About best practice

This idea of best practice continue to pop up. For me is disturbing, in the sense that some practices are being introduced just because near them appears the word “best”. And this is bloody dangerous. In the lines below, I try to decompress what I have just said because I realized that even though we speak about 2 words, is important to say what those 2 words might mean and all the thoughts regarding them.

As a premise, I have to say that language is very important because it can trigger different kind of thoughts( I thought is not true, but after reading some of Antonio Damasio work, I am sure now) . For example, I told to a Scrum Master that her role is like an attractor ( attractor idea from Lorenz attractor). I could have used a word which Scrum Masters are accustomed, but I might have failed to communicate what I wanted to say.

So, for me there are 2 things/directions to think about this topic, at least to try to clarify or see if is some coherence. First what some might really mean/think by “best practice” and second, see what exactly really is “best practice”.


1. what some really might mean by “best practice” -> to really see what might they mean, I imagine there can be lot of inference.

Here I identified 2 directions also, there might be others:

1.1 some people want to say something else, but don’t have the words.

For example: We want to use a new JS framework, like Vue/React/Angular. No one knows the particulars. So someone might say: “What are the best practices for developing with Vue/React/Angular?” By saying this, in a way, is trying to see how to deal with uncertainty regarding that framework. For sure the official documentation/video trainings give a way to handle the work with Vue/React/Angular. This means that, by best practice that person might mean some common norm/recipes to tackle that UI framework. If that person will follow the patterns exposed by the documentation of others then he/she will be able to learn/do/search for the work – at least this might be his/her thoughts.

With this example I have noticed some things:

– Is one thing to learn a specific syntax. Then those, introductory, videos/documentation might help;

– Is a totally different story to structure/architect the code and not being biased by the UI framework. I have to say that I do not search for a Vue/React/Angular developer, I search for a craftsperson developer which is a totally different thing(1) – by the way this is becoming a serious problem, especially in outsourcing;

– Framework designers will not ask for my or yours permission to modify or retire a framework;

So, in this case, “best practice” might mean something like finding a way to deal with unknown. But, but, if by solving this unknown thing ( at least having the impression it can be solved) in the end might use the, so called, best practices. These best practices are like universal rules for that specific framework and might be  in the thoughts of a lot of people. And here, at least for programmatic stuff, trouble might begin.

Small conclusion of this point: Best practice, maybe, is used in two ways. First to tackle/understand the unknown in a tacit way and then as universal receipts, but with no context awareness. And maybe the person using it not being aware of this.

1.2 Mechanistic thinking, truly believing/hoping there exists this “best practice” applicable everywhere.

For example, an upper manager who has to handle multiple projects, new and existing, in an outsourcing regime.

This kind of person might look to see best practices in architecture, again as recipes. He/she will want this because once it has the list of the so called architectures, then the  architecture subject will no longer be a problem. He/she can really concentrate on hiring React/Vue/.Net/… developer, not a craftsperson developer.

With testing, he/she will ask the infamous ‘lets automate all’, and a common approach to all projects of what automation to be done, like using Gherkin and that’s it.

This kind of person, thinks that everything is the same. But he/she forgets that “God is in the details”(2).

These kind of persons will come with a checklist, to be sure that some practices, they already have in their head or imposed by others, pardon best practices, will be implemented. What is sad is that those practices are the same for each project, context is not considered at all. I have noticed that some use “good practice” , not “best practice”, but their dynamic of actions is the same, they only changed “best practice” with “good practice”.

Small conclusion of this point: Here “best practice” is much more dangerous than the one described in section 1.1, because it imposes a certain way of doing things no matter what. The dynamic of actions generated by this can and, I think, will generate lot of strange and unwanted things.

2. See what exactly really is “best practice”

Here I’m influenced by the book “Tools of Critical Thinking: Metathoughts for Psychology”, in the sense that is there a chapter which gives examples of a concept being true and wrong in the same moment – I hope I recall ok this.

Also I have to say that I’m influenced here by the work of Alicia Juarrero and Dave Snowden.

I like the idea of seeing context from the point of view of constraint and casualty. For contexts were best practice might be applicable the constraints are governing ones. They are very rigid. And regarding casualty it should imply that it can be identified very clear and without doubts that event B appeared because of event A and nothing else.(3)

So, the context must be very very rigid. But when we deal with human systems, generally speaking, if it will be constraint in such a way, it will crash or find hidden ways to do the work.

