Monday, January 7, 2013

Demba Ba and transfer planning

I try not to get involved in too much "will-he-won't-he" speculation in these pages with, for example, the weekly statistical rankings being provided regardless of rotation risk. I'm not saying, of course, that this shouldn't be a factor, only that I'm in no position to offer a worthwhile opinion, other than collating other news stories (which are often based on nonsense quotes from managers anyway). There are plenty of other sites who offer opinions on whether Tevez will play or if Vidic has recovered from injury but in almost all cases it's pure guess work and so I try and stay away from such discussions.

With that in mind, my initial reaction to Ba's move was negative. I have no idea whether he will play every week at Chelsea, and while you'd think it would extremely odd for a player to leave a good team to go and ride the bench, other players have done the exact same thing before. All I therefore see is a player who has gone from being the elite forward with perhaps the safest job in the league, to one who could very well only play every other week.

Medium term then, Ba has fallen a long way down my personal rankings, as for his price tag I'd prefer not to be taking on large amounts of risk unless I'm getting a player who has the potential to be dominant (such as with Rooney or Aguero). However, short term, Chelsea have a good looking double gameweek coming up and thus we get into the situation where even if you don't believe in Ba for the long term, he becomes an interesting pickup for one or two weeks (in GW23 he faces ARS at home which projects to be a useful game too).

If you're looking for Ba analysis, this is essentially the end of this piece. I love him this week and will be monitoring him medium-long term, but the presence of Torres (still a good player) means I wouldn't be happy simply plugging him in this week and assuming his production will continue as it was at St James' Park. Ba's potential for this week gives us an opportunity to look at one part of wildcard strategy which generally doesn't appear to get much attention: planning your future transfers.

The raison d'etre for playing your wildcard is often to set your team up for the significant future and replace players who might be injured or out of favour. While this is obviously the starting point, simply picking players you like for the next 8-10 weeks (or beyond) can leave some short-term value on the table. This is especially true for the next two weeks, where we have double gameweeks from elite teams and so anyone wildcarding in either of these gameweeks should consider how they will use their next two or three transfers carefully.

I'm going to use the example of forwards only here, but the logic can obviously be applied to midfielders or a mixture of both. Given the specifics of the individual double gameweeks, I'm not overly concerning myself with defenders as Chelsea are very ownable regardless of their DGW, while Arsenal, West Ham and Southampton each face a fixture in which they look likely to concede so defensive value may be limited.

Let's look at the model forecast for the next eight weeks for the elite and widely held forwards:

I don't wish to get into the specific merits/drawbacks of the model (and freely admit there are many of the latter), suffice to say that facing weaker opponents is a good thing and two games are better than one (van Persie's P90 of 8.0 for the season would be matched merely by someone like Pogrebynak (4.0 P90) having two average games).

If you were settling on a team for the next eight gameweeks, and placed little emphasis on the double gameweeks, you might settle on a front trio of van Persie, Suarez and Berbatov. Indeed, these three options are the most widely held trio available without getting into extreme circumstances where you're spending 35m+ on your front line. That would set you back 31.5m at today's prices and give the below forecast points for the next eight gameweeks:

This looks a very strong platform to build on every week, picking up double digit points from a captained van Persie every week (who will obviously outperform these numbers by a large stretch every now and then), along with strong production from your other two forwards. In total we get 174 points for our 31.5m, which translates to a P$ of 0.690. If we were at a normal stage in the stage I'd say this situation is a great one to be in. However, wildcard season affords unique opportunities, especially when combined with double gameweeks, and thus we can quickly find different use of assets to maximise our points returns. I haven't run an analysis to see which would be optimal, but take the below as an example:

This strategy allows you to take advantage of Ba's double gameweek, while eliminating his long term risk, as well as grabbing Podolski for his two games next week as well. We also get the benefit of cutting Suarez loose during his tough run of games, and enjoying some extra time to decide whether to invest in Rooney or Aguero, depending on recovers from their injury first. You'll also note that in GW28 I move back to Ba (assuming he's proven himself by this point) but the point was really just to illustrate that by that point we're in a great posittion of flexibility and could easily stick with Berbatov or move to someone like Ba, Cisse, Defoe or any new arrivals in the transfer window. This strategy gives us 190 total points for an average investment of just 29.1 (with a peak of 29.9m) for a P$ of 0.816. The price we pay is the loss of future flexibility with transfers in GW23 (Ba>Podolski), GW24 (van Persie>Rooney), GW25 (Podolski>Aguero) and GW26 (Rooney>Suarez) all accounted for, but there would of course be some flexibility to cut Rooney out altogether if his recovery is slow or Kagawa has a negative impact on his positioning

Now I understand that maximising double gameweeks is not exactly a novel idea but this is a unique opportunity that I believe too many people are going to bypass purely because they simply refuse to part with van Persie. His presence in your team doesn't preclude you from engaging in this kind of strategy completely, but it does of course limit your flexibility and while, for example, flipping Berbatov to Lambert for this week is a nice idea, it's still not going to give you the kind of ceiling (and higher floor) you get by playing the above game. You'll note that I haven't even included Lambert in the above strategy, instead sticking with Berbatov throughout. This is due to a desire to avoid paying for transfers at a later point as I would want the Bulgarian long term, but if you (a) are happy to fork out the four points or (b) believe Lambert to be Berbatov's equal after the double gameweek, his inclusion could make the strategy even better.

