Dual threats

I like my information to be visualised, probably due to some kind of learning preference - the kind of thing you didn't listen to in school, but actually might have been quite useful. Anyway, for that reason I'll often plot some data - even if fairly simplistic - to better understand it and I thought I'd share a couple of simple graphs here.

We've plotted shots in the box per 90 minutes (SiB H90 or SiB A90 for home/away splits) against chances created (CC H90 or CC A90) to split players between those who have shown an ability to contribute both goals and assists and those who look like a one-category player. That, in of itself, isn't necessarily a problem, but it might indicate some concern if a player's value to date has come from, say, just goals, and the supply of those chances can be called into question.

The size of the square plotted represents each player's ownership percentage, which illustrates a couple of players who are clearly held by too high a portion of managers right now, based on their stats to date. The line represent the average SiB/CC racked up by that position at home/away:



I don't want to get into any real analysis of anyone here, as this is just a quick tool to raise questions rather than provide answers, but a couple of players do jump out:
  • Fellaini's shot totals are dominant and more than makeup for his average assists potential
  • Taarabt is certainly a statistical outlier whose dominant underlying stats simply refuse to turn into points with any consistency whatsoever. I feel like understanding why this is could potentially unlock the next piece of the puzzle for any prospective model.
  • Silva, with just 6% ownership could be a key player for the coming weeks, as based on these basic stats he's as good as anyone at home, and actually has much better stats away from home, but doesn't appear on the graph due to limited minutes played (3.0 SiB A90 and 4.4 CC A90). With City bounced from the Champions League, we'd expect his playing time to stabilise and he could be a key differential of sorts among the elite midfielders.
  • I haven't given Suarez a great deal of attention this year but if nothing else we can see just how dominant he's been this year. The sustainability of these numbers is another issue, as is the question mark over his teammates' ability to convert his created chances, but he can boast the best data among his peers so far.
  • The Rooney versus van Persie date is nicely highlighted here as while Rooney has shown more potential at home, the Dutchman has been substantially better on the road. I can see the argument that this potential inconsistency makes van Persie easier to own (as you can captain him every week) but I'm not sure that allowing for 'easy' decisions is a massive plus for a player. That said, those road numbers for Rooney are a concern and do need monitoring closely in the coming weeks. 
  • Aguero versus Tevez is clouded more by sample size concerns as the observant reader will notice that the former is absent from the away chart (due to a lack of playing time). If you like Aguero you can point to the fact that he's been outstanding in that limited time (4.5 SiB A90, 4.5 CC A90) where as Tevez fans could claim that those numbers are unsustainable if, or when, Aguero does rack up more minutes. For the record, I'm in the first camp and would rate Aguero ahead of all other forwards, if not for the next couple of fixtures which look tricky.

Comments

I have looked at your model, FFS and SoT and have come to the same conclusion on Silva. With Ben Arfa out and Fellani suspended today and some hard games coming up I figured the time to move on from them both for Silva and Morrison had come.

My only real headache coming up is if Foster is out much longer. The temptation might be to go with Mingolet with a double game week coming up for Sunderland but they have Chelsea in that week and looked ropey this lunchtime.

Amazed at Aguero's low ownership as well and figured a price rise is on thew way for those two in next few weeks.

Popular posts from this blog

Expected goals plus-minus

Selecting defenders within the same team

A brief update