THREAD
Last week I asked everyone what there there Overall rank was and how many hits they have taken this season. Over the weekend I put together all the 165 results and this is what I found:

If this is usefull/interesting please consider Retweeting!
#fpl #FPLCommunity

So is there a correlation between overall rank and hits taken? Well, sort off. When all the results with less than 16 hits taken; this is what the results show on a standard scatter diagram with a logarithmic scale representing overall rank and hits taken on the x axis.

So the diagram itself does not show a clear correlation and this is representated by the product moment correlation coefficient (R value) of 0.4. The R number is used to judge how strong a relationship between x and y values. This R value suggests there is a weak positive link.

After not being satisfied with just this result I dug a bit deeper. I wanted to look at how many hits those in the top 10k in my sample were taking compared to those who were not ranked in the top 10k.

Below is a visual representation of hits taken. The top 10K had an average of 4.9 and a median of 4. This was considerably lower than those managers outside the top 10k with an average of 7.8 and a median of 7. Another observation is that it was extremely rare...

For a manager in the top 10K to have taken more than 12 hits however this was much more common for those managers outside the top 10K. These graphics do suggest a different result to the initial scatter diagram and that less hits gives you a better chance of being in the top 10K.

I decided to put all the ranks in order and take the average hits raken of each 5 consecutive ranked managers. Below is a graphic (the second one is the first with a trendline) that qalthough fluctuates there is a clear consistent increase in hits taken when looking ...

At managers outside the top 25K. However this graphic does not conclusively show a strong correlation, at least when looking inside the top 25K.

Another graph I put together is this simple one which just shows the average hits taken for selected rank groups. A clear increase in hits taken as the rank decrease is shown although this would be expected. Nevertheless this does advocate that there could be a stronger link.

I then wanted to look at the average rank of all the managers to have taken X amount of hits. I thought it would be appropriate to take a truncated mean for this graph as this would reduce the affect outliers have as it takes out the higest rank and lowest rank from each category

This graphic does definitely suggest that managers that take less than 6 hits are at a better chance of having an improved rank. The average rank of those that had taken more than 5 hits was much more inconsistent and were at greater risk of a worse rank.

To conclude:
These graphics all suggest a correlation between Overall rank and hirs taken to a limited degree. Although these results in some part are inconclusive they’re is a strong possibility that if you take less hits you’re chances of a better finish are not guarantee...

However the odds do improve in your favour. Some other things I noticed was one manager who had taken 0 hits was around rank 150K and another manager who had taken 21 hits was also about 150K OR.

There are always managers who will never ever want to take a hit no matter what and others who will want to play more aggressive and happily take hits. The truth is there is no definitive right or wrong approach as both can work out.

Even after this I will remain a player who likes to take a -4 or even -8 when I want to; and I would encourage all managers to continue doing what they find fun and enjoyable as thats what fpl is about!

Dont take these findings as scientific evidence at all but it was definitely fun to put together as its often a heated debate in the FPL community. If you found this somewhat interesting please consider Retweeting! #fpl #fplcommunity