Do you have the possibility to determine how many of those low probability shots, below 0.17, could also have been an assist looking for a deflection or rebound opportunity?
If I understand correctly you are talking about the sort of ball that is played into the area in front of goal for, as you say, a deflection or maybe forcing the goalkeeper to come and intercept which might result in a rebound? If I have that right I can say the criteria for a shot is quite strict. The ball for a deflection is not counted as a shot nor is a ball that forces the ‘keeper to come and intercept - they are categorised as passing actions. It is obviously difficult to tell what the players intent was sometimes and there can be some ambiguity when a player plays a hard ball across the face of goal (was it a shot or a pass for deflection - did the player even know?!) but in general those attempts are defined by where they ended up. If it made the goalkeeper save (rather than intercept) it was a shot, if it was available for a deflection it was a pass attempt.
That said I think it is interesting to see if these lower value shots cause other opportunities (a second shot, a PC or a PS perhaps) just from the defensive mayhem the initial shot creates. We get into conditional probability here (the value of a shot, followed by the value of a shot from the rebound for example) but it is worth looking at to see how often a second positive action may occur from an initial (low value) shot. I hope I can put something together on this in time for the talk next month😁!
Exactly... I want to give strikers a bit of credit here :) And think at least some of the low value shots were taken with the idea: "if it goes in... great, but at least there is a good chance for a tip-in or a rebound"
A low vaue shot taken without a team mate in front of the goals for a tip-in or a rebound remains a low value shot. But one where that team mate is there maybe deserves better than the low value opportunity it's getting now ;)
In some ways I am surprised we don't know this already. I mean, the data isn't difficult to collect - what happened after the initial shot was taken? - but it doesn't seem to be something hockey has characterised in any meaningful way. I am doing the analysis now and it will be interesting to see if there really is, as you say "...a good chance of a tip-in or rebound".
Hi Simon. Thank you for the great insights. I am a dad of a fairly young striker and we are always looking at how she can improve her goal scoring as well as her positioning in the D (the sweet spot seems to be from the p flick spot all the way to just in front of the keeper?). Just two quick questions: 1. What is the goal conversion rate for a forward under the age of 18 (in general)-shots taken versus goals scored? 2. From the data above is it possible to see how many goals were scored from cutbacks? Kind regards. Jeremy (South Africa).
Hey Jeremy, thanks for the comment. The easy, but I suspect somewhat unsatisfactory answer is that I don’t know what the conversion rates for junior players would be. Though I do coach U18s I haven’t collected this data as I have been focused on senior women’s hockey. The modelling approach does scale quite well across senior hockey levels but I am not sure how good it would be a predicting junior hockey probabilities. It may well be that the juniors need their own specific model and that is quite a lot of work to put together!
Yes it is possible to narrow it down to outcomes from cutbacks and in fact one of the things on my long list is to look at in-circle pass direction and subsequent scoring probability by the player receiving the pass. So that would include cutbacks and passes from a range of angle. Something to do when the season finishes!
Sorry I couldn’t give you more precise answers but this, along with modelling what goes on between the circles, is very much work in progress so there is a lot to work on still.
Hi Simon. Greatly appreciate the response and the work that you do. I suppose we focus on the senior players and try and beat their stats or at least use them as a benchmark. In South Africa we are neglecting our female strikers/forwards at a junior level in terms of increasing their goal scoring abilities as field goals and the build up thereof are a problem (an eye test confirms this). The in circle pass direction with XG will be quite interesting when you do have time 🙏Have a great weekend and maybe one day you will see my girls on an international stage. All the best👌🏽
Do you have the possibility to determine how many of those low probability shots, below 0.17, could also have been an assist looking for a deflection or rebound opportunity?
Hi Ernst,
If I understand correctly you are talking about the sort of ball that is played into the area in front of goal for, as you say, a deflection or maybe forcing the goalkeeper to come and intercept which might result in a rebound? If I have that right I can say the criteria for a shot is quite strict. The ball for a deflection is not counted as a shot nor is a ball that forces the ‘keeper to come and intercept - they are categorised as passing actions. It is obviously difficult to tell what the players intent was sometimes and there can be some ambiguity when a player plays a hard ball across the face of goal (was it a shot or a pass for deflection - did the player even know?!) but in general those attempts are defined by where they ended up. If it made the goalkeeper save (rather than intercept) it was a shot, if it was available for a deflection it was a pass attempt.
That said I think it is interesting to see if these lower value shots cause other opportunities (a second shot, a PC or a PS perhaps) just from the defensive mayhem the initial shot creates. We get into conditional probability here (the value of a shot, followed by the value of a shot from the rebound for example) but it is worth looking at to see how often a second positive action may occur from an initial (low value) shot. I hope I can put something together on this in time for the talk next month😁!
Exactly... I want to give strikers a bit of credit here :) And think at least some of the low value shots were taken with the idea: "if it goes in... great, but at least there is a good chance for a tip-in or a rebound"
A low vaue shot taken without a team mate in front of the goals for a tip-in or a rebound remains a low value shot. But one where that team mate is there maybe deserves better than the low value opportunity it's getting now ;)
In some ways I am surprised we don't know this already. I mean, the data isn't difficult to collect - what happened after the initial shot was taken? - but it doesn't seem to be something hockey has characterised in any meaningful way. I am doing the analysis now and it will be interesting to see if there really is, as you say "...a good chance of a tip-in or rebound".
Hi Simon. Thank you for the great insights. I am a dad of a fairly young striker and we are always looking at how she can improve her goal scoring as well as her positioning in the D (the sweet spot seems to be from the p flick spot all the way to just in front of the keeper?). Just two quick questions: 1. What is the goal conversion rate for a forward under the age of 18 (in general)-shots taken versus goals scored? 2. From the data above is it possible to see how many goals were scored from cutbacks? Kind regards. Jeremy (South Africa).
Hey Jeremy, thanks for the comment. The easy, but I suspect somewhat unsatisfactory answer is that I don’t know what the conversion rates for junior players would be. Though I do coach U18s I haven’t collected this data as I have been focused on senior women’s hockey. The modelling approach does scale quite well across senior hockey levels but I am not sure how good it would be a predicting junior hockey probabilities. It may well be that the juniors need their own specific model and that is quite a lot of work to put together!
Yes it is possible to narrow it down to outcomes from cutbacks and in fact one of the things on my long list is to look at in-circle pass direction and subsequent scoring probability by the player receiving the pass. So that would include cutbacks and passes from a range of angle. Something to do when the season finishes!
Sorry I couldn’t give you more precise answers but this, along with modelling what goes on between the circles, is very much work in progress so there is a lot to work on still.
Hi Simon. Greatly appreciate the response and the work that you do. I suppose we focus on the senior players and try and beat their stats or at least use them as a benchmark. In South Africa we are neglecting our female strikers/forwards at a junior level in terms of increasing their goal scoring abilities as field goals and the build up thereof are a problem (an eye test confirms this). The in circle pass direction with XG will be quite interesting when you do have time 🙏Have a great weekend and maybe one day you will see my girls on an international stage. All the best👌🏽