Small conclusion to this point: Even if best practice might mean and have a sense, for sure it has a bounded applicability and this is not understood by a lot of people.

Conclusion: When I hear the words “best practice”, for me is like a heuristic for sloppiness/anomally/something to raise the guard, which if is imposed, it might hurt a lot. For sure is a good indication to see the situated present and how analyze/see/make a sense of it .

(1) David Schmitz, “10 Tips for failing badly at Microservices”, https://www.youtube.com/watch?v=X0tjziAQfNQ

(2) James Coplien, “Lean Architecture: for Agile Software Development”, https://www.amazon.com/Lean-Architecture-Agile-Software-Development/dp/0470684208

(3) Dave Snowden and Mary E. Boone, “A Leader’s Framework for Decision Making“, https://hbr.org/2007/11/a-leaders-framework-for-decision-making

How to make people hate the idea of estimations

Once in awhile the topic of estimations pops out in my activity. I said once in awhile because I am thinking to a special kind of way to do estimations or better said how the idea of estimation is put on the scene in some contexts.  This topic I have had it in my mind for a long time, which became obvious again these days.


1. Usually, when you ask people whether certain items can be made in a sprint ,people can say, maybe, also after some calibration, a clear answer like: “Yes, we think is ok for this sprint to do these” or “Yes, it makes sense to have those items for this sprint”. This means that although not explicitly, but implicitly, an estimation is being made.


–  Let’s suppose that some unknowns were already clarified because of some spikes/investigations done a priori;

– “calibration” is an important word[1];

But, and maybe I am wrong,  the same people when asked in detail(hours/points in the sprint) they will not feel comfortable with the estimation and with the meeting(s), because of doing this. I think this happens also because humans are messy[2] and is ok.

What is not so ok, I think, is why to ask more specific estimation detail, for a sprint, when already the team said those things can be made in the sprint, and of course with the information they have for now is the maximum of work they can accept as a team?

There are some possible answers I can think of:

– maybe someone wants control in the sprint( maybe the managers of the manager of the project impose on him , or lack of trust, or outsourcing context where client  paying by day involuntary triggers this need or because he/she knows that the setup of the team is not ok(let’s say the competency of team members)).

– I saw how tools like Jira entice some managers to ask these things;

– they do not understand why sprints/iterations were created;

2. Then are those estimations when getting a project and it has to be finished in, let’s say, 6 months. And the estimations “must be ok, they must not be wrong”. So, we have an initial “must be ok” estimation. But then after 4 sprints again a new estimation is done – of course I imagine people joy to do that because the estimation wanted to be done, strangely, must be “ok”…

So that kind of project is being managed in a way by the “must be ok” estimations, if I can say so – just to make me clear: I am speaking about ( and I am influenced by Alicia Juarrero[3] ) using estimations as a placid background, like an equilibrium structure, like an indication of stability( small deviations around equilibrium).

Note: I am assuming that a project is not in an obvious state, it can be also in complex situations.


Estimations – in a way or the other – we use them. We use them implicitly or explicitly or deduced and in various forms( relative, time, distributions…) and it makes sense as a concept.

I think that the estimations, more often, are being used by managers in a wrong way, for example like a pressure consciously/unconsciously/unknowingly. And this is actually the problem. Maybe this is happening because of the mechanistic way of their thinking or lack of knowledge or … They choose the wrong metaphor, they do not deal with a refinery/factory, but with an ecology.  Also, probably at least what I saw, most of them do not take a look at what psychology, neuroscience have to say about this and adapt their actions.

I do not think is ok to encourage the dichotomy ProEstimate and No Estimate, I think here is a continuum between them.

(1) Adrian Lander, “Linked In discussion”, https://www.linkedin.com/feed/update/urn:li:activity:6426405267491164160

(2) Dave Snowden, “Humans are messy”, http://cognitive-edge.com/blog/humans-are-messy/

(3) Alicia Juarrero, “Safe-Fail, NOT Fail-Safe  ”, https://vimeo.com/95646156

An example of what it means for a deeply experienced person to held an interview for hiring

Yesterday I wrote about why it matters that a deeply experienced/senior person held an interview.

