Gameweek forecasts: a couple of case studies

Producing forecasts is a tricky business. Even with hindsight it is tough to predict the expected outcome of a given game (i.e. how shots transform into goals) and that problem increases exponentially when you also need to try and forecast the underlying data. Throw in uncertainty around how much players will play and the issue of small sample sizes and you have a recipe for some funky results over these early weeks of the season. First, to make sure the model isn't totally off track, let's look at how it performs retrospectively over the first seven weeks of the season (using actual shot totals as inputs):

Learn About Tableau

Though we can see outliers in the above chart (especially at the top end of the market), the overall trend is promising and the r-squared of 57% for players with a risk factor of 2.5 or less is encouraging enough. That's not to say it's infallible, but it's a good start and for the majority of the extreme outliers we can point to specific factors which have led to their strong results (Yaya Toure is unlikely, for example, to convert his shots on target into goals at a 66% clip for the rest of the season).

So if we're happy that the model is working relatively well over a medium length period, let's get into the forecast side (which include predicting how many shots a player will get and how they will be converted) and look at a couple of actual examples from this week's forecast:

Loic Remy 7.3 points (you'll need to slide the risk slider to 2.7 or more to see him included)
Remy represents one of the dangers of forecasting in the early weeks; a problem that is compounded by the fact that Remy has only started four times. When he has played he's been nothing short of spectacular, averaging 4.4 shots per 90 minutes, hitting the target 50% of the time (a very useful rate). Despite this success, I imagine most people who asked about this forecast were confused given Newcastle's opponents. With just four goals conceded on the season, Liverpool have appeared to be a very useful defensive side and thus don't immediately jump out as a team you want to start your forwards against. However, digging a bit deeper we see a team with a +/- SiB rate of 22% away from home, surrendering the same number of SiB as teams like Villa and Cardiff (this despite playing AVL, SWA and SUN). This isn't to say, of course, that Newcastle are set to put Liverpool to the sword this week, but looking purely at the data, the Magpies' prospects this week are better than most would probably think (the model estimates them to notch around 9 SiB this week). With Remy accounting for 40%+ of his team's SiB, the model likes his chances this week, even if the intangible factors (playing time risk, wide role) suggest more caution.

Daniel Sturridge 4.6 vs Luis Suarez 4.1
If Remy is giving us trouble with five appearances, things get even tougher with Suarez who has played just twice. I suppose we could lean on prior year data but without wanting to reinvent the whole model for every idiosyncrasy, we'll just live with it. Long term I would personally back Suarez to top his English colleague in the scoring charts, though as this is a stat based site, I've nothing to base that on other than past events so it doesn't make it into the model. As always, these forecasts should support our decisions and if something seems off then we should simply ignore it.

Stevan Jovetic 7.3 (risk rating 3.9)
This is something of a "damned if you do, damned it you don't" situation as if I exclude players from the weekly listings I inevitably get questions where Player X is, and then when they're in, it seems ridiculous to rank the scarcely-used Jovetic as the top option. For clarity, all players make the listings and you can filter down to more reliable options using the risk slider. Jovetic's sample size is almost certainly too small to be particularly reliable and thus the 7.3 points ranking isn't wholly useful, but if you think this then simply ignore him. I prefer this approach rather than me deciding who is/isn't relevant and having readers miss out on a sleeper prospect they are targeting. This is the same approach used by publications like Baseball Prospectus who generate PECOTA forecasts for minor league players on the assumption they get major league playing time (and it's to the manager, player and luck whether the youngster gets his shot). For what it's worth, in his limited time Jovetic has been excellent (6 shots, 4 SiB and 2 SoT in just 85 minutes) and thus the model likes him to succeed were he to be given time in this talented City side.

Julian Speroni 5.7
This is probably the strangest forecast for the week and I must admit, I went back to recheck the data myself before posting these numbers. The fact that it's Speroni (and hence Crystal Palace) is somewhat surprising though when you consider their opponents' pathetic efforts away from home to date, it doesn't seem overly unusual to see Palace ranked well this week (indeed they would rank 3rd in the standard "goals per game" projections from last season). The issue then, is more that a goalkeeper is the week's highest ranked player with a low risk rating which doesn't feel quite right. Part of this is simply the perception of fantasy players and part of it is a lack of sophistication with the model.

