Shameless plug here - I wrote a blogpost about our defensive issues this season and NareysToePoke blog kindly published it

http://nareystoepoker.blogspot.co.uk/2017/01/caley-thistles-case-for-defence.html

Hope people find it interesting, any feedback welcome.

Great post. Interesting to read about how new signing Brad McKay is doing too. 

Needs more graphs.

Any idea on what source data is being used?  I only ask because it's became increasingly apparent this season that the Opta data is very unreliable.

I actually suspect it's maybe not Opta data as the info looks almost identical to analysis that the club has access to from elsewhere.

Strike the bit about thinking it may not be Opta as I see on screenshots from the source twitter account for the graphs being produced that it is Opta based.

14 minutes ago, CaleyD said:

Any idea on what source data is being used?  I only ask because it's became increasingly apparent this season that the Opta data is very unreliable.

I actually suspect it's maybe not Opta data as the info looks almost identical to analysis that the club has access to from elsewhere.

The data for the radar charts was supplied by @TheSPFLRadar - he can give you more details if you want to message him directly on Twitter.

All the other data I did myself by going through all our conceded goals for ast season and this one. A fun way to spend a Friday night.

Just now, ictchris said:

The data for the radar charts was supplied by @TheSPFLRadar - he can give you more details if you want to message him directly on Twitter.

All the other data I did myself by going through all our conceded goals for ast season and this one. A fun way to spend a Friday night.

Not doubting the methodology (I like the concept of the article), but I have highlighted big flaws in the Opta data several times this season; where numbers have been hugely out of sync with reality. So much so that it's what lead me to look into it in the first place.  They're struggling to get some of the simple things right like number of shots, fouls, offsides etc (they've even missed substitutions more than once) and that leaves me in doubt as to how accurate the rest of what they're doing is.

Even more disappointing when you consider that the SPFL are paying them large sums of money for licenses to use their data and information....as are the likes of Sky, BBC, BT....which gives the perception that they should be considered an authority on the matter and as such most people don't question the output....especially when that output fits with their untested perceptions.

There's definitely a paucity of data on Scottish football, I suppose to be expected for a smaller country. There's a lot of dedicated amateurs, which can be better, but there's a lot of measures you see for the English leagues that aren't available easily in Scotland.

Having checked with the @TheSPFLRadar the data for the radar charts isn't directly from Opta.

 

 

 

15 hours ago, ictchris said:

There's definitely a paucity of data on Scottish football, I suppose to be expected for a smaller country. There's a lot of dedicated amateurs, which can be better, but there's a lot of measures you see for the English leagues that aren't available easily in Scotland.

Having checked with the @TheSPFLRadar the data for the radar charts isn't directly from Opta.

Agreed.

Even with the amateurs, a lot of them are pulling date from sources which use Opta and just presenting it differently.

Because it's such an expensive and time consuming process there's very few companies doing it and price means that sources tend to stick with a single source....so you don't often get cross comparison of the data.  As I said, Opta's accuracy only really raised alarm bells with me this season when they made a few big errors and I then went to other sources, as well as running some checks on stuff by analysing a couple of games myself.

Admittedly, missing a couple of passes when looking at dozens of them is not going to cause the same level of discrepancy as missing a couple of shots in a game where there's a much fewer number, so it's possibly not as big a deal....but, as a statto, I hate it when people give you crap data and expect you to extrapolate information that is then presented as fact.

Taking the figures at face value...

 

Looking at the areas of the opposition goal scoring moves, it looks like we need a sweeper to cover moves into the 18 yard box.  The figures don't reflect the number of times players are caught out of position resulting in a goal which I suspect might be a contributing factor, combine this with the "40% of through balls" fact.  A sweeper might help this too.  Our "defensive dual" success seems low for everybody so someone sitting deep would be able to double up on tackles would help this although at the expense of a deeper back line, Meekings would have the pace to track through balls more then any other options.

 

Suggest;

                          OFW

                       Meekings                        

Raven    McKay    Warren   Tremarco

                           Draper

              Doran   Tansey  King

                        Doumbouya

48 minutes ago, Northern_jaggie said:

Taking the figures at face value...

 

Looking at the areas of the opposition goal scoring moves, it looks like we need a sweeper to cover moves into the 18 yard box.  The figures don't reflect the number of times players are caught out of position resulting in a goal which I suspect might be a contributing factor, combine this with the "40% of through balls" fact.  A sweeper might help this too.  Our "defensive dual" success seems low for everybody so someone sitting deep would be able to double up on tackles would help this although at the expense of a deeper back line, Meekings would have the pace to track through balls more then any other options.

 

Suggest;

                          OFW

                       Meekings                        

Raven    McKay    Warren   Tremarco

                           Draper

              Doran   Tansey  King

                        Doumbouya

I think a lot of the problems aren't restricted to the defence - if it was one player who had got significantly worse then it'd be easy to pinpoint but all defenders performing at a lower level, in the measures used, suggests that the issues are wider.

I didn't do any analysis of other players further up the pitch but the fact that this season more of the moves leading to goals against us have been in the 'centre' circle area stuck out for me.  I don't think we are keeping the ball up front well enough and when we lose it I'm not sure we are doing the best job of recovering it, or even pressuring the opposition and covering space.  I can't say anything definitive about it though.

Regarding crosses and through balls, those figures haven't really changed much from last season to this.  40% of goals from through balls, 20% from crosses from the left, 20% from crosses from the right and the rest from 'other' (long shots, dead balls, OGs, direct passes to the opposition).  

One thing that would've been useful is to have had the figures for the entire league to see how we compare, especially with the pitch grid stuff.  I'm not doing that though because I think my wife would leave me if I spent hours watching all Hamilton Accies goals.

I don't really like that formation - I don't think adding another defender would deal with the issues and I can see that midfield getting put under a lot of pressure.  I'd have Polworth starting as well.