Today is time to write about such kind of person(although to say about him that is senior is too little, he is much much more) . Is about a special and important example of an action, made by him, in context of interviews . Is special and important because he helped people which now are ok because of him.

His name is Flaviu Boldea.

Almost 4 years ago we held lots of interviews. I was sad because I saw good/nice/pleasant people which did not passed the interview. With his calm voice he told me that this does not mean it should stop there. If I see a candidate which is ok, as a person, but is not passing, yet, the evaluation to offer him/her my time to help them prepare. Is important to say that he had done this several times before, so his words were not just some empty words.

I liked that he did not transform this in a process to do in the company. No no, it was about our willing, as simple persons, to help from our free time. It was about taking the responsibility of doing something and not to abdicate after saying no. It was not easy at all, but it was rewarding.

Experience/type of people involved/seniority/professionalism/craftsmanship/… matters.

About an anomaly in interviews and evaluations within companies

Too often I began to see a strange thing regarding people who make interviews for hiring or evaluates other people in companies. I saw that those persons are not actually seniors(deeply experienced) although there are , in the company, also seniors, who can handle this part. As I said, is a thing I experienced – maybe is not generally applicable, but it made me think deeply about this.

Another detail I did not mention is that these evaluations are made after a checklist, sometimes having a Dreyfuss representation.

I am not speaking regarding the practice that, at interviews for hiring, team members can join the interview just to see the type of the person being interviewed.

I try to be careful about dichotomies in the sense that more often  people use dichotomies when dealing with situations when is not the case. So, in our context, I imagine that is possible that at a certain moment a person with not so much experience can hold an interview/evaluation because of a special context( for example he/she is alone and is an urgent need).

Why I  thought that this might be a problem? Because:

  • experience matters. It matters because of the: different types of situations that person had, different types of skills he/she has acquired, the knowledge he/she has gained, the bad experiences he/she lived(all those lessons learned the hard way), familiarity of situation(s),…;
  • is hard for the evaluator to transcend the things. What if that checklist will not be of any direct help, does it mean the candidates will not be evaluated ok?;
  • It will be hard for him/her to see not just the letter of the law, but also the spirit of the law;
  • the language will be different. And now, I remembered Alistair Cockburn work. A Shu level has a different language than the Ha level or the Ri level or the Kokoro level. A Kokoro level having passed through all stages, if I can say so, will be able to understand and recognize the other levels even the Kokoro level itself. This means it will be a problem to decodify/interpret/analyze/acknowledge what the respondent will say;
  • Go to other domains like medicine. Is it normal for an intern to decide a certain level of an experienced surgeon?

Is important to mention that this evaluation can have an important impact on the person being evaluated( maybe is about money, image, dreams, promotion which can be affected in a bad way) or on the company also. Should an evaluation be let in hands of the not experienced? I hope not. I hope to see human judgment prevail, not to see some actions inspired by a checklist used by inexperienced people.(note: checklists are good, but when I see that for each type of problem the answer is to make a checklist then, maybe, something is not ok).

Forming a team, an agile perspective

I’m seriously concerned about this topic because I see the disastrous effects of its bad handling every day.  It’s not enough to put some people together and ask them, usually in a perverse, detached and manipulative way, that no matter the difficulties encountered, they should make it work because this is what professionals do – in a way this is like an abdication of the job a manager should do rigorously. I know it’s hard, damn hard, I made my share of mistakes …

I wrote about this topic before.

But this time I would like to recall two important perspectives, from the agile world, that unfortunately seem to be ignored:

1. Early this year a nice interview(1) with Jerry Weinberg was published. In it, he described one method to form a team:

“…We put together teams. What I recommend is people put together their agile teams by what I call incremental consensus. You start with one person who you feel has the personal qualities. Stop emphasizing very soon with what I read about is people emphasizing: you got to have the most skilled coder and you can put all the best coders together and then you have an agile team. Now you find someone who has got these personal qualities: they know how to communicate , they’re willing to communicate, they’re willing to admit their mistakes and learn. And you start with that person. And you say: okay now among all the people we have available to us who would you like most to team up with to do a good job; you understand the agile principles. And they choose of course… it’s one person so it’s a consensus and the person has to want to be on it … Now you have two people. Now you say to them: okay now come to an agreement on who else you’d like to have. If you started with the right person the process goes on and they put together a team that really does well; and they take awhile to learn but they can learn the skills, technical skills, more easily than they can learn the personal skills, the people’s skills as we call them…”

