Player Analysis: By The Numbers Week 8
Some weeks just knock you for six in the Premier League and last week was just one of those. It started with Spurs thrashing City, a result that I doubt anyone in the football community could have predicted and continued to just get better and better. I don’t know if any of you were counting but there were 41 goals this weekend, that’s an average of 4.1 per game which considering it’s usually 2.5 shows just how much fun it was!
Of course, all of these goals led to fantasy points:
I managed a score of 198.2 in the Fantrax AMCR league moving me up to 7th overall and thrashing my highest week to date.
I’ve been trying to finalise a strategy for Mondogoal and managed a profit of £27 in my first week – It doesn’t sound like much but with two non-players it’s pretty spectacular.
As I was focussing on Mondogoal I had to abandoned my Beat the Bloggers team, who promptly scored 143 – my highest score of the season.
I hit a score of 208.75 in Perfect XI – the second highest in the Rotoworld league, moving me up to 17th in the overall game – I don’t recognise any of the names above me but please let me know if you’re a reader.
So what am I bringing you this week? Well, it’s time for a change.
Goalkeepers
The writing of my weekly column is the easy part, the difficult part is crunching all of the numbers to gain some insight. The most time consuming of these insights is the goalkeeper prediction. This wasn’t so bad when the focus was on YFF but with Mondogoal, multiple Fantrax leagues, PL.com, etc it’s too much to do. The goalkeeper predictor is being returned from service. If you know your way around Excel then I’m happy to send readers a copy to use for their own teams – easiest way is to contact me either through twitter (@funtasyfootball) or through the contact form on my website www.imawinner.co.uk.
So, with that mainstay retired, what will I be bringing you? Well, to begin with please welcome back ‘The Predictor’.
The Predictor
This predictor, as opposed to the one used in recent weeks, isn’t my work at all. The data behind it belongs to Joe over at football-data.co.uk, If you want downloadable .csv files of everything that happen in European games over the last 20 years then that’s the place to go.
Joe effectively modelled the performance of teams in the premier league and compared it to their previous home/away 6 game performance. It turns out that it’s remarkably good. The keen eyed amongst you will immediately note that there haven’t been 6 home games and 6 away games for each team yet. That’s true so please take its predictions with a pinch of salt for the next few weeks.
Next up is……The Poisson Predictor
The Poisson Predictor
Readers from last season will be well aware of this prediction tool but for new readers here’s a recap. The Poisson is a statistical probability distribution named unsurprisingly after Simeon Denis Poisson. Whilst there are a number of probability distributions the Poisson binomial is used for a number of successes in a series of yes/no experiments where the probabilities of success differ. Or as we use it, the probabilities of goals where two teams are of varying strength. When I last introduced the Poisson I showed its performance against the previous premier league and it’s pretty good. The trouble is that it is at its most accurate in the last game of the season when the average goals per game is completely accurate. Every other week of the season it is a good indicator of where goals (and equally importantly clean sheets) should be coming from.
What does it tell us this week? Well to start with Palace vs. West Brom and Villa vs. Stoke look to be dull, dull, dull – Both games have over a 65% chance of 2 goals or less. Bournemouth vs. Watford isn’t a whole lot better. City look ridiculously strong against Newcastle which is to be expected. The reason the graph looks odd is because Newcastle have yet to actually score an away goal and I doubt one will come this time.
Norwich vs. Leicester looks like a far more fun affair to me with Jerome, Vardy, Mahrez etc offering good choices – avoid the defence though. Similarly, West ham look like they’ll be scoring for fun although they do seem to be great against the best teams and average against the worst.
Chelsea look to be 1-0 or 2-0 against Southampton which felt right a few weeks ago but lately I’m not so sure. The score line looks the same on Merseyside but that’ll depend on Lukaku and Sturridge I guess.
The final two games I believe suffer from the early lack of data. United are listed as favourite and maybe they are but 1-1 seems the right score to me. Similarly Swansea are favourites at home to Spurs but you have to question the impact Spurs developing streak will have on them.
So, that’s me done for this week. I’ll speak to you next Tuesday.
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.
