The Week Ahead - Season Opener

Rotoworld
The Week Ahead - Season Opener
.

View photo

The maiden voyage of 'The Week Ahead' on RotoWorld.com with some input from a statistical modeling expert …

So, here we are in our new home but with that familiar thrilling feeling of another new Premier League season about to kick-off.  We've been getting used to our new surroundings through a long summer of over-hyped transfer rumors (otherwise known as "every summer" in the Premier League) and the occasional meaningful evaluation of moves that have actually happened.  Now, it is time to get back down to business and start writing about actual matches and actual fantasy decisions again.  

As you are all aware by now, Yahoo has done some renovations to their fantasy game this summer.  There are things that I really like about where it is and there are things that I am less excited about.  At the end of the day though, the thing that I'm most excited about is that Yahoo seems committed to continually examine their game year-over-year with the purpose of taking it to new heights.  I know there has been grumbling in some corners about this or that specific change to the game, that is inevitable with any change.  Hearing feedback is a good thing for the evolutionary process. I would like to assure everyone that we do pass along any constructive feedback that we receive to the Yahoo team so that they have it to consider for future iterations and evolutions.  For those of you who haven't been thrilled with all of the changes, it is my hope that you choose to a) give them a chance before you render a final decision; b) provide constructive suggestions (and some reasoning behind your suggestions) as the season unfolds; and c) stick with the game because trust me when I say that what I have heard about long term vision has me very excited.

Follow RotoWorld_PL on Twitter | Follow the team on Twitter: Neal | Nik | Jeremy | Galin 

For those unfamiliar with my column from it's previous homes, we are in the middle of the part of the column where I write a lengthy pre-amble that may or may not have anything to do with the fantasy player picks that you're probably here for.  As it turns out, the schedule-makers have undermined my first Week Ahead column a bit by having Chelsea and Aston Villa play two matches this match week while everyone else plays but one.  That makes for a relatively uninteresting player picks section of the column because there are relatively few players to choose from.  In fact, given Chelsea's home match-ups with Hull City and Aston Villa I'm going to limit myself to a quick Chelsea-only rundown by position and leave the more expansive Player Picks column to Nik tomorrow. 


To fill up the many words that I would ordinarily have written about various player options (and I do go on from time-to-time as long-time readers will no doubt tell you) I am going to add an extra section to this week's column.  That section will be comprised of a summary of statistical projections of season-long player performance from Aaron Nielsen of SoccerMetrics who has adapted a number of player projection models from North American sports (baseball, basketball, etc.) and appied them very effectively to soccer.  He has been particularly effective at predicting performance for players moving from one league to another regardless of whether they are moving "laterally" (from Germany/Italy/Spain to the Premier League) or moving "up" (from Holland/Belguim/France/MLS/Scotland to the Premier League).  He has created projections for every player in the Premier League and I'll share some of the ones I found particularly interesting on the eve of the new season.

Team-by-Team Season Previews: ARS | AST | CHE | EVE | FULMAC | MAN | SOU  | TOT

Chelsea Player Picks

Forward - Juan Mata didn't travel with Spain and seems very likely to start both matches along with Eden Hazard who played 74 minutes against France in Brussels.  The big question is which forward between Romelu Lukaku, Fernando Torres, and Demba Ba starts and whether anyone has the balls to think they know the answer for both matches.  It will probably be easier on all of our hearts to pick either Christian Benteke (who has two difficult matches at Arsenal and Chelsea but is almost guaranteed to start both) or Roberto Soldado (who only has one match but it should be a profitable one for the debutant).

Midfield - What will Jose Mourinho do with the "Frank Lampard spot"? Will he start his old pal Lamps and, if so, will he get both matches? Will he slot Oscar (who played 59 minutes in Basle in mid-week) and/or Andre Schurrle (who played the final 10 minutes in Kaiserslautern) in for one of the matches in deference to Lampard's age? I'm not particularly interested in any of Chelsea's other midfielders even with two matches so my decision here will depend on how good I'm feeling about one of those three guys.

Defense - The easy decision is Ashley Cole here.  He isn't the attacking force that he used to be but it seems likely that Chelsea will get at least one clean sheet between the two matches and phantom points should make ACole a good buy over two starts.  The rest comes down to how Mourinho aligns his defensive resources.  My inclination is to make Ivanovic my second choice figuring that he'll start at either center back or right back.  

Goalkeeper - Petr Cech seems like a no-brainer given the way budgets are working out, doesn't he?  