In 2016, after a horrible month of solving some issues, with a delicate suite of products, I realized that this is the time when the core problem could be solved. It could be solved because the key persons were prepared to allow the kind of work I had in my mind for almost 2 years. In that terrible month, I have worked with some wonderful people to fix the problems. I knew that the kind of work I wanted to do needed special people. Even before I spoke with the management, I spoke with each of them separately. I told them what I had in mind and the persons I wanted to do the work with. I needed their green light regarding the work that needs to be done, their future colleagues and if they were willing to join the team. I left them time to reflect on all those things. In this time I talked with no one else. After several days we spoke again, separately, and each of them gave the green light. Then I spoke with management and it was hard…but I was determined, if one of them could not join the new setup, I would not take in charge that new project/product. I wanted from each of the future team members their consensus. And as Jerry Weinberg said, it was not only about the technical skills of these people, it was much much more. Please watch the Jerry Weinberg interview.

2. Scrum Plop(2) – I am amazed that so many Scrum people, from my circles, are not aware of this. It’s about patterns that show and help on how to implement Scrum.  If Scrum guide seems abstract and hard to be decoded, well these patterns help a lot in making sense on how to implement Scrum. There is a pattern called Stable Teams(3) which accurately articulates the problem and also gives a solution:

“…Keep teams stable and avoid shuffling people around between teams. Stable teams tend to get to know their capacity, which makes it possible for the business to have some predictability. Dedicate team members to a single team whenever possible. Members of Stable Teams get to know each other. The team members experience each other’s work style and learn how much work they can do together. A Stable Team grows in familiarity and consistency of meeting mutual expectations and starts developing a Community of Trust…. what is often forgotten when measuring a team’s velocity is that the only way that a team can get to know their velocity is by having the same team members over a longer period of time…”

Trust is a key/serious/important word. Trust takes time because is arising from people interactions over a certain period of time. You cannot create a roadmap or a checklist on how to build trust within a team.

Since I mentioned Scrum above, it is important to also mention 2 important related details about it:

– an important detail, also applicable to teams, is about competence(4)/skill/mastery/expertise. Although this is an aspect that is not explicitly specified in the Scrum guide, it is a sine qua non condition for Scrum to work;

– management plays an important role when forming the team. Not an easy thing to do. But after they’ve formed the team, they should trust them and let them accomplish their mission(5).


Every time I think about this subject, I think about families. Look at families, at how delicate  things can get sometimes, because family members can’t get along with each other. And it’s not like this problem can be easily solved by a parent, for example, by just saying: “hey, you are brothers, you are supposed to get along”. Some will say, when speaking about team members, that it’s not about friends or family members. And I say “Exactly. It’s even harder and it’s not that easy to impose things”. So, I imagine that for some managers it’s easier to pretend that they don’t know things/techniques  on forming a team, and then use some fancy words and clichés to “solve it”, thus tossing the effort they should have done onto the shoulders of other people. But the reality has its own style to deny certain approaches/ideas/opinions.

What was a surprise for me is that it can be really really dangerous to desire having a Stable Team. The surprise was even bigger that the danger came from people which actually, at least declaratively, supports and promotes Scrum and agility.

(1) Jerry Weinberg, “Agile for Humans #26: Agile Impressions and Errors with Jerry Weinberg”, https://www.youtube.com/watch?v=I_AOoSnhE-8

(2) Scrum Pattern Community, Jim Coplien – Product Owner for the Scrum Patterns effort, http://www.scrumplop.org/

(3) Stable Teams Pattern, https://sites.google.com/a/scrumplop.org/published-patterns/product-organization-pattern-language/development-team/stable-teams

(4) Jim Coplien, “Ten things scrummers do wrong”, https://vimeo.com/42772592

(5) Alistair Cockburn, “Core scrum, barnacles, rumors and hearsay, improved version”, https://www.youtube.com/watch?v=AuUadPoi35M

Forming a team discussion metaphor

Shu(1): I need your help.

