Mid-season report

[PerfICT]

We're just past the halfway stage of the season, so what are our thoughts on the shape of it so far? Here are some statistics to be bored by or to chew over to help the discussion along:

The league stats show a certain amount of symmetry:

P21 W7 D7 L7: 28 points. Goals For: 30; Goals Against 31

Given our terrible start, I'm delighted with the current league position and am fairly optimistic for the remainder of the season, if we can keep everybody fit. We could, however, do with Meekings and Doran back!

Some other notable statistics - not many surprises!

We have the lowest number of shots/game (8.56), the 3^rd lowest shots on target (3.48), but the 3^rd best conversion rate (goals/shots on target; 39.73%). This might reflect our reluctance to shoot unless it is a clear-cut opportunity!

We also have the lowest corner count (won) which might be linked to the lack of shots on goal.

Our average home crowd is 3,959 (260 better than the average for whole of last season) - 10^th in league above Hamilton and St J and within 100 of PT and Killie. This suggests that our attendance is holding steady, but sadly there isn't much evidence of a bounce from last year's heroics.

Our average possession is just less than 50% (49.9%) and there is a very clear inverse relationship between goals scored and possession in each game: for every 10% more possession we have, we score ~1 less goal. In 5/7 of the games we have won, we have had less possession than the opposition and in 5/7 of those that we have lost, we have had more possession.

This might sound crazy and could just be a trick of the numbers, but I don’t think so on this occasion. Perhaps it indicates that we really struggle to break down organised teams that allow us to play in front of them (St Johnstone at home and Stirling Albion in the cup are good examples). Possession, particularly in the middle or defensive third does not win games and can often contribute to losing them.

Cups:

Going out of any cup is disappointing, no more so than when it is at the hands of County. Likewise, our performance in the Scottish Cup was not great against Stirling Albion, but at least we have a chance to put that record straight next week. It would be great to have another cup run and I see no reason why we can't this season.

Good buys, goodbyes and star performers:

Clearly, the best signing of the season was Miles Storey and it is fantastic that he's with us until the end of the season. Some of the other signings have not been so great - Lopez in particular. The most improved player has been Liam Polworth by a distance. It is also great to have Gary Warren back, but it is a real shame that we haven't yet managed to reunite Warren and Meekings. Draper, Tansey and Vincent are starting to show the consistency that we're used to from them. It was sad to say goodbye to Ryan Christie who was a star for much of the early season - I wish him well for the future.

Thoughts?

[+DoofersDad]

Very interesting!  The most interesting thing must be the inverse relation between possession and goals scored, together with the number of shots and corners per game.  On the face of it, this would seem to reflect what many of us have been saying about the emphasis on possession that the manager has focussed on.  The fact is that trying to play the ball out of defence whenever possible and patiently retaining possession in midfield until some opportunity emerges, simply allows the opposition to get players behind the ball and makes it harder to create chances.  It can also be very boring to watch. The other interesting thing is that despite the emphasis on possession football, the stats would suggest we're not actually that good at keeping the ball.

The lesson from all this would appear to be a need to move the ball out of defence quicker and to take a few more chances with possession further up the field in an attempt to create chances.  In other words we need to look to play a few more through balls and we need to have players running at defenders more.  If we do that we will inevitably lose possession pretty often, but if we lose possession in our opponent's half, it is far less dangerous than messing about at the back and losing possession in our own half. And if we take the risks and we don't lose possession, well, that is when we create the goal scoring chances.

To be fair, it is now beginning to look as though the lessons are being learnt.  We are unbeaten in the last 5 league games but have conceded 8 goals in that period (despite not having played Celtic!) At the start of the season, we only conceded 5 goals in the first 5 games against opposition other than Celtic, but failed to win. However, we have scored 12 goals in the last 5 games compared to only 2 in the earlier games.  Significantly, we averaged 52% possession in those 5 early games but this has dropped to a shade under 50% in our recent unbeaten run.  Clearly it is not keeping the ball that matters, it is what you do with it that counts.

 

[Charles Bannerman]

6 minutes ago, DoofersDad said:

  Significantly, we averaged 52% possession in those 5 early games but this has dropped to a shade under 50% in our recent unbeaten run.  

Are the possession stats and the means of measuring them accurate enough for that magnitude of difference to be significant? You sometimes see some really bizarre ones.

[PerfICT]

14 minutes ago, Charles Bannerman said:

Are the possession stats and the means of measuring them accurate enough for that magnitude of difference to be significant? You sometimes see some really bizarre ones.

Totally agree. The possession stats can be bewildering sometimes, so I'd take them with a pinch of salt.

Another stat I meant to include - approximately half our goals have been conceded in the last 10 minutes or in injury time. This has not escaped the notice of the supporters, players or management. Knowing the problem is one thing, identifying the cause and rectifying it is quite another!

[PerfICT]

For anybody who is interested, here's the relationship between goals and possession. Statisticians will get all upset about the fact that "goals" is discontinuous data and therefore doesn't suit this kind of analysis......but I still think it shows something interesting! I wasn't expecting any kind of pattern from such a small data set. Red symbols are losses, green are wins and orange are draws.

 

[AlexJones]

[+DoofersDad]

1 minute ago, Charles Bannerman said:

Are the possession stats and the means of measuring them accurate enough for that magnitude of difference to be significant? You sometimes see some really bizarre ones.

Not in themselves in a strictly statistical sense, but in the context of other stats I think it would be a mistake to ignore the implications.  Bottom line is that if you want to score goals, you have to risk losing possession.  

[Yngwie]

Some fascinating stats on this thread.  I'm sure Sneckboy will be keen to comment, just as soon as he's finished wiping down his keyboard and monitor!