On the first point, Artur Boruc ranks 9th among all players this year and three 'keepers place in the top-15 so in reality it isn't particularly surprising to suggest a 'keeper will score well (remember that they tend to earn more points than defenders per clean sheet due to the presence of saves). They lack the upside of defenders who can also score or notch assists, but for most players the chance of those events is fairly low and thus doesn't have a huge factor on a weekly ranking. The second point is a bigger issue and that's how 'keeper points are calculated. Right now, saves are awarded on a very crude average basis which doesn't take into account the propensity for earning saves in a given game. Thus, while Speroni has racked up decent save totals to date, if his chances of a clean sheet are higher this week then his save totals should go down, but the model doesn't make such an adjustment. This is unlikely to result in more than a half point variance in a given week though.

The wider point on 'keepers and to a lesser degree defenders is that the forecast is more likely to be wrong on them in a given week as their chances of success is a little bit "all or nothing" (where as midfielders can earn points for clean sheets, assists, goals and are more likely to get the bonus nod). I still feel fairly confident is using this data for ranking purposes but wouldn't suggest captaining Speroni (5.7) over someone like Michu (4.9) given the exponentially higher upside enjoyed by the Swansea man. To repeat, this data is based on logic but a simple model cannot account for every possibility so personal judgement is still required.

Hopefully these examples gave a bit more detail to how the weekly rankings are made and we'll continue to check back to see how they are performing as the season goes on (we'll hopefully see some of the stranger outliers disappear as sample sizes start to increase). Thanks for reading and for sticking with the blog during the quiet opening weeks and please continue to send your questions to @plfantasy, on Facebook or in the comments below.

Gameweek 8 forecast

Learn About Tableau

Risk slider - this season each player is assigned a risk rating, based on their playing time and current injury status (as explained here). There's no reason to definitely exclude all risky players, but you should consider their playing time before pulling the trigger with them.

I've had a couple of questions about a couple of the 'odd' forecasts above, namely Loic Remy's lofty status and how Sturridge out ranks Suarez. I'll address those in more detail tomorrow but wanted to get the data out now.

Player Dashboards explained

Hopefully you've had a chance to look at the new player dashboard which has launched this week and can be found here or by following the link on the menu bar above. I'm sure that for the most part the data is self explanatory but I thought it might be useful to quickly run through the new features so you can get the most out of them. Let's start with the points section:

1. Here we simply see the players' actual points by week plotted against their expected total. One key to note here is that the expected number is based on their actual shot data rather than the forecast number that will be given each gameweek starting with this one. Point being, the expected number shows how many points we'd expect a given player to score given all the other events observed from his performance.

2. This is a somewhat crude depiction of how each players' points total was earned. You can hover over each slice of pie for an explanation, namely:

• Appearance (less yellow and red cards): blue
• Goals: green
• Assists: orange
• Defense (clean sheets less points lost for conceding 2+ goals): red
• Bonus points: Yellow

3. The +/- score quickly shows whether a player is under or over-performing his expected points total. A positive number suggests he has under performed his total and thus should be due for some positive regression should he continue to produce shots / create chances etc at a consistent rate. Note that we're not saying that this gap will necessarily closed, only that we'd expect his future totals to match more closely his expected numbers.

4. As with the points chart above in (1), this chart plots actual goals scored against expected goals, again based on the actual shots registered by the player in a given game. Comparing these on a one-game, weekly basis is probably not a great idea, but over a longer period we can identify players who are perhaps getting a bit unlucky and under-performing their underlying stats and thus might be undervalued by the market.

5. These simple pie charts show the split between:

• pSiB% - the percentage of a team's shots inside the box that the given player has accounted for (adjusted for the time actually spent on the field)
• pSoB% - the percentage of a team's shots outside the box that the given player has accounted for (adjusted for the time actually spent on the field)
• SoT% - the percentage of the player's shots that hit the target. Most research I've performed suggests this is a sustainable skill and won't regress to a league average rate, though we might expect it to regress to a player's own historic rate (a player can't for example, hit the target with 80% of his shots over a sustained period).
• SiB% - the percentage of a player's shots that were taken inside the box. 

