Hi Monte
You are certainly right when you say that everything can be found in the SD documentation.
But the problem is about the volume of the documentation.
In a specific case, where in the documentation must one refer?
I think that there is a difference between applying concretely a method and just reading cases. Cases are specific as is the case you want to solve, and one solution good in one case can be wrong in the other one.
For example, you see very rarely, eventually never, a concrete provisional study of the expected cost in money, effort, time of the modelling effort, nor the expected benefit, even if these evaluations are between a confidence interval and a relatively low probability.
You see extremely rarely when somebody is reporting a modelling application, what the model costs and what was the benefit estimated by the 'CLIENT'.
Just browse the papers at the last SDC, and tell me if you find. papers working this way, with a provisional evaluation and with the real data after the work is finished.
Or this is what any reasonable business manager will ask prior to engaging into that sort of work.
Other organisations, may not think this way being less concerned with costs and benefits and being sensible to other considerations.
I do not understand your last paragraph that seems contradictory.
I have read quite a lot about revenue management a field that deals with how Airlines company fix their rates, but too hotels, hospitals, cruise company, and about all service companies, and too the distribution sector.
I read too a lot about science marketing.
These two fields, generally always deal with dynamic problems.
In the more recent book about revenue management, that is considered as the bible, there is no mention of SD, even not in the books referenced. The technique used is mainly O.R., probability and statistics, mathematics and that’s all.
Same for books about marketing research.
I wanted to explain that you can work with many tools to solve one problem, and that the difference is not if one method can or cannot do the job, but what method will dot it with the less cost and will deliver solutions more efficient.
Regards.
JJ
[Edited on 22-9-2007 by LAUJJL]
why not build a model of the growth of the field
...not more people attracted by question?
Hello,
it seems that at the present time just about 4-5 people are contributing to Jean-Jacques' question of "Modeling the Growth of SD" especially under the view of why it is not spreading deeper into companies and the business world.
From my personal experience I can tell you that the mental model of managers concerning SD is that this has solely to do with complex models where it needs lots of equations and numbers that are not easily found in companies.
Too bad -in my eyes- as SD starts much earlier in the process and the model would be -in a way- the final point of examination and decision making.
Why aren't the experts taking part of the discussion yet?
Looking forward to seeing more input on a really essential question that everybody of us can experience oneself.
Best regards
Ralf
it seems that at the present time just about 4-5 people are contributing to Jean-Jacques' question of "Modeling the Growth of SD" especially under the view of why it is not spreading deeper into companies and the business world.
From my personal experience I can tell you that the mental model of managers concerning SD is that this has solely to do with complex models where it needs lots of equations and numbers that are not easily found in companies.
Too bad -in my eyes- as SD starts much earlier in the process and the model would be -in a way- the final point of examination and decision making.
Why aren't the experts taking part of the discussion yet?
Looking forward to seeing more input on a really essential question that everybody of us can experience oneself.
Best regards
Ralf
Ralf Lippold
SD Growth
Hi Ralf
Changing little things can have big impact.
Yes but you often see it only when things have happened.
About modelling the Sd growth, I just think that I am absolutely not able to bring something concrete to that problem. I mentioned that question because I thought that it was the job of the system dynamics association to do that with whatever information they had or could gather in the future.
The only thing that I can say, from my point of view necessarily biaised is:
For 50 Years the SD growth was week.
Why?
1.
I think that Bob mentioning in the sd mailing list, that SD
was too scientific to be used by managers is right.
One could argue that businesses hire a lot of scientists and using science is not problem for them.
But the distinction is the subject of investigation.
As long as the study is dealing with technical matters, it is ok, but it the subject is strategic, science is in competition with the way traditional managers think that is not scientific.
So managers will not accept that somebody else do their job, using a method that they do not master.
And to use such a method, it should be widely recognized to be a very efficient one, which seems not to be the case.
2. There is probably a strong competition from other methods, using static methods to solve dynamic problems.
For instance using equations with the time included in them,
the tradtional 't' symbol.
To what extent is SD way of portraying the effect of time, better than these traditional methods, would be interesting to know.
3. What is the real number of problems where there is sufficient dynamic? probably a lot, but how much compared to other more static problems and if there are many problems are there enough competent modellers to solve them? It is a bit like the chicken egg problem.
Are there no competent modellers because there is no market, or is there no market because there are no competent modellers.