Ri: About what?

Shu: Help me with a metaphor. I see teams formed just by simple association of people, at least this is my feeling.  I want to tell my boss, or anyone who does this, that is not quite like this. Sometimes is hard for me to explain a certain subject without being to verbose. But now I need something simple, but powerful to explain what I feel.

Ri: Hmm….[smiling]

Shu: Is hard to start a new battle and a lot of problems will appear because of this …. help me please.

[A moment of silence appears. Shu is looking at Ri puzzled. Suddenly the Ri’s face became serious and his right palm suddenly and firmly is put on the table.]

Ri: Look at the fingers. All have their place.

[Shu looked for several seconds, and put his right hand near Ri’s hand]

Shu: And my fingers are not on your hand and your fingers are not on my hand. Each have its place.

[Ri smiled]

Ri: You see, with this hand you can do a lot of things. It depends on you. The cohesion of the fingers helps in doing different things. They work for a purpose/goal. The same is with people, they are different and they have to work for a common goal/purpose/direction.

[Shu interrupts Ri, like continuing what Ri says]

Shu: And you cannot do that if you are not careful on how you choose those people. If you have a broken/abnormal finger it affects the things you can do with the hand.

[Ri smiled]

From a discussion had with clinical psychologist Dr. Ion Zamfir regarding management.

(1) Alistair Cockburn, “Shu Ha Ri”: http://alistair.cockburn.us/Shu+Ha+Ri

The problem of connecting the dots

Recently I read an article (1) . The article is about Fred Brooks, his book (“‎The Mythical Man-Month‎”’) and the fact that the overhead of communication is direct proportionality with the number of people and  the number of different communication channels, also increases rapidly with the number of employees.

I liked a lot the drawings from the article. It was a very nice representation of this connection between increasing the number of people and increasing the number of communication channels. After I read the article, I realised those images don’t fully represent the whole story.

The following things activated in my head, maybe I am wrong, but still:

1) Max Boisot made similar drawings(2) and observed an interesting thing – yes, at that moment, he was speaking in the context of international terrorism. So, depending on the number of dots he have:

-for N=4 we have 6 possible links and 64 possible patterns;

-for N=10 we have 45 possible links and 3.5 trillion possible patterns;

-for N=12 we have 66 possible links and 4.700 quadrillion possible patterns;

For him dots were the data, links the information and patterns the knowledge. And an important point is to process the patterns.

2) Then I remembered what Dave Snowden said regarding identity in humans, inspired by anthropology :”Identity is fluid in humans, we move between roles depending on context and have developed rituals by which we can temporarily align our identity with a role for collective purpose.”(3)

So, maybe one dot, from the drawing, actually  can be more than one dot because people can shift identities.

3) For sure Fred Brooks was right in the sense that we should not add people late in the project. And yes “‎The Mythical Man-Month” is a wonderful book.

Conclusion: In a sense those drawings, from the article mentioned, made me visualize how nasty things can be. Those drawings are beautiful, but somehow incomplete, is just the tip of the iceberg. They made me understand better how tangled complexity can be.

Regarding the conclusion “les managers doivent s’assurer d’optimiser le nombre de salariés travaillant sur une tâche, ni trop, ni trop peu…” –> I think that, if the manager (first line manager) is not involved in the work(does not interact with the system, is not with the ‘skin in the game’)  the chance to make bad decisions is increased a lot. I think is possible for a project with just 3 people to give a lot of headache. The manager should be aware that:

– there will always be hidden surprises;

– there will be volatility;

– when there is prediction to be sure unpredicted things will pop-up.

Note: The title of this blog post is inspired by (2)

(1) Cadreo, “La loi de Brooks ou pourquoi la multiplication des collaborateurs fait perdre du temps”, https://www.cadreo.com/actualites/dt-loi-brooks-ou-pourquoi-multiplier-les-collaborateurs-sur-un-projet-fait-perdre-du-temps

(2) Christopher  Bellavita, “Nidal Hasan and the problem of connecting the dots”, http://www.hlswatch.com/2009/11/12/nidal-hasan-and-the-problem-of-connecting-the-dots/

(3) Dave Snowden, “Three or five?”, http://cognitive-edge.com/blog/three-or-five/