6. A player's goals per shots on target rate is somewhat complex but in the majority of my research I've found that for the most part it tends to regress towards something of a mean. Some players - though not particularly the first ones you'd think of - have show an ability to exceed the league average with some consistency, though I haven't done a full enough test to determine if these are simply expected statistical outliers. For now then, we have two different graphs to show. For lower profile players we can see their G/SoT rate for this season against the league average rate for their position. For the more established players, or those who bring a strong pedigree of success from other elite leagues, we've highlighted their historic rate as the comparison. Where available a bias is given to (i) time played on the current team, then (ii) time played in the Premier League, but we've used other league data for players like Ozil or Soldado on the assumption that these elite few need some recognition as being better than simply average.

7. These charts highlight:

• pCC% - the share of his team's created chances for the given player
• Final third passes - the percentage of a player's passes made in the final third of the pitch

8. The assist per created chance rate works similarly to the G/SoT rate described in (6) above. The key difference is the average rate comparison which for this metric is the rate for the player's team as a whole. For example, the above shows Aguero enjoying a 38% rate while City as a whole have seen their created chances converted at just an 18% clip. Allowing for some variance given the quality (and position) of pass made by Aguero versus some of his teammates, we might suggest that he's due for some regression in this area as the season progresses.

Clean Sheet Conversion Rates

Alright, enough is enough. While I'm concerned about delving into data too soon and reaching all sorts of ridiculous small-sample-driven conclusions, I'm equally conscious that people want to start making big decisions with their respective teams and thus it's time to launch the weekly rankings and forecasts (still with that small sample asterisk though).

Before that, let's look at a new addition to the weekly forecasts. With most of the forecasts we do on this site, there are two distinct parts to the puzzle:

1. What is the expected volume of an underlying event (normally we focus on shots, shots on target etc)
2. What is the impact of those events on actual footballing events (i.e. goals, assists and clean sheets).

With goals we've spent a reasonable amount of time talking about how we forecast shots and how we convert those expected totals into goals, but I've tended to neglect the defensive side of the game. This had led to the unfortunate position where team totals are given each week but you end up having a tough time comparing, say, your 4th defender and your 4th midfielder, when deciding whether to go 4-3-3 or 3-4-3. With that in mind, this post will attempt to lay out some analysis of how different defensive shot totals are converted to clean sheets, and provide an example of how future forecasts will be made for the defensive side of the pitch.

Converting shots into goals
This is the reverse analysis of the time we've spent looking at player shots becoming goals, though we're going to be looking at it slightly differently as we're dealing with game totals for a team rather than an individual player. The below data is taken from the 2012-13 season only (plus a little bit of the new season), which isn't an ideal sample size, but is all I have to work with for now. If anyone has any differing conclusions stemming from prior years, please share them in the comments below or via email.

The below tables show the percentage of games which became clean sheets after a team conceded the noted number of shots inside the box (SiB) and shots on target (SoT).

The first point to note of course is that (thankfully) the data makes sense - especially in the larger 2012-13 sample - and a team's chance of keeping a clean sheet falls as they concede more shots. Further on that point, you can see that, generally, the first few chances you concede greatly diminish your chance at a clean sheet, where as once you get into the higher ranges, you don't see such variance as each shot is registered. As an example then, if a team is forecast to concede five SiB and two SoT, the data suggests they have a 43% - 50% (35% - 50% based on prior year) chance at a clean sheet, depending on which metric you look at.

Stop the presses! Less shots conceded equals a better chance at a clean sheet! Admittedly this isn't new news, but by putting some actual percentages to the problem, we can hopefully get to a point where we're getting a usable number each week. Right now, the forecast shot data is converted into goals and we'll say a team is forecast to concede 1.2 goals, but as we know, for fantasy purposes, defensive outcomes are very much all or nothing (plus some small impact of playing someone who concedes a lot of goals but you rarely be in that circumstance). A forecast will never suggest a team will score zero goals so you end up with weekly rankings which are useful when determining whether to play goalkeeper A or B but not so good when assessing whether to play defender A or midfielder B or whether or not to sign a defender for the next five weeks.