About that problem Kim Warren mentions at the end of his first book to the reader that has gone so far, that after his big effort, he is not able to build a serious model, and
that finding an expert to do it, will be very difficult, the competent modellers being very few in the whole world!
4. One question that seems quite evident for me, is that SD is not enough practical oriented.
I know only one book that trains the reader to build models.
Because to be a good modeller, one needs training and applied knowledge.
For what reasons is it so academic?
Maybe because it is intellectually very gratifying and has been monopolised by academics, or because people competent at modelling have no interest in divulging their
own methods, leaving the field of learning to academics?
I have no answer to this question, but there may be some other possibilities.
Regards.
JJ
[Edited on 8-10-2007 by LAUJJL]
Changing little things can have big impact.
Yes but you often see it only when things have happened.
About modelling the Sd growth, I just think that I am absolutely not able to bring something concrete to that problem. I mentioned that question because I thought that it was the job of the system dynamics association to do that with whatever information they had or could gather in the future.
The only thing that I can say, from my point of view necessarily biaised is:
For 50 Years the SD growth was week.
Why?
1.
I think that Bob mentioning in the sd mailing list, that SD
was too scientific to be used by managers is right.
One could argue that businesses hire a lot of scientists and using science is not problem for them.
But the distinction is the subject of investigation.
As long as the study is dealing with technical matters, it is ok, but it the subject is strategic, science is in competition with the way traditional managers think that is not scientific.
So managers will not accept that somebody else do their job, using a method that they do not master.
And to use such a method, it should be widely recognized to be a very efficient one, which seems not to be the case.
2. There is probably a strong competition from other methods, using static methods to solve dynamic problems.
For instance using equations with the time included in them,
the tradtional 't' symbol.
To what extent is SD way of portraying the effect of time, better than these traditional methods, would be interesting to know.
3. What is the real number of problems where there is sufficient dynamic? probably a lot, but how much compared to other more static problems and if there are many problems are there enough competent modellers to solve them? It is a bit like the chicken egg problem.
Are there no competent modellers because there is no market, or is there no market because there are no competent modellers.
About that problem Kim Warren mentions at the end of his first book to the reader that has gone so far, that after his big effort, he is not able to build a serious model, and
that finding an expert to do it, will be very difficult, the competent modellers being very few in the whole world!
4. One question that seems quite evident for me, is that SD is not enough practical oriented.
I know only one book that trains the reader to build models.
Because to be a good modeller, one needs training and applied knowledge.
For what reasons is it so academic?
Maybe because it is intellectually very gratifying and has been monopolised by academics, or because people competent at modelling have no interest in divulging their
own methods, leaving the field of learning to academics?
I have no answer to this question, but there may be some other possibilities.
Regards.
JJ
[Edited on 8-10-2007 by LAUJJL]
Hello everybody,
as the discussion has come to a stand still I would like to make it run again.
I just got an interesting book which is about "Kaizen - the Art of Creative Thinking" by Shigeo Shingo (one of the "architects" of the Toyota Production System (TPS) and as I started to read it stuck who much SD thinking there was.
Kaizen is by the way -for the ones of you not familiar with TPS- the incremental improvement of processes in a given system (such as an organization). That makes it possible to change things over time (sounds familiar again;-)) without great disturbance of the other parts of the system.
He talks a lot about mental models (even though he doesn't call it that) and how they influence decisions.
So perhaps instead of looking what could be done to grow the acceptance of SD outside the purely SD orientated academic world and a few fields where the benefits have been already recognized we should focus on what mental models (our own) are hindering us to bring a better understanding of SD to public?
A starting point would be to make people curious about their own assumptions (such as "SD is complicated and expensive simulation stuff that takes forever!") and connect to their emotions and there personal experiences.
Right on the beginning of the above mentioned book there is the question on how many squares can you see (there is a layout given with 16 equal squares:
1. 16 = the obvious solution is 16 (just the small squares)
2. 17 = 1. plus the surrounding (the 16 small one) square
3. 30 = 1. plus 2. plus 4 9-block-squares and 9 4-block squares
As one can see that makes a huge difference and still everything is the same.
So what are we all missing? Is there something in front of us that is not yet seen by us?
Jean-Jacques has pointed out some great shortcomings that I realized myself as well.
