteta currently sidelined (three weeks, calf), those two spots look settled, though their potential value is less clear and something we can touch on in the upcoming player stat review.

Despite the depth, the attacking trio was also shaping up to be relatively predictable with Ozil, Ramsey and Sanchez getting the lion's share of starts to date. Ozil (three months, knee) and Ramsey's (three weeks, hamstring) injuries throw that group into question with Oxlade-Chamberlain, Cazorla and Rosicky currently competing for time while Walcott and Gnabry will provide further options in a matter of weeks, or even days. One would expect Walcott to be given every opportunity to win a healthy share of playing time once he's fit, though unlike in previous years, he's unlikely to earn too many minutes in that lone forward role given the arrival of Welbeck. Ramsey looks close enough to a return that his replacement (probably Cazorla) shouldn't really be relied upon, but Ozil is out long enough to be ignored for now, and that should open the window for a returning Walcott to offer a differential option. That said, for 0.2m less, Ramsey is now held by a more manageable proportion of managers (18%) and there's no real reason to think his excellent start to the year wasn't legitimate and so he should remain a key option to consider.2

Up front, the intrigue about how Welbeck and Giroud can fit together is a moot point with the Frenchman sidelined until the new year with an ankle injury. Welbeck looked very promising in his first three games (the team understandably struggled at Chelsea) and will be well placed to excel if he makes that role his own. Perhaps Walcott, Podolski or Campbell vulture minutes here and there but from what we've seen and based on the transfer fee paid, Wenger clearly thinks Welbeck can lead the line for his team on a weekly basis and thus he becomes extremely intriguing for just 7.3m.

Despite Wenger assembling one of the best squads during his tenure at Arsenal, and indeed one of the best around the league, there remains plenty of reliable options here, aided of course by key injuries to Ozil, Giroud and Debuchy. In the next post we'll dig a bit deeper into exactly how players have contributed so far, and try and identify a couple of targets to focus on for the coming weeks.


1. Ramsey can of course play this role too, as he did in GW5, but it looks like his preferred role is behind the front man and he should generally be free to push forward rather than provide defensive support
2. We'll look at a comparison between him and Sanchez in the player stat post to see which player should be targetted going forward for the coming weeks and months

Thursday, October 9, 2014

Adopt a team

With a busy schedule of late, I didn't see much action last weekend and then with the international break in play I've been struggling for inspiration. Like all resourceful bloggers I have stolen settled on an idea for the next few weeks of action. Taking a page from my favourite NFL podcast (Grantland), I am going to "adopt" two teams each week and focus coverage on them in the lead up to their games (ideally against each other which allows a more comprehensive review after the game). The idea being that trying to cover 20 teams on your own is (a) difficult and (b) leads to average coverage of all teams rather than great coverage of some. I would suggest this idea works best in all media, though of course a lot of local news is terrible, but that tends to be because it panders to ridiculous local interests and biases rather than because it's a flawed principle.

The schedule for each team will run roughly from Tuesday - Tuesday and try to feature (as much as possible), a review of the season to date, player highlights, future forecasts and then a review of their game with a light touch on their tactics.1

At least for the first few weeks, I have tried to pick two teams who are facing each other, with the below being the initial schedule.2 I will either give this week's teams a bit more attention or look at a more general topic in the coming days if anything jumps out as a I catch up on the data:

Gameweek 8 - Arsenal and Hull
Gameweek 9 - Burnley and Everton
Gameweek 10 - Newcastle and Liverpool
Gameweek 11 - QPR and Man City
Gameweek 12 - Chelsea and West Brom
Gameweek 13/14 - Man Utd and Stoke




1. I am far from a tactical expert, and so while that doesn't seem to stop many throwing around their "trequartistas", "registas" and other sexy, international names for fairly mundane concepts, I will try and keep things in my wheel house

2. The teams have been selected on which games I have televised here in Canada