Team specific data
As we've seen with converting shots into goals, different teams achieve this at different rates, and based on yearly data, there's reason to believe that the difference is sustainable (better players create better chances which lead to better shots and thus a higher chance of conversion). A similar hypothesis can also be put forward then for goals conceded (better teams limit opponents to either shots outside the box, from tight angles or contested efforts and thus will see them converted at a lower rate). Thus, we should look at how different teams saw their surrendered chances converted into goals, before assuming we can use a standard league-wide average. This analysis is prone to small sample size issues, given that a team might only register a particular event once all season. For example, in the last game of last season, Fulham surrendered 16 SiB yet were still able to keep a clean sheet. Given that this was the only time they conceded such a total all year, the data would suggest they have a 100% conversion rate to apply to future games where they concede such a huge haul of shots. This is clearly a perverse situation but is combated by:

1. Grouping shot totals together into ranges, such 0-4, 5-7 and 8+ SiB conceded rather than looking at each total individually
2. Regressing totals back to league average based on the number of occurrences in the population. If, for example, Arsenal have 10 games where they've conceded 4 SiB and posted a clean sheet conversion rate of 50% in those games, we have a lot more confidence in that total than the fact that West Ham are 1 for 2 when conceding 8+ shots. We will therefore use a weighted average between the team rate and the league rate to get our overall expected conversion rate.

For reference, the conversion totals (regressed) for each team in each SiB and SoT range for the current year to date are as below:

After just seven weeks, it's to be expected that the team-by-team rates are all very close, as we simply don't have enough data for anyone to really distinguish themselves from average. These rates will likely diverge a bit as the season progresses and we can check back in a few weeks to identify any trends or issues here.

A worked example
Let's look at this week's contest at the Etihad to put the above into context:

1. As in past projections, let's try and get a forecast for the number of SiB and SoT for the coming week. Going into the week City had registered a -32% +/- score (holding opponents to 32% less SiB than they have averaged elsewhere), so with Everton averaging 8 SiB on the road, we get an expected SiB total of 5.4. On the other side, Everton had posted a +/- of 4% on their travels and City had conceded just 4 SiB at home. This gives an expected SiB total of 4.5. Taking an average of these two totals gives an expectation of very close to 5. 
2. Using the data from above we see that an expected SiB total of 5-7 is converted into a clean sheet 38% of the time (specific to Man City).
3. Moving on to SoT, we see that City had posted a very impressive -50% SoT +/- mark at home and so with Everton adding a healthy 4 SoT away from Goodison, we'd give them an expectation of 2 SoT for this game. On the reverse, City have surrendered just 2 SoT per game and Everton's +/- rate of 5% doesn't have much impact here. Again, taking an average of the two marks we get an expectation of 2 SoT for the game.
4. Using the aforementioned tables again, we see that 2 SoT converts to a clean sheet at a 44% clip.
5. Long term we can look at which of these two forecasts correlates better to the actual data observed, but for now we will simply take a crude average and get a mark of 41%.
6. For individual players, the final step is simply to say that a 41% chance of a clean sheet is worth 3.6 points (2 appearance points plus 1.6 for the clean sheet) and that is the number which will then be added to their offensive threat to be included in the eight week forecast / captain rankings. 

So that's the new defensive forecasting system in a nutshell and I'd appreciate any comments / feedback in the comments below or via email / Twitter / Facebook. Regularly scheduled programming will start to resume this week. Thanks for sticking around!

Lineup Lessons: Gameweeks 1 -4

As we all know, getting too bogged down in data analysis at this stage of this season can be dangerous, as the sample sizes on hand lead to one anomaly having a material impact on our conclusions (whether or not that changes after just 8-10 weeks is an argument for another day, but we can agree that three of four games isn't enough to make too many concrete claims).

One factor, however, that is equally essentially to fantasy success that can be discussed now is playing time. Regular playing time at this stage of the season doesn't guarantee a future spot in the side, but it is indicative that a given player sits atop the depth chart at his position and is at least in a position of privilege for the coming weeks.