Before the modeling stands the understanding - that seems to be the present task. Myself I try to bring the systems thinking view into the open in various discussions and it is interesting how people sometimes tend to hook on the presented ideas even though they don't know anything about system dynamics. It is great that one can have an impact with the "small" but nevertheless effective changes:-)
Best regards and looking forward to your comments and ideas. It will certainly enrich the "field" and the connecting points to other topics.
Ralf
as the discussion has come to a stand still I would like to make it run again.
I just got an interesting book which is about "Kaizen - the Art of Creative Thinking" by Shigeo Shingo (one of the "architects" of the Toyota Production System (TPS) and as I started to read it stuck who much SD thinking there was.
Kaizen is by the way -for the ones of you not familiar with TPS- the incremental improvement of processes in a given system (such as an organization). That makes it possible to change things over time (sounds familiar again;-)) without great disturbance of the other parts of the system.
He talks a lot about mental models (even though he doesn't call it that) and how they influence decisions.
So perhaps instead of looking what could be done to grow the acceptance of SD outside the purely SD orientated academic world and a few fields where the benefits have been already recognized we should focus on what mental models (our own) are hindering us to bring a better understanding of SD to public?
A starting point would be to make people curious about their own assumptions (such as "SD is complicated and expensive simulation stuff that takes forever!") and connect to their emotions and there personal experiences.
Right on the beginning of the above mentioned book there is the question on how many squares can you see (there is a layout given with 16 equal squares:
1. 16 = the obvious solution is 16 (just the small squares)
2. 17 = 1. plus the surrounding (the 16 small one) square
3. 30 = 1. plus 2. plus 4 9-block-squares and 9 4-block squares
As one can see that makes a huge difference and still everything is the same.
So what are we all missing? Is there something in front of us that is not yet seen by us?
Ralf's commentOriginally posted by LAUJJL
For 50 Years the SD growth was week.
Why?
1.
.....
So managers will not accept that somebody else do their job, using a method that they do not master.
And to use such a method, it should be widely recognized to be a very efficient one, which seems not to be the case.
Ralf's comment: Of course it seems that managers reject the "high class" of SD with simulation as they can't understand it and it seems to produce no instant benefits.
From what I have learned so far, SD starts way earlier with the unbreaking of mental models through either easy simulations as the beer game (where everybody can see what is going on underneath the surface of current beliefing) or asking questions about what could be the causes of problems (to catch people in there known territory where they feel familiar) in recent settings.
2. There is probably a strong competition from other methods, using static methods to solve dynamic problems.
For instance using equations with the time included in them,
the tradtional 't' symbol.
To what extent is SD way of portraying the effect of time, better than these traditional methods, would be interesting to know.
Ralf's comment: causal loops can be a good way to open up minds, even though it is difficult to make people open for discussions like that (especially managers are focused on other things and tend to believe that they have no time to being taught this kind of stuff).
3. What is the real number of problems where there is sufficient dynamic? probably a lot, but how much compared to other more static problems and if there are many problems are there enough competent modellers to solve them?
Ralf's comment: What kind of static problems are you thinking of? Isn't everything dynamic as time goes by, even though the human mind can't capture the changes?
It is a bit like the chicken egg problem.
Are there no competent modellers because there is no market, or is there no market because there are no competent modellers.
About that problem Kim Warren mentions at the end of his first book to the reader that has gone so far, that after his big effort, he is not able to build a serious model, and
that finding an expert to do it, will be very difficult, the competent modellers being very few in the whole world!
Ralf's comment: That makes me feel like trying to learn more and more about SD in order to gain the necessary insights to either beccome a modeler or stick together with other SD learners to get to higher points of knowledge (what I am currently doing).
4. One question that seems quite evident for me, is that SD is not enough practical oriented.
Ralf's comment: From first sight that can be really true and as managers tend to see the "big picture" there is a tendancy to look at the end product (the end product of SD is for sure a good simulation). As they don't understand what is really going on in the functions and the connections of the model they think, "That is superficial and not practical for our use. So let's not use it."
I know only one book that trains the reader to build models.
Because to be a good modeller, one needs training and applied knowledge.
For what reasons is it so academic?
Maybe because it is intellectually very gratifying and has been monopolised by academics, or because people competent at modelling have no interest in divulging their
own methods, leaving the field of learning to academics?
Ralf's comment: So let's connect academics and practioners. Both have a similar interest in the field and have different points of view which makes learning a great opportunity in order to forward the field for the next 50 years (and beyond;-))
[Edited on 8-10-2007 by LAUJJL]
Jean-Jacques has pointed out some great shortcomings that I realized myself as well.
