Monday, November 21, 2011

Playing Your Card Right Part I: Chasing Points

It's getting to that time of year when fantasy squads start to develop too many common threads and managers begin to longingly look at that 'play wildcard' option when making transfers. With the second wildcard window looming in GW20-23, realistically the two strategies are to:
  • play your first wildcard now, re-load in GW23 and hope to survive the remainder of the year with that team, or
  • hold off until GW20, maximise your squad for the busy Christmas period (possibly taking advantage of the inevitable DGWs) and then switch out to a new team in or around GW30 for the final stretch.
There are advantages to both strategies and in all honesty, the fact you have made it this long without using your wildcard puts you ahead of many other managers (even if you have used your first wildcard, you might still pick up a few things to thing about below, so keep reading!) 

The choice of when to use your wildcard is going to greatly depend on the idiosyncrasies of your league. 
  • How far behind/ahead are you? 
  • Have all your competitors used theirs already? 
  • Are injuries/suspensions starting to rack up for your team? 
What won't change however are the kind of things you should consider when deciding when, and how, to use your wildcard. The first thing we are going to look at is the distinction between chasing points, genuine stars and those players due for a strong second half.

Chasing Points vs. Genuine Stars
I would suggest that the most common mistake when making transfers in general, but more specifically when playing the wildcard, is placing too much weight on what has happened (particularly recently) and not enough on what will happen. Of course, none of us know exactly what will happen in the future, but the data available to us can give us a good idea, at least to the point that will be better than simply ranking all available players by points scored to date.

We've touched on the principle before, but it will form an important part of this analysis so it bears repeating again. Key passes made (KP) and shots attempted (Sh) have been good indicators of assists and goals respectively and so a player who has done well in these areas but has yet to see any production should be in for a improvement in fortunes in the future (or vice versa). It is worth noting though that, generally, their production will not 'over correct' to compensate for the weeks of below average production, but normalise to expected levels from now on. For example, if we say that on average one of every 10 key passes is converted into an assist, and a player has 30 key passes with just 1 assist, we would not expect 3 of his next 10 passes to be converted (thus giving him 4 from 40) but rather that his production will regress to a normal rate of 1 from 10 rather than his current rate of 1 from 30 in the future. That is still however an increase in production from the season to date and thus could indicate that a player is currently undervalued (and therefore under owned).

I had some concerns about being overly simplistic with this metric but looking at last season we see that the vast majority of players who played at least 10 games had their key passes converted at a rate of between 0 and 15 key passes per assist (for a league average of 10.5). We can therefore use this measure (with some care) to suggest that players who are seeing their key passes converted at a substantially lower rate will be due for an increase in fortune in the future (or vice versa).

It's a similar story for shots, which enjoyed an 80% correlation to goals last season. On average a goal was scored for every 9.8 shots, so again, if a player is taking shots but not getting returns, we might forecast him to have success in the future and thus be a target for your wildcard transfers.

On the flip side, you've got those players where everything they have touched has turned to gold (or goals as the case may be) and history tells us their productivity will (likely) decline. Of course, there are exceptions every year and statistics won't necessarily average out over 38 games (Mulumbu scored 6 goals from 8 shots last season) but if you're still with me, you probably have at least an interest in playing the odds.

With that in mind, below are a sampling of the top players whose underlying stats have not yet been converted into fantasy points, and those who appear to have been overly successful so far:

xG is the expected goals based on the shots taken to date and xA is the expected assists based on the key passes made. xFP is the expected fantasy points 'missing' from a player's scoresheet based on the underlying stats. Now, to be clear, we are not saying that a player will 'recover' these points, but that Moses is playing more like a 4 point player than a 2 point player and should play at that level in the future, all things being equal.

I don't want to reproduce too many of the stats from FFS as they pay Opta for the privilege but anyone who has a subscription can see the above data for all players and the logic should stand up for most players. The one group who this probably doesn't apply to are the elite class of forwards. The elite strikers have tended to convert at a much better rate than the average player, generally knocking in a goal every three or four shots. This holds up this year with Van Persie (13 from 43), Aguero (10 from 33) and Rooney (9 from 40) but these players are less important for this kind of analysis anyway so we won't dwell here.

I will reproduce the short list later in the week to highlight widely held players who fair badly on this, and other, analysis and thus should probably be avoided when playing your wildcard. Also stay tuned for the next few themes to consider when playing your wildcard including how to build a squad rather than a team, differentiating your team and playing the fixture list.


vanilla said...

I love this blog... it's now a bookmark I check every day, and it's a selling point for signing up to Twitter

CDI said...

Excellent blog sir. Rooney's underline stats and his fixtures are whats stopping me from kneejerking him out of my team but at what point do you have to cut your losses? Also can you see Walters and klansic keeping up their PPG?

Ratchet said...

love it!

I suspected Suarez is due for big scores soon and your analysis confirms it.

Daniel said...

Well that comment about RVP, Rooney and Aguero, meaning the elite strikers have tended to convert at a much better rate than the average player, generally knocking in a goal every three or four shots, should be applied and to Lamps so he is not over achiever, he is just better finisher 6 from 17. And regarding Suarez his finishing is awful, he'll never be close to average. He'll always need more shots to score. And I wonder what will happen when you are gonna include Chicha and Bent in this analysis.

Cheers and keep up with the good work.

WS said...

Which one is a better player for long-term? Nani or Young?

PUl615 said...

Love the writing, love the stats. Any way to make similar over/underperforming lists on defenders?