One of the charts I included in the recently released e-book showed the defenders for every team for each week of the season, assigning them a status based on whether they were rotated by choice or unavailable for selection. The idea being, that we can deal with injuries and suspensions but seemingly random rotations can ruin a player's fantasy value and thus he shouldn't be considered for ownership, even if he is a 'cheap link' to that defense.

For this season, I've extended this to all players and will track these movements as the season progresses. I will leverage from these charts when performing the 'lineup lesson' analysis that we discussed last season and it will hopefully make it easier to spot trends and avoid potential mistakes.

A couple of quick notes:

• For simplicity, if a player is on the bench then I will classify him as 'rotated', even if there is some question over his fitness.
• For player positions, it can sometimes be hard to pinpoint exactly whether a player is playing as an attacking midfielder (AM) / forward (FW) or a central midfielder (CM) / attacking midfielder, but the idea behind these classifications is to show whether someone like Michu is playing as a lone front man or just off a teammate, or whether Jack Wilshere is playing in a holding role or a more advanced position in the front four, so the definition doesn't have to be too precise.
• Sometimes it is hard to identify whether a player has been rested or is injured, with an injury only revealing itself a week or so later. I will endeavour to get the charts as accurate as possible but if you ever spot an issue, please let me know and I can get that updated.

Here then, are the lineups for the first four weeks of the season:

Learn About Tableau

Arsenal
For the most part, Arsene Wenger seems to be quite settled on the defensive side of the pitch with Szczesny, Gibbs, Mertesacker and Koscielny all playing every game in which they were available. Jenkinson got the nod over Sagna in GW3 and this week the Frenchman deputised at centre back, so we're not 100% clear as to who owns that right back spot at the moment. Unless we see solid evidence that Jenkinson has made the right back spot his own (and thus become an intriguing cheap link at just 4.5m), Mertesacker or Gibbs look like the best choices here.

Going forward, Arsenal enjoy the best strength and depth they've had in recent memory with nine capable players (plus the injured Arteta and Diaby) competing for just five spots. The addition of Ozil will likely limit the number of minutes Wilshere gets in a more advanced role, assuming Cazorla returns to fitness sooner rather than later. With so much competition paired with a couple of heavy hitters (who are likely to start whenever healthy) in Ozil and Walcott, the rest of this midfield looks tough to own, even with their tempting price tags. Perhaps Ramsey - who has done an excellent job at getting forward from deeper positions - is the exception and remains ownable at under 6.0m.

Aston Villa
A very settled team here, with only a couple of optional rotations all season. Whether or not there's any value outside of the excellent Benteke is questionable, but at least we know which players will collect their two points every week!

Cardiff
Not too much excitement here either, though the arrival of Theophile seems to have ended the run in the side for Connolly. Regardless, if you did want to access this defense, Turner at 4.0m looks just as safe yet cheaper. Bellamy missed out this week due to a lack of overall fitness and his ongoing country-vs-club debate should be a red flag for the 4%+ managers who currently own the Welshman.

Chelsea
With this week's round of rotations, only Hazard and Ramires are left among all midfielders and forwards as being ever present for Mourinho (in just three games, remember). The front line looks particularly messy with four legitimate options vying for just a single spot. Given the prices involved that entire group might well prove to be unownable and could well be joined by several of the midfield group if Hazard suffers the same fate as his colleagues in the coming weeks. How to deal with such a situation is tricky and you need to assess if 80% of Hazard is better than 95% of someone like Michu who will start virtually every game when available.

At least things are looking settled along the back line with Cahill and Luiz the only rotation issue to date. Now that Luiz is staying at Stamford Bridge and has regained fitness, we can expect him to feature more often than not and if you have any concerns about Terry's ability to play every week, Luiz could offer a nice option given his discounted price tag compared to the safer Cole.

Crystal Palace
Aside from new signing Mariappa sliding in at right back, Palace have been unchanged for the last three games and look to be one of the more predictable sides in the league. Whether or not we can find anyone to squeeze any fantasy value out of is a question better suited to the upcoming 'fanning the flames' piece.