Before the modeling stands the understanding - that seems to be the present task. Myself I try to bring the systems thinking view into the open in various discussions and it is interesting how people sometimes tend to hook on the presented ideas even though they don't know anything about system dynamics. It is great that one can have an impact with the "small" but nevertheless effective changes:-)
Best regards and looking forward to your comments and ideas. It will certainly enrich the "field" and the connecting points to other topics.
Ralf
Ralf Lippold
Sd growth
Hi Ralf
<as the discussion has come to a stand still I would like to make it run again.
JJs comment:
I personally think that I have not the information to give a valuable idea about the problem of the development of SD.
There have been many trends on that subject on the SD mailing list.
There are multiple point of views that lead to different and often opposite ideas.
One should first define what is meant by SD.
There are to my opinion different kinds of SD.
You have the theoretical SD, and you have the practical SD.
There are too different kinds of practical SD.
For me a model built for the public sector has less chance to be practical than one built for the
private one. The public sector is less adamant about cost and utility than the private one.
For me the smaller the private customer the greater the chance it has to be useful, because smaller
businesses have no budget to spend on research.
I am interested in practical SD, the one that delivers verifiable results with verifiable costs.
From what I have seen so far, that last SD is marginal in books I have read, courses I studied, and forums
I frequent.
One must then before studying SD, precise what SD you are talking about.
<I just got an interesting book which is about "Kaizen - the Art of Creative Thinking" by Shigeo Shingo (one of the <"architects" of the Toyota Production System (TPS) and as I started to read it stuck who much SD thinking there <was
<Kaizen is by the way -for the ones of you not familiar with TPS- the incremental improvement of processes in a given <system (such as an organization). That makes it possible to change things over time (sounds familiar again;-)) <without great disturbance of the other parts of the system.
<He talks a lot about mental models (even though he doesn't call it that) and how they influence decisions.
JJS comment
For me Kaizen was introduced by USA after the war to help the Japanese to recover from it.
The exceptional results come from the mentality of Japanese where the individual is totally at the service of the
community.
Sd is not process oriented and could be if it was more practical oriented.
For 50 Years the SD growth was week.
Why?
1.
.....
<So managers will not accept that somebody else do their job, using a method that they do not master.
<And to use such a method, it should be widely recognized to be a very efficient one, which seems not to be the case.
<Ralf's comment: Of course it seems that managers reject the "high class" of SD with simulation as they can't <understand it and it seems to produce no instant benefits.
<From what I have learned so far, SD starts way earlier with the unbreaking of mental models through either easy <simulations as the beer game (where everybody can see what is going on underneath the surface of current beliefing) <or asking questions about what could be the causes of problems (to catch people in there known territory where they <feel familiar) in recent settings.
JJS comment
I am after 6 years of SD, not yet clear about where and how to use SD and if it adds a real value.
2. There is probably a strong competition from other methods, using static methods to solve dynamic problems.
For instance using equations with the time included in them,
the tradtional 't' symbol.
To what extent is SD way of portraying the effect of time, better than these traditional methods, would be interesting to know.
Ralf's comment: causal loops can be a good way to open up minds, even though it is difficult to make people open for discussions like that (especially managers are focused on other things and tend to believe that they have no time to being taught this kind of stuff).
JJs comment.
Managers concentrate their time on anything they find useful. If they do not want to be taught this kind of stuff, it is because they have no examples close to them that would prove the utility of doing it.
3. What is the real number of problems where there is sufficient dynamic? probably a lot, but how much compared to other more static problems and if there are many problems are there enough competent modellers to solve them?
Ralf's comment: What kind of static problems are you thinking of? Isn't everything dynamic as time goes by, even though the human mind can't capture the changes?
JJ’S comment.
It it true that everything is dynamic, but it is most of the time, simpler and more useful to consider it static.
It is a bit like the chicken egg problem.
Are there no competent modellers because there is no market, or is there no market because there are no competent modellers.
About that problem Kim Warren mentions at the end of his first book to the reader that has gone so far, that after his big effort, he is not able to build a serious model, and
that finding an expert to do it, will be very difficult, the competent modellers being very few in the whole world!
Ralf's comment: That makes me feel like trying to learn more and more about SD in order to gain the necessary insights to either beccome a modeler or stick together with other SD learners to get to higher points of knowledge (what I am currently doing).
