Thursday, November 7, 2013

Points versus expected points

The below chart is something I've posted before and am tweaking for the new website so I thought I should post it now and get your feedback while also hopefully providing a quick view to identifying some players to research further:


3 comments:

Pulma said...

Love this article, shows guys who have underperformed. Need to watch who are perennial underperformers though like Walters.

Tony S said...

When I look at this chart I don't see underperformers or overperformers all I see is players on form who are capitalising on their chances (Ramsey and Rooney) and players who are of such quality that they will always out perform the normal (Aguero and Yaya) ........ and on the otherside are players out of form / low on confidence who should be avoided.
I would personnally be looking to take the players over achieving as opposed to those 'due a score'.
I believe that if you reconstructed this table after xmas with a start date of tomorrow then the same players would occupy the two sides. This is where I believe the value is in this chart. This chart highlights players who will consistantly score and players who will ALWAYS underachieve.
Make sure you have a team of overachievers :) they always come out on top :)

Chris Glover said...

Thanks for the comments guys.

There are some players who could be considered consistent "underperfomers" though that's more to do with the lack of quality of the model than any real science. For example, players who shoot from stupid angles/distance like Taarabt will never regress to the mean and thus could be labelled "underachievers" every year.

That said, this logic applies to VERY few players.
- No one saw Ramsey coming
- Suarez has been a big "underachiever" in recent years - I wrote a post on this at FFS last Christmas
- Michu outperformed his stats by a mile last year. Long time readers will recall me forecasting a slow down for about 15 straight weeks!
- van Persie WAY outscored his underlying stats last season.

Some players have done well in consecutive years (Yaya is the main one who jumps out) but the same could be said for a coin which comes up HHHHH in two separate trials. Toss enough coins and every outcome will eventually happen.

My examples above are obviously cherry picked to prove a point so I will try and plot last year's P vs xP +/- from, say, last Christmas with this year and see if we spot a trend.

My hypothesis is that some players will absolutely have a chance to outperform their stats every year, namely those with superior skills and especially those on a better team. However that only goes so far. I've found very little evidence for example that a player's G/SoT is sustainable and will generally regress to the league wide mean within roughly one StdDev. So while someone like Aguero might be able to sustain, say, a higher SOT% rate or continue to receive a lot of chances, if he's hitting the net with every shot, the regression monster absolutely will come for him.

I'll try and get that post up by the weekend. Good idea!