Everton
Martinez has been incredibly consistent in his team selection, with the departure of Fellaini being the only change of note. He has been replaced in the side by Gareth Barry, which should allow the impressive Ross Barkley to move into Fellaini's old attacking role: a factor which hasn't gone unnoticed by the 24% of managers who have already nabbed the Everton midfielder. In a strange quirk, he actually failed to register a shot this week in that advanced role, after averaging over four in the first three games, though the quality of opponent obviously played a role there too.

With three goals in four games, this Everton side haven't grabbed many headlines to start the year, yet they've managed to register 66 shots (3rd in the league) including 40 inside the box (t1st) and 22 on target (4th). Part of that will be driven by their relatively weak schedule to open the season, but this is shaping up to be a useful team again and given the consistency of players on offer, should offer some fantasy value as the season progresses.

Fulham
The story of this lineup so far has merely been a continuation of last season: stability at the front and back with a total mess filling the midfield. Well, in truth, the middle of the midfield is set with Parker and Sidwell holding down the fort, but while that pair have some value as minimum price plugs, it's the wide positions where we might find some value in this side yet Duff, Kacaniklic, Ruiz, Kasami and Taarabt have all been rotated at will. Given the fact that three players (van Persie, Giroud and Lambert) have taken as many shots inside the box as this entire team, this mess is one to be avoided for the foreseeable future.

Hull
With the new signings settled into this team, there's very little to discuss here with the only question mark being what happens when Sagbo returns (in all likelihood not very much for fantasy managers).

Liverpool
This is an interesting side whose lineup became increasingly complex as the transfer window went on. Though the team has depth at all positions, the 'keeper and full back positions look settled and that's probably where our focus should lie for now when trying to capitalise on this excellent defensive unit. Sakho looked somewhat erratic on debut Monday night, though his physical skills and comfort on the ball were clear to see (perhaps to a fault) and given the large outlay to bring him to Anfield, it's likely that over time he works himself in the starting lineup here ahead of both Skrtel and Toure. For a 0.5m saving though, that risk is currently not worth taking with Jose Enrique and (when fit) Johnson representing the best value here if you want to double down with Mignolet.

The midfield has been fairly predictable to date but with Moses making a scoring debut and Suarez nearing a return from suspension, this situation might start to get messy. Coutinho has quietly emerged as a force within this team and Rodgers noted after the game that his departure really impacted his team's ability going forward. That served as a clue as to how highly Coutinho is valued and he should be able to start most games when healthy. The other spot could be mired in rotation all year so expectations for old favourite Moses should be tempered.

Edit: Coutinho is now said to be out for six weeks which will open the way for someone else to enjoy a brief spell in the side.

Man City
Manuel Pellegrini has already given 19 different players a start this season, with Micah Richards coming close to fitness to become the 20th (and likely end the run of one of only two ever presents in this side - Zabaleta). Pellegrini's approach to rotation was rather strange, with the manager naming an unchanged front six for the opening three games of the year then dropping all but Yaya Toure this week (Silva was injured) for the trip to Stoke. Normally we'd expect a team to cycle through their options, dropping one star every other week or so to keep everyone fresh, and given the disappointing display this week from City (two SiB and three SoT), we might see that next time Pellegrini wants to shuffle his deck.

This rotation probably doesn't a huge impact on City player's value, as we all knew this was coming and it's largely built into their prices already (if we thought Aguero would play 38 games his price would be much closer to van Persie's 14.0m than his current 11.1m). Though his playing time is certainly a plus, I'd encourage Yaya Toure owners to look at the Ivorian's two SoT's and four touches inside the opponents box and then take a long hard look at players like Walcott or Michu who can be had for a similar price.

Man Utd
David Moyes has run out a very settled lineup during his brief time at Old Trafford with most changes being forced on him rather than his desire to rotate. He always favoured a stable lineup at Everton and that seems to have carried over to his new team, a fact that should help reasonably priced options like Welbeck and Valencia become more reliable than in past years.

