Player Analysis: By The Numbers Week 2
Wow, what a start to the season! Arsenal lose, Chelsea draw and lose Courtois, Everton scrape a point against a newly promoted team and West Ham have a 16 year old with a 95% pass completion percentage. Who’d have guessed? Well, I did predict the fate of the newly promoted teams in my earlier post…
Of course the fallout from these performances is that the vast majority of players started their premier league campaign poorly. Personally, I did about average in PlayTogga, PL.com and Fantrax. The positive thing about starting slowly though is that generally you’re going to improve going forward!
A starter for 10
Now regular readers will be aware that this early part of the season is a nightmare for predicting results from a statistical performance. There’s very little in the way of reliable recent form and statistical forecasting using things like the Poisson distribution are wholly inaccurate until later in the season. Even the great ‘predictor’ we used last season required the goal difference from the last six games. So what can we do? Well, my regular readers will also know of my love for getting bookmakers to do the work.
These statistics come from Bet365 but not in the way you’d expect. Instead of converting the odds into percentages, this forecast comes from analysing ten seasons of actual performances and comparing them to the odds offered on the home team. I have found this to be an excellent approach because it removes any concerns regarding the bookmakers over-round and their desire to balance their books.
The table shows that an abundance of Liverpool, Spurs, City and United players are the order of the day for this week – no surprises there. The interesting thing from a statistical point of view is that the predictions of Watford against West Brom are identical to the big match-up of the weekend between City and Chelsea. You could argue that an investment in Watford is just as likely to pay off but at a substantially lower player cost.
For those readers who’d like to be able to work these out yourselves then you can download this fantasy football cheat sheets from my website.
Goalkeepers
It would appear that everyone’s favourite feature last season was the goalkeeper predictions – It was one of mine also. Unfortunately the graph does little to help in PlayTogga’s Perfect XI as there’s no requirement to look for value for money. It does help however for the ‘Beat the Bloggers’ league in Fantrax. Here it is:
As it’s still early in the season there are a lot of goalkeepers still valued at 5000 which rather skews the figures. Looking at it though, the best value appears to blatantly be from Romero who has the lowest price and the second highest expectancy.
Moving on to the premier league game, we get a slightly differing view. Please note that these figures are based on my guesstimate of performance over a number of weeks due to the transfer restrictions in the game, therefore the right hand axis is for comparison only:
The pricing is far less intense in the premier league game but every little helps so it would appear Mignolet, Adrian, Stekeleburg, Pantilimon, Myhill, Gomes, Butland, Schmeichal and Guzan in that order. Personally, I have Myhill in my team after week one and he has the largest gap between cost and expected performance. You may wonder why Chelsea and United are scoring so low – It’s based on the expectation that their scores will be divided amongst their goalkeepers. Why are Crystal Palace scored so lowly? Well they fancy Arsenal, Chelsea, Man City and Spurs in the next five matches – you’ll have to be crazy!
If you chose well at the start of the season, you may wish to stick with what you’ve got. Creating a bank of transfers you can call on is a good idea to combat potential injuries and rotations.
That’s it for me – see you next week.
This entry passed through the Full-Text RSS service – if this is your content and you’re reading it on someone else’s site, please read the FAQ at fivefilters.org/content-only/faq.php#publishers.
