“Wise is the one who chooses the questions well, not the one who tries to solve them all”

Securing a permanent position in science often implies you invest an increasing amount of time in managing people (and projects, and budgets) rather than doing science. Unfortunately, scientists are not trained in managing. While some people might have natural (or learned) skills, most struggle and have to learn them the hard way (i.e. by trial and error). As a mixed strategy, I read a bit about managing people during the last few years (and also “tried and err” a lot). Rather than a proper post, I am using this entry to bullet point things I found useful (and try to practice with more or less success). This is written mainly for me, as a place to go back and refresh some of the ideas, because most of these ideas are easier to say than to do and need revisiting and practice, but maybe it is also useful for someone else. 

• A good group leader is the one who cares. It doesn’t matter if you mess up often, as far as you care and try hard. You can be better or worse at it at the beginning, but most problematic PI’s simply don’t care about their group. If you care about it, you will be fine, and it’s just a matter of time to get better. 
• Learn from your heroes. Science fiction and fantasy books are full of leaders. Which ones excel at leading groups? For PhD advising, I love the Gandalf analogy. You set up the plan and bring up the motivation, then you disappear and let the student lead their adventure, but if there is trouble, you always arrive on time to support and bring reinforcements if needed. Kelsier from Mistborn I is also a good example of forming a united group built on mutual trust and a common goal. Ned Stark (Game of Thorns) nails it when suggesting not to impose a task on anyone (the task can be beheading a person, but also working 8 straight hours under the sun) that you won’t like/accept to do yourself.
• Disagree and commit. In a team, you need to be able to disagree, but if a decision you do not support is made as a group, then you fully commit to it. Looks easy, but many people fail either to disagree (so you are not heard) or to commit (boycotting the project on purpose or not). If you care about a group, then you explain your reservations but respect the group’s decision.
• Define who is making decisions. There is nothing more frustrating than a meeting where you don’t know who calls the final decision. Would this be a consensus decision? Someone will decide pondering all the discussion? Or this is simply a brainstorming meeting and no decisions will be made at this point? Clarify that and your meetings will relax a lot.
• Learn when to talk first or talk last. Talking first sets the tone, can bias your team’s ideas, but ensures the discussion does not go in a totally unwanted direction. Talking last promotes candor in the participants, you practice listening and allows you to talk with a balanced and informed position. Of course, ask a lot of questions in between.
• Ask a lot of questions. This is the only way to understand.
• If you have drawn an idea, write it up and vv. This is a recurrent piece of advice I give. If an idea works in a graph, try writing it down. If your paragraph makes sense, draw it, make a figure, a causal path, or a sketch. Often they complement and enlighten.
• Start with why: Always give context. Why we want to do a project is more important than how can we do it, or what should we do to implement it. Explain your feelings.
• Give credit, take the blame. As a group, any credit goes to the team, and all the blame to the leader. Period.
• If in doubt, talk in person. Any friction or sensitive topic should be dealt with in person (or by telephone if in-person is not possible). Emails/Slack are great for everyday exchanges, but if the topic is complex, make it personal. 
• Assume good faith. Do not try to read behaviors. If something botehrs you, assume good faith and talk it in person as soon as possible. 
• Choose your teammates. Working with generous, fun people is better.
• Be process-oriented and not goal-oriented. People who play chess for fun, keep playing even when they lose, get better in the process, and have fun (and become infinite players). People who play to win stop when they lose, don’t want to play with stronger rivals, and are stressed. In science, we play for fun.
• Trust. It’s your team, you trust them, and you give them agency. Things will never be exactly as you initially envisioned. But often they will be better.
• Let people up-manage you. You work for them, not the other way around. When you want a task to be done, ask people how you can help. They know better what they need from you in different situations (e.g. resources, talk to key people, redefine objectives). Don’t try to guess and let them tell you what is blocking the task, and what you can do. If they only need you to do nothing, do nothing.
• Discuss ideas, not sides. Never confront what person A proposed against what person B proposed, even if it’s not your intention and you just use the person’s name for simplicity. Discuss idea X and idea Y, both emerging from the team. Wording matters more than we think.
• Part of your job is listening about and solving problems not work-related. This is time well invested. 
• Use retrospectives to reflect on what worked great in a project and should be repeated, what worked just ok and can be enhanced, and what did not work and should be avoided next time. Don’t miss learning opportunities. Focus on what we learned and not on blaming. Acknowledge the contingencies, and be aware that you now have all the information, but at the time of making many decisions, our knowledge was partial.
• When conflict emerges play “convince-me”, and give the floor to whoever thinks differently and just listen and ask questions. Genuinely aim to be convinced. 
• Assume you are a leader, and take risks, and take responsibility.

Overall, invest in team culture. This does not mean your lab needs to be all super good friends, they only need to work well together and trust each other. For example, I liked the radical candor approach, but there are many more approaches. Nothing of what I said is easy to do, or to do well, and not always work as expected, but is a good start.

Feel free to add what works for you in the comments.