Newcastle
We're seeing some movement in this team but at the key spots there's enough stability to offer fantasy value. At the back, it isn't clear whether Steven Taylor will be sidelined for good or if this week's benching was a one off, so for now the well priced and versatile Yanga-Mbiwa looks like a cheap link here. Assuming Cabaye is back, this midfield has a good amount of depth which will unfortunately limits everyone's potential aside from the impressive Ben Arfa.

Norwich
Chris Hughton tinkered with his midfield options during the first couple of games of the season but with Snodgrass now fit and Elmander brought in as another forward option, he seems to have settled on a lineup, naming an unchanged side in GW3 and GW4. Defensively the team looks set now that Bassong is fit, though at 4.2m Whittaker looks well placed to offer good value as a cheap link there. Attacking wise, it's been the youngster Redmond who's impressed the most with 10 shots already registered (though it should be noted that just one of these came inside the box).

Southampton
It's not really clear what Southampton are doing at the full back positions, which is significant because Clyne (4.4m), Fox (4.4m) and particularly Chambers (4.0m) come at a discount to the central options in this team. One would expect Shaw (4.7m) and Clyne to eventually take those spots but in the mean time, Boruc (4.5m) and Fonte (4.6m) offer the best combination of reliability and value here.

The midfield appears to be fairly stable, though the arrival of record signing Osvaldo puts Rodriguez's playing time in serious jeopardy. It looks likely that he will split time with young Ward-Prowse to play on the opposite flank from the likely entrenched Lallana (with Lambert, of course, locked in up front). Osvaldo has shown some promising flashes in his limited minutes so far, though it seems almost unconscionable that he comes in with a higher price tag than Lambert, who's league leading seven SoT speak to a player who could (and perhaps should) have a better points haul than he already does.

Stoke
As always Tony Pulis Mark Hughes has rolled out his side with extreme predictability, with the rotation of Crouch and Jones the only real move of note. There's really no reason to get involved at all there, and the best bets in this side could be a couple of minimum priced options in Marc Wilson and Jermaine Pennant. With Glenn Whelan lurking on the bench, it isn't 100% guaranteed that Wilson will be able to hold down that defensive midfield spot, but for now he makes a nice addition as a 5th defender type.

Sunderland
This team is a total disaster right now, and while there is certainly scope for something to rise from the ashes, I wouldn't want to try and pinpoint any one player to back right now. Colback and Johnson are the only ever-present outfield players through four weeks, and considering Colback is a midfielder being deployed as a defender, there's not much to dwell on there (Johnson is a reasonable prospect but feels more like a 5.5m player rather than one costing 6.9m). We'd expect Fletcher and Altidore to get the lion's share of starts up front, but Borini is not without talent and could still upset things, so given the lack of chances going around for this team already (just 11 SoT in four games) it seems prudent to steer clear and see if Di Canio can right this sinking ship.

Swansea
Given the knocks to several midfielders already this season, we still don't have a clear answer as to which of the talented group will be benched with any regularity, though De Guzman can probably be the first name thrown into that dubious category. That would leave Dyer, Hernandez and Routledge fighting for two spots and with returns somewhat limited so far, one is minded to stay away until someone emerges from that group as a legit fantasy threat. At least this defense is nice and settled with four straight games with an unchanged back five.

Tottenham
Let's start with the easy bit. The defense has been unchanged for all four games with Kaboul relegated to backup duty behind Vertonghen and Dawson (who is excellent value at just 5.1m and likely due for another price rise when people realise it). The centre of the park sees Dembele, Paulinho and Capoue scrapping for minutes with the former two looking favourites to play more often than not. Paulinho has seen time in a more advanced role too, but with Eriksen now joining the team and Sigurdsson enjoying a very good game this week, a repeat of that deployment could be limited and thus some restraint needs to be exercised before getting carried away with his promising statistical start. No one has really distanced themselves from the pack here (other than Paulinho) with Chadli probably edging it thanks to his solid assist threat (eight created chances) and affordable price tag (7.5m). Aside from Lamela (9.0m), there are some very promising price tags on offer here and at 7.0m and 5.5m, there's even scope for the likes of Paulinho and Dembele to succeed even if they are held back in deeper roles. Despite the presence of Defoe, it looks as though Soldado will be the clear number one forward, a fact which should set him up nicely to deliver strong returns on his reasonable (though by no means cheap) price tag.