[Charles Bannerman]

43 minutes ago, DoofersDad said:

Not in themselves in a strictly statistical sense, but in the context of other stats I think it would be a mistake to ignore the implications.  Bottom line is that if you want to score goals, you have to risk losing possession.  

Agreed, but given the inaccuracies in possession stats, is the difference between 52% and "a shade under 50%" significant enough to support a distinction between the first five games and the rest?

The graph is very revealing indeed though, as a general indication and in this case, the trend it shows seems quite robust enough to survive the use of a discontinuous variable.

Edited January 13, 2016 by Charles Bannerman

[+DoofersDad]

Another interesting statistic is the results when Ryan Christie has been in the team compared to when he hasn't.  In the 13 games he was involved in we got 14 points and scored 13 goals.  In fact, 2  of those goals were scored after he went off injured in the Motherwell game.  In the 8 league games he has not been involved in we have also got 14 points and have scored 16 goals.  Are we a better team without him in the side?

[Yngwie]

More likely because Christie's games were early in the season when our injury list was something of a crisis.  His appearances also largely coincided with Lopez playing upfront instead of Storey.

[IMMORTAL HOWDEN ENDER]

Suppose this means from a statistical perspective that we have been shitier than usual this years

[Robert]

This season was always going to be tough after the last three years, and the achievements of last year in particular.

The recent good run has helped, but stalled last weekend and today. 

Hopefully Tuesday's replay goes to plan, but we then have a tough run in the league: Partick, Hearts, Aberdeen, Celtic, Dundee and Hearts (first three at home, next three away), so the next couple of months could be traumatic!

Fingers crossed we keep keep close to or in the top six and well away from the play off place dog fight, but it could be tough!

[Yngwie]

Here are some more stats to add into the mix, inspired by this comment from the Partick thread.

5 hours ago, PerfICT said:

Ironically, I think we're more likely to get a result against Hearts and Aberdeen because they'll have more of a go and force us to be more counter-attacking.

Here are our points out of the possible achievable against each team so far:

Celtic 0/6

Aberdeen 4/6

Hearts 3/3

County 6/6 

St J 1/6

Dundee 2/6

Partick 2/9

Motherwell 3/6

Hamilton 3/6

Killie 3/9

D Utd 2/6

Now, splitting this into performance v Top6 and Bottom 6:

Top 6:  14/27, achieving a very credible 52% of points available

Bottom 6: 15/42, a mere 36%.

This would appear pretty consistent with what PerfICT shows earlier in this thread, whereby the more possession we have, the fewer goals we score and the less likely we are to win.

It also reiterates the importance of getting into the top 6, because given our apparent inability to beat the poorer sides, 5 post-split games against those sides and you can see how we might easily slip into serious relegation contention - unless something changes or we adapt our tactics against weaker sides. Hmm.

 

Edited January 24, 2016 by Yngwie

[+DoofersDad]

Surely this evidence tends to suggest that we do have the players to match the better teams in the league, but are tactically inept against the poorer sides.

[bdu98196]

I think I agree, we seem to be able to sit in and keep possession deep letting team play infront of us, but when we have to step up against teams who sit in we lack a cutting edge or creativity to open them up. When the 'better' teams come to us they try to go for the 3 points where as the 'poorer' teams tend to be more defensive and dont want to loose - with our lack of width, no overlapping runs and slow approach play its easier for them to stiffle us and hit us on the break, where thats what we tend to do against attacking teams.

 

[rocky1]

1 hour ago, bdu98196 said:

I think I agree, we seem to be able to sit in and keep possession deep letting team play infront of us, but when we have to step up against teams who sit in we lack a cutting edge or creativity to open them up. When the 'better' teams come to us they try to go for the 3 points where as the 'poorer' teams tend to be more defensive and dont want to loose - with our lack of width, no overlapping runs and slow approach play its easier for them to stiffle us and hit us on the break, where thats what we tend to do against attacking teams.

 

Pretty well sums it up so Mr Hughes the solution is? 

[bdu98196]

Missing creative players like Doran & Roberts through injury hasnt helped added to we lost Watkins in the summer and now Christie is gone too. Roberts may come good but he needs to get fit, I do think we have Williams who is a great attacking threat showm last year when he played on the left and got forward, playing balls in the box and going past players yet somehow Yogi see his best postion at LB, when we have a capable player in Tremarco who can play there. Unfortunately injuries to Mutumbo and Sho-silva never helped as these guys never got going?

Unfortunately unless we add width through Willaims (L) and Doran/Roberts (R) we will continue to struggle to create. Through the middle we have too many defensive minded players so nobody to unlock the door there either.

[+HawkeyeTheGnu]

6 hours ago, DoofersDad said:

Surely this evidence tends to suggest that we do have the players to match the better teams in the league, but are tactically inept against the poorer sides.

I think that the "better" teams try to play football from the off, which suits our style of play. Against the other teams we have to earn the right to play football, and we are not capable of doing that. Also, the "poorer" sides are more defensive, and we lack the movement or width to get through them.

Having said all that, that is 2 more points from Partick than we managed last season!

[PerfICT]

It is fair for Yogi to bemoan the loss of players like roberts and Doran to injury and Christie to Celtic. HOWEVER, the complaint only hold water if, when these players are available, he plays them! Doran rarely got a start under JH and Roberts is keeping the bench warm - is it just that he doesn't actually value width and creativity? Further fuel to this fire was his public denouncement of Mutombo's "showboating": one man's showboating is another's creativity and attempt to actually beat an oppentn with skill. I'm no supporter of flash for the sake of it, but I crave seeing an ICT player skinning a defender with skill and/or pace this season - the message seems to be that these tactics are banned in preference for passing a team to death??