The same principle applies in midfield of course, with both key double gameweek teams (Chelsea and Arsenal) boasting two great options (better than the front men indeed). I have focused on forwards for two reasons:
  1. I believe the likes of Mata, Cazorla and Walcott to be good enough to own and play every week without the risk or downside observed in the forward options like Ba, Torres, Podolski or Giroud.
  2. If you were to heavily invest in Chelsea and Arsenal midfielders for the double gameweeks, planning to move them at a later date to, say, Bale and Fellaini, you are then left with the tricky proposition of who to play up front. It's reasonable to be spending 33-34m on your back seven if you so wish, but then investing the 44m in midfield required to grab four elite prospects would leave you with just 25m or so up front, leaving you with just one elite forward. It can be done of course, but for me I'm much happier projecting excellent second halves for the mid level midfielders like Fellaini and Michu than I am for the likes of Benteke or Lambert, who would be integral to such a strategy.
In midfield then, Mata ranks as the top midfielder per the model by a distance so would be my number one target for this week, given that you can grab him then happily plug him into your lineup for the foreseeable future (I'm okay with the slight increase in risk that Ba brings, who could vulture some minutes in one of the three attacking midfield slots). Hazard, I'm less keen on and while he should be an asset for this week, I wouldn't pay to get him given his somewhat shaky future prospects (his underlying stats have been concerning for a while).

As for Walcott and Cazorla, the model gives them almost the exact same forecast, which would lead one to conclude that Walcott is the best pick, given his cheaper price and the fact that the model is still using some data from when Walcott was deployed in midfield. They don't place much higher than Hazard in average forecast points but I'd be happier with this pair's risk level and thus, like Mata, I'd again be happy plugging either of this pair into your team without the need to play around for the double gameweeks. My only issue is how to accommodate Bale, who still looks like the best combination of upside/security, but with three away games and United at home in the next four games, I'm okay staying away for a while, then probably making that move if/when we know Ba's impact and the sustainability of Walcott's minutes up front.

I hope the above gives those of you wildcarding a few ideas as to how a little bit of planning can go a long way to maximising your budget and will hopefully lead to some sounds decisions and transfers in the next few weeks.


CDI said...

Great stuff as always Chris. I was wondering if you wouldn't mind giving me some input on my current W/C team. I like your rotation ideas in a vacume but I don't think I'd have to go to such lengths with my current TV.

Jaaskelainen, Begovic

Gibbs, Harte, Turner, Azpilicueta, Davies

Fellaini, Mata, Bale, Walcott, Michu

RVP, Ba, Lambert

I'll have a real benching headache this week but I plan to downgrade Lam soon enough and just play 3 5 2 every week. If Ba does face alot of rotation then I'll somehow jam Suarez back in for him.

stooshermadness said...

This is a real eye-opener, Chris, thanks. You have crystalized a vague sense I had with respect to using my WC early. I am on my WC because I really needed to strengthen my shambolic defense and the DGWs offered a chance to get some extra juice out of a CHE and ARS defender; and because I was holding only Hazard from Chelsea, too little for what looks like a profitable DGW for CHE. But, also, as you say, these DGWs offer a really unique opportunity and some relatively easy transfer options ahead. I had moved for Podolski 2 weeks ago in anticipation of the DGW23, but I swapped him out last night on my WC for Ba, perhaps for only a 1 week flyer. I've had RvP since GW2, but your chart above makes a convincing case that it may be time to let go soon. Great analysis, thanks.

Hitman said...


Super analysis. Wanted to tell you that I am a big fan and never finalise my GW transfes without reading your blog.

Am a decent hand at excel and make financial models for a living. I would love to be your apprentice if you ever are looking for one. :)


malini ecorp said...

Thanks for sharing this useful info. Keep updating same way.
Regards,Siddu Corporate Training

Raju Kumar said...
This comment has been removed by the author.
Raju Kumar said...

such a good website and given to more information thanks! and more visit
sas online training

sss lokesh said...

Very important article on Tableau.... this is the great course for all to get skill updates
Websphere Training In Hyderabad

Srinu Vasu said...

great article
SAP GTS Training In Hyderabad

Cse It said...

I am highly thankful for this post. I’ve been looking everywhere for this! Thank kindness I found it on bloging effort. You’ve made my day! Thank you again!

sap isu billing&invoice training

sap ehs training

sap tm training

sap fica training

sap hana development training

Rakesh S said...

valuable information..
Big data training .All the basic and get the full knowledge of hadoop.
Big data training

Raja B said...

Hadoop bigdata online training,with all the basic knowledge.
hadoop online training

vignes waran said...

I want to learn from something useful topics our blog site post , such a fantastic , I definitely read enjoyed your post.

Android Training in Chennai
Android Training Institute in Chennai

ricky branch said...

Thanks for sharing the post. Really a useful post and can be acceptable for every tableau users.

Tableau Training in Hyderabad

Ohhani Baek said...

you are written the Nice information, Thanks for share the useful post for the tableau people.
Tableau Course In Hyderabad