West Brom
With the recent signings only just coming into contention, it's tough to forecast how this team is going to shake out. With just one goal to their name through four games, it's clear that something needs to change and thus we probably can't write any of the incumbent attacking players into Clarke's future team sheets in anything more than pencil (the defense looks settled but haven't performed overly well so offer little right now). There is talent here and it's a team worth watching but there are simply too many question marks to get involved with old favourites like Sessegnon, Sinclair or Anelka just yet.

West Ham
Sam Allardyce has opted for a consistent lineup again so far this season with only a couple of voluntary changes made to date. The biggest long term question is who will miss out when Downing, Jarvis, Cole and Nolan are all healthy, though the most useful of those options - Nolan - also looks like the most secure so there's nothing to lose sleep over here. 

Shot data: a terrifyingly early look at team totals

I simultaneously love and hate writing posts at this time of year. On the plus side, there is so much happening with new transfer signings bedding into their respective teams, the promoted sides showing what they're made of and familiar faces emerging or fading as fantasy stars. With all that excitement though, comes the nagging issue of sample size, which needs to be trotted out with boring regularity during these early stages of the season. I don't think anyone reading these pages would be naive enough to assume that Liverpool have emerged as the best defensive unit of all time, yet to totally ignore their impressive start seems a bit too conservative in the other direction and might lead us to miss out on emerging trends.

With this in mind, I'm going to have a very tentative peak at the shots created/conceded by each team compared with their 2012/13 totals and see if we can we start to spot any flags which need to be watched more closely in the coming weeks. Remember, any variance here is absolutely not a reason to conclude anything, just a mere flag to look deeper as to what might be causing that difference in the coming weeks.

Attacking data

Learn About Tableau

One point to note, which will have been picked up by anyone who's even casually glanced at the weekly scores is that goals are down so far this year and this is backed up by lower shot/created chance numbers. The variance isn't huge (it amounts to about one less shot inside the box per team each week) but it could be an early indication of a more defensive league, in part due to the arrival of a number of generally conservative managers at the league's best teams (no matter the opponents, it was rare for United to get shutout in back-to-back games under Ferguson's leadership. Indeed, United only threw out two zeros in all of 2012-13 in the league).

Liverpool have the greatest variance to date, averaging just 12 shots through three games compared to 19 for last season, though that is largely driven by their woeful display against Villa (five total shots) and one game should not cause too much alarm just yet, especially with their talisman Suarez due back in a couple of weeks.

If we look at the volume of created chances we see that all but three teams come within a variance of two chances from last season - a sign to suggest that using prior year data is probably a reasonable approximation during these early stages of the season.

Defensive data

Learn About Tableau

As noted above, defensive data is generally improved across the board so it's no surprise to see a couple of strong performers here. Sunderland's improvement is as much to do with their terrible record last season (17 shots conceded) as their strong play this year (11), though in fairness, continued play like this would push them into ownable territory; a marked improvement from last campaign. Everton's data looks outstanding, though the strength of their opponents needs to be considered. Still, beating up on weak teams is a definite asset for a fantasy defense and with three strong performences from three attempts, this is probably the most significant finding to date.

As we said, there's nothing here to overreact to, with the main conclusion being that prior year data is probably a useful benchmark and thus will continue to dominate the mechanics behind the weekly rankings for the coming weeks.

Coming up

I'm back on solid ground and back to reality so the blog will be getting back into full swing very soon. Over the last three weeks I haven't had much of an internet connection and didn't have a laptop, so it's going to take me a couple of days to get the data recorded and catch up with the latest news, but I hope to have a limited data set ready for GW4 and everything up and running by GW5. In reality, this is fine as two or three weeks of current year data isn't going to help us too much anyway, so the only real value I can offer in the early weeks is some rambling notes over who is starting and who isn't.

The priority then is to get all the underlying data work done and then put together a bumper "lineup lessons" to discuss the new transfers and a couple of unexpected surprise names.

Thanks for sticking with me through this quiet period and thanks again to everyone who bought the e-book.