Thanks for all the work you put into this guys, really appreciate it and find it very interesting. I did some work on PPG analysis back in October and put it in the Stats thread but it occurs to me that you might find it interesting and applicable to your work here and can even use some of the data for benchmarking. Sorry if it's stuff you already know but I thought it was worth including just in case you missed it.
Basically, I was interested in 2 things at the time:
1) Analysing PPG data w.r.t. a team's final positions over the course of the Premiership.
2) Seeing if it was possible to split the PPG into smaller groups very much like jp does here with the top and bottom teams' PPG.
I collected all the data from 2000 onwards to generate a big enough sample size and got to work. I broke the teams into quarters, thirds and halfs to see if segmenting the teams could be useful, the first table here is the summary output from my work. What I found interesting was that a PPG of 1.83 (1st table: No Groups Column) on average was enough for CL qualification and not the 2 PPG I originally thought was the norm. I was quite concerned my data was wrong so I re-did the work looking at median values instead - enabling me to ignore any outliers that may have thrown the dataset - but I still got very similar results (differing from the average by 0.05). I then had a look at the last 3 seasons (the 2nd table here) and realised that my average results seemed about right.
Just to be on the safe side however, for one last check I decided to plot the Points data on a graph and added some trendlines to them to see if the historic data had anything to say and came up with this:
Basically for the top three positions there's a minimal trend with regards to Final Points over the course of the last 17 years, however the next three positions shows a more pronounced upwards trend, pretty much reinforcing the fact that the EPL has become more competitive over the last few years, especially for that last CL spot.
That means that for positions 1-3 we can pretty much state with a reasonable level of confidence that 2.3 PPG will,
on average, win you the league, 2.1 PPG will get you 2nd and 2 PPG will get you 3rd.
That 4th spot is a little bit more ambiguous and should in effect be slightly higher than the 1.83 I recorded. Looking at the 2nd table ( ie the last three seasons) the data here implies it should be 1.86 but I suspect it might be slightly higher, possibly closer to 1.9.
With regards to segmenting the teams into smaller groups, I found some small anomalies in bracketing the teams into quarters and thirds that has me 2nd guessing their usefulness. JP here breaks the EPL into two groups which is about the right amount of segmentation that one would want especially when teams can easily flip from one group to another based on their results throughout a season. You can use the data from the first table (2 groups) for benchmarking with a similar level of confidence here, as per before position 1-3 should be about right with the 4th position maybe being a little bit higher than recorded here.