Channeling Aaron Nielsen: Player Predictions for the 2013-14 Season

We here at the blog (I guess we were at the old blog at the time), were all over Michu and Santi Cazorla as exceptional bargains last season and we had generally positive feelings about Rickie Lambert but were dubious of an older player coming up from the lower divisions for the first time.  The reason I am paying particular attention when Aaron gives me his predictions is that he not only called Michu and Cazorla but he was also emphatically behind Rickie Lambert and Christian Benteke before they started showing their stuff in the Prem.  With that in mind, I've poured through his projections and come up with some interesting ones that I'll break up into two lists - New Players to the League and Returning Players:

New Players 

  • Roberto Soldado (Projection 16 goals; 4 assists) - Aaron has Soldado as the most impactful of the new arrivals by a hair over...
  • Wilfried Bony (15 goals; 4 assists) - Imagine if Swansea had had a real forward with those stats to pair with Michu last sesaon.
  • Alvaro Negredo (13 goals; 3 assists) - Given City's likely rotation between Negredo, Aguero, Jovetic, and Dzeko much more would seem implausible.  The surprising part is how close that is to...
  • Gary Hooper (12 goals; 4 assists) - Norwich regains the production lost when Grant Holt fell off a cliff last season.
  • Jozy Altidore (11 goals; 2 assists) - After watching the highlights of his second half hattrick against Bosnia today this feels a little low but this is the difference between emotional analyses and statistical ones. 
  • Stevan Jovetic (10 goals; 5 assists) - This isnt an indictment of the player so much as a reflection of a projected minutes played total under 2000 (compared to 2700+ for Robin van Persie or Gareth Bale).  If you can pick out when he'll start, he'll be a very effective selection on a per match basis.
  • Rickey Van Wolfswinkle (9 goals; 3 assists) - Not stellar but when combined with Hooper's projection, that's some solid work for Norwich over the summer. 
  • Andreas Cornelius (9 goals; 2 assists) - Given his age, that bodes well for him in future seasons whether Cardiff stay up or not.
  • Matej Vydra (8 goals; 5 assists) - He's only projected to be playing about 1500 minutes so the projected totals could climb to impressive levels if he gets more regular playing time.  Like Jovetic, the key here is knowing when he'll start and pouncing. 
  • Andre Schurrle (7 goals; 4 assists) - I was surprised by this one because of the presence of Lampard and all of the other attacking talent at Chelsea but Aaron sees Schurrle producing very strong numbers with only limited playing time. 
  • Nicklas Helenius (5G; 3A), Nacer Chadli (4G, 3A), Jesus Navas (2G, 6A) - Aaron's projects were particularly unkind to these three bigger names that have been the source of some optimism among fantasy managers. 
Returning Players
  • Luis Suarez (15G; 5A) - Part of this will be due to missing matches and part is likely moderating between his conversion rate on shots from two seasons ago (he was terrible at it) and last season (when he was much better).  
  • Andy Carroll (14G; 5A) - Goals and assists aren't everything in the real game on the field but wouldn't it be something if the man Liverpool sold at a huge loss was this close to the man they apparently won't be selling to Arsenal for an inflated price?  
  • Olivier Giroud (14G; 3A) - So, the model has Giroud performing very similarly to Suarez in a similar number of minutes.  If Aaron turns out to be right then that offer for Suarez is going to look pretty silly. 
  • Shinji Kagawa (10G; 6A) - Aaron's model shares my optimism over Kagawa's potential under David Moyes, it isn't the reason I like Aaron's work but it certainly doesn't hurt.
  • Kevin Mirallas (9G; 2A) - Apparently Aaron's model doesn't think much of the Belgian's chances of staying healthy because he has him down for a mere 1500 (ok 1532) minutes.  The production is quite good given the assumption on playing time but the implication is that he'll be a huge headache for fantasy managers again. 
  • Michu (10G; 5A) - So, here's the downside of Bony's projection. My read is that Bony will take goals from Michu rather than adding much to Swansea's overall total.  There could also be some regression to the mean in this calculation as Michu's total last year surpassed any previous season in Spain. 
  • Christian Benteke (11G; 4A) - Speaking of some regression, Benteke's totals are projected to do the same after an exceptional debut season at Villa.
  • Robert Snodgrass (8G; 8A) - I don't know what Aaron thinks of Norwich's defense but between Hooper, RvW and Snodgrass he has Norwich as a pretty formidable attacking side. (FYI, he also has Becchio and Pilkington chipping in with 5 goals each - good times at Carrow Road).  
  • Frank Lampard (8G; 3A) - This model has Lampard in what amounts to a job-share with Andre Schurrle that will produce 15 goals and 7 assists - not bad production from that spot, huh?  Too bad for fantasy managers that we'll have to be guessing each week which one will get the start and the points. 
  • Marouane Fellaini (7G; 6A) - He definitely has the big/big-haired Belgian regressing signficantly from last season's production. 
I could go on all night and I want to leave you with the caveat to all of the above that the projections are based on current information about injuries to the players in question and other players at the same position on their team.  If Frank Lampard were to get hurt then Andre Schurrle's projection would likely take a dramatic turn for the better due to an expectation of more playing time while Lampard's would take a corresponding nosedive. There are also elements of managerial preference involved (I didn't discuss the three Chelsea forwards but the model attempts to predict what a new manager will do with 3 players who bring entirely different things - both good and bad - to the table).  I suppose what I'm saying is that as with any predictive models, even ones like Aaron's that have proven very effective in the past, use your good judgment when circumstances change. 
 
I'd like to close out by thanking Aaron for discussing his model with me over the phone and providing me his projections to use in this column.  If you want more from Aaron (and you should) follow him on Twitter.  
View Comments