JJs comment.
I try to learn more about using SD practically.
To do that, I try first to study any problem using only Word and a simple calculation machine.
It helps me stay close to the problem.
One does not imagine what one can do with these simple tools and a lot of thinking.
I add more sophisticated tools only if I can see the added value of doing it, and trying to add these sophisticated tools in light touch, always staying close to the problem.
I will not jump to SD first, and will first eventually work with Excel.
I can then see the real added values of any tool and its relation to the original problem.
I think that if SD was taught this way, it would ease the spread of the field, as people would see the read added value
of the method.
When one studies SD books, everything is always related to SD, as if it was a panacea.
Computation is never a substitute to thinking and close observation of reality.
4. One question that seems quite evident for me, is that SD is not enough practical oriented.
Ralf's comment: From first sight that can be really true and as managers tend to see the "big picture" there is a tendancy to look at the end product (the end product of SD is for sure a good simulation). As they don't understand what is really going on in the functions and the connections of the model they think, "That is superficial and not practical for our use. So let's not use it."
I know only one book that trains the reader to build models.
Because to be a good modeller, one needs training and applied knowledge.
For what reasons is it so academic?
Maybe because it is intellectually very gratifying and has been monopolised by academics, or because people competent at modelling have no interest in divulging their
own methods, leaving the field of learning to academics?
Ralf's comment: So let's connect academics and practioners. Both have a similar interest in the field and have different points of view which makes learning a great opportunity in order to forward the field for the next 50 years (and beyond;-))
[Edited on 8-10-2007 by LAUJJL] [/quote]Ralf's comment
Jean-Jacques has pointed out some great shortcomings that I realized myself as well.
Before the modeling stands the understanding - that seems to be the present task. Myself I try to bring the systems thinking view into the open in various discussions and it is interesting how people sometimes tend to hook on the presented ideas even though they don't know anything about system dynamics. It is great that one can have an impact with the "small" but nevertheless effective changes:-)
JJs comment.
You have completely expressed the resume of this discussion.
There is a minimum of understanding before one wants to model something, and there is absolutely no use of jumping to any method if one is not sure of the utility of doing it and if one does not have already a good knowledge of the problem, having already tried to use simple methods.
This is why I privilege a continuous method, starting with no method at all.
I will attend the next UK SDSIG conference in London, in February.
It is short enough, one day and a half.
It will give me the opportunity to see SD’ers. I have not yet seen any before.
In France, SD is completely unknown.
Regards.
JJ
<as the discussion has come to a stand still I would like to make it run again.
JJs comment:
I personally think that I have not the information to give a valuable idea about the problem of the development of SD.
There have been many trends on that subject on the SD mailing list.
There are multiple point of views that lead to different and often opposite ideas.
One should first define what is meant by SD.
There are to my opinion different kinds of SD.
You have the theoretical SD, and you have the practical SD.
There are too different kinds of practical SD.
For me a model built for the public sector has less chance to be practical than one built for the
private one. The public sector is less adamant about cost and utility than the private one.
For me the smaller the private customer the greater the chance it has to be useful, because smaller
businesses have no budget to spend on research.
I am interested in practical SD, the one that delivers verifiable results with verifiable costs.
From what I have seen so far, that last SD is marginal in books I have read, courses I studied, and forums
I frequent.
One must then before studying SD, precise what SD you are talking about.
<I just got an interesting book which is about "Kaizen - the Art of Creative Thinking" by Shigeo Shingo (one of the <"architects" of the Toyota Production System (TPS) and as I started to read it stuck who much SD thinking there <was
<Kaizen is by the way -for the ones of you not familiar with TPS- the incremental improvement of processes in a given <system (such as an organization). That makes it possible to change things over time (sounds familiar again;-)) <without great disturbance of the other parts of the system.
<He talks a lot about mental models (even though he doesn't call it that) and how they influence decisions.
JJS comment
For me Kaizen was introduced by USA after the war to help the Japanese to recover from it.
The exceptional results come from the mentality of Japanese where the individual is totally at the service of the
community.
Sd is not process oriented and could be if it was more practical oriented.
For 50 Years the SD growth was week.
Why?
1.
.....
<So managers will not accept that somebody else do their job, using a method that they do not master.
<And to use such a method, it should be widely recognized to be a very efficient one, which seems not to be the case.
<Ralf's comment: Of course it seems that managers reject the "high class" of SD with simulation as they can't <understand it and it seems to produce no instant benefits.
<From what I have learned so far, SD starts way earlier with the unbreaking of mental models through either easy <simulations as the beer game (where everybody can see what is going on underneath the surface of current beliefing) <or asking questions about what could be the causes of problems (to catch people in there known territory where they <feel familiar) in recent settings.
JJS comment
I am after 6 years of SD, not yet clear about where and how to use SD and if it adds a real value.
2. There is probably a strong competition from other methods, using static methods to solve dynamic problems.
For instance using equations with the time included in them,
the tradtional 't' symbol.
To what extent is SD way of portraying the effect of time, better than these traditional methods, would be interesting to know.
Ralf's comment: causal loops can be a good way to open up minds, even though it is difficult to make people open for discussions like that (especially managers are focused on other things and tend to believe that they have no time to being taught this kind of stuff).
JJs comment.
Managers concentrate their time on anything they find useful. If they do not want to be taught this kind of stuff, it is because they have no examples close to them that would prove the utility of doing it.
3. What is the real number of problems where there is sufficient dynamic? probably a lot, but how much compared to other more static problems and if there are many problems are there enough competent modellers to solve them?
Ralf's comment: What kind of static problems are you thinking of? Isn't everything dynamic as time goes by, even though the human mind can't capture the changes?
JJ’S comment.
It it true that everything is dynamic, but it is most of the time, simpler and more useful to consider it static.
It is a bit like the chicken egg problem.
Are there no competent modellers because there is no market, or is there no market because there are no competent modellers.
About that problem Kim Warren mentions at the end of his first book to the reader that has gone so far, that after his big effort, he is not able to build a serious model, and
that finding an expert to do it, will be very difficult, the competent modellers being very few in the whole world!
Ralf's comment: That makes me feel like trying to learn more and more about SD in order to gain the necessary insights to either beccome a modeler or stick together with other SD learners to get to higher points of knowledge (what I am currently doing).
JJs comment.
I try to learn more about using SD practically.
To do that, I try first to study any problem using only Word and a simple calculation machine.
It helps me stay close to the problem.
One does not imagine what one can do with these simple tools and a lot of thinking.
I add more sophisticated tools only if I can see the added value of doing it, and trying to add these sophisticated tools in light touch, always staying close to the problem.
I will not jump to SD first, and will first eventually work with Excel.
I can then see the real added values of any tool and its relation to the original problem.
I think that if SD was taught this way, it would ease the spread of the field, as people would see the read added value
of the method.
When one studies SD books, everything is always related to SD, as if it was a panacea.
Computation is never a substitute to thinking and close observation of reality.
4. One question that seems quite evident for me, is that SD is not enough practical oriented.
Ralf's comment: From first sight that can be really true and as managers tend to see the "big picture" there is a tendancy to look at the end product (the end product of SD is for sure a good simulation). As they don't understand what is really going on in the functions and the connections of the model they think, "That is superficial and not practical for our use. So let's not use it."
I know only one book that trains the reader to build models.
Because to be a good modeller, one needs training and applied knowledge.
For what reasons is it so academic?
Maybe because it is intellectually very gratifying and has been monopolised by academics, or because people competent at modelling have no interest in divulging their
own methods, leaving the field of learning to academics?
Ralf's comment: So let's connect academics and practioners. Both have a similar interest in the field and have different points of view which makes learning a great opportunity in order to forward the field for the next 50 years (and beyond;-))
[Edited on 8-10-2007 by LAUJJL] [/quote]Ralf's comment
Jean-Jacques has pointed out some great shortcomings that I realized myself as well.
Before the modeling stands the understanding - that seems to be the present task. Myself I try to bring the systems thinking view into the open in various discussions and it is interesting how people sometimes tend to hook on the presented ideas even though they don't know anything about system dynamics. It is great that one can have an impact with the "small" but nevertheless effective changes:-)
JJs comment.
You have completely expressed the resume of this discussion.
There is a minimum of understanding before one wants to model something, and there is absolutely no use of jumping to any method if one is not sure of the utility of doing it and if one does not have already a good knowledge of the problem, having already tried to use simple methods.
This is why I privilege a continuous method, starting with no method at all.
I will attend the next UK SDSIG conference in London, in February.
It is short enough, one day and a half.
It will give me the opportunity to see SD’ers. I have not yet seen any before.
In France, SD is completely unknown.
Regards.
JJ