Are you overwhelmed by work?
Do you feel like being constantly under time pressure?
As I’m talking with colleagues in nearly all organizations, there seems to be an overall increase in workload, shorter timelines, a reduced number of people to help, and also an increase in complexity of the projects.
This leads to the breakdown of the system.
There’s some interesting research by a German mathematician, who wanted to understand why the German railway system sometimes completely breaks down and delays accumulate quickly over the day.
The answer comes with the buffer in the system. We always have disturbances in terms of unplanned events in complex systems – like the German trains. If there isn’t enough buffer built in, these minor problems lead to a break-down as the system cannot adjust. Consecutive problems eat up the little buffer and delays accumulate.
Listen to this episode to learn about ways to build in buffer!
Alexander: You’re listening to the effective statistician podcast, the weekly podcast with Alexander Schacht, Benjamin Piske and Sam Gardner, designed to help you reach your potential to lead great science and serve patients without becoming overwhelmed by work.
Today, we are talking about The Lack of buffers in complex systems and what this has to do with your day-to-day work. Sometime ago. I was thinking something about why train systems break down and it was from a mathematical perspective, very interesting. And that was some further research about it. Then that was, generally about complex systems and This episode is about this topic because we work in complex systems. We can’t see what’s going on. We can’t foresee what will change? There’s always something new coming, there’s styling updates coming, whatsoever and your supervisor might chime in and give you a new task. So we have been working on a very complex environment and what does buffer have to do with it? Well, obviously, a small buffer is in the system, the better it works, but do we really plan for it. How much buffer do we need? Stay tuned for these topics to be discussed later on.
I’m producing this podcast in association with PSI, a community dedicated to leading and promoting the use of statistics within the healthcare industry for the benefit of patients. Join PSI today to further develop your statistical capabilities with access to the ever growing video on demand content library, free registration to all PSI webinars, and much, much more. Head over to the PSI web.org, to learn more about PSI activities and become a PSI member today.
Welcome to another episode. And again, it’s Benjamin! How are you doing?
Benjamin: Very good! Thanks Alex. How are you? Do you have a complex day today?
Alexander: Actually, it’s a time It’s not that complex. But usually when we work there’s a lot of Shifting things around, isn’t it you’re for a different project you have, things that move out, things that move in obviously and usually everything at the same time. And these people need to go crazy to meet deadlines and work long hours and everything becomes hectic. I was recently talking to someone and she was so hectic but I think if the database is done then it will get better and I was thinking, didn’t you tell me this, the last three times we talked with each other? It never became really better. And you see people in these organizations becoming crazy because there’s one deadline after the other. There’s one emergency after the other and people are like fire fighting all time. Have you seen that?
Benjamin: Absolutely. Well, I don’t think it’s necessarily fire fighting. What I see is more like a constant threat in the complexity of the studies that we are working on, so it is not necessary that we are not aware of things or that. I mean, there’s often something moving in, as you said, usually moving in rather than out, but that again is messing up the whole plan because you usually have quite a good plan and overlapping responsibilities or studies. And so it fits nicely, and as long as it fits, there’s no need for any support or escalation and everything is running smoothly. People are busy. It’s not that they’re not busy or they’re getting bored. It’s just that they have a plan for everything. But if one tiny piece falls apart, that’ll be timelines or maybe something about unexpected data quality issues, anything that’s messing around your plans. It may get up to fire fighting. I agree, but also just purely to unexpected overload and whatever new task coming in that you haven’t been planning? You haven’t been thinking about it before.
Alexander: And these kinds of little changes. If you have these complex systems, where everything works at top speed and seeing them, the system’s crash. I once had this when I was going to the PSI conference in Amsterdam, And I went by train and the first train, I had 11 minutes delay. Not a big thing, if you think about a three hour train ride. Unfortunately, I only had 10 minutes to switch trains at the station. And then instead of having an 11 minutes delay, I headed for a two hour delay. Then I needed to wait for the next train. And so this, if something little goes wrong in complex systems that can multiply quite a lot.
Benjamin: Yeah, giving the example of the trains, it’s not that the system ideally doesn’t work. So if there’s nothing breaking down if there’s no coal on the tracks or whatever. Everything is going smoothly, it works. So it’s well planned and well organized, and there is probably room for improvement. And it’s always funny if for example, if you go to Switzerland and they start apologizing for a 3 minute delay, where you always think coming from Germany. It’s like why, what exactly are you apologizing for? So it’s their different system. It’s different but in general, it really works in general. But in your case, if it’s 11 minutes late, instead of 10 minutes than you usually had, it’s breaking down.
Alexander: But if you compare the Swiss Train system to a German one. The German ones are much more complex. And so this increase in complexity means that there are much more things that can go wrong. So it’s just much more room for error. And for things like a crash, an accident, technical problems, weather problems or whatsoever. There are thousands of things that can happen, and then these kinds of delays add up. And the same thing happens at work. It’s completely the same way, if you have a lot of moving targets and everything is working together. And you have these different studies that influence each other and hand over those different people working at such an overall study is complex and the overall clinical development process even more complex. So there’s lots of moving targets. In terms of the German Railway system, That’s actually an interesting thing because there was a mathematician in Germany that looked into these types of complex systems and He did more research in it and he discovered that when you have a system and you try to get it as efficient as possible, by getting everything to maximum capacity there is something interesting happens like up to 80% – 85%, small problems don’t add up. Between 85-90 percent something interesting happens. Any small problems, then directly shoots the capacity of the system beyond 100 %. Then all these different things add up. Like in my two-hour delay to Amsterdam, one thing that fits to the next and you get this domino effect. That’s not stopping you. And so the same happens at work if you have a full plan for everybody to their full capacity. You basically need to assume that everything works exactly like your plan.
Benjamin: Did it ever work exactly like you have planned and will we see where we’re getting to?
Alexander: Of course, it never works like you planned. There’s this famous quote by Malka. Malka, a German general who said, your plan of attack never survives, the first contact with the major force of the enemy. Because things always change and planning is really important. But completely dependent on exactly implementing your plan leads to disaster.
Benjamin: Hmm. you might be wondering why, if you say, it adds up and will stress the system and even overcome the capacity of the system, so how do we then actually manage to do it? And if you recall that? That we quite often think about the Peaks of our workload. And so, we are in a lucky position, unlike the Deutsche Bahn or any other train that we have over capacities, natural over capacities, which is overtime usually or maybe just, putting something aside which is less important for the moment and saying that it is not as important and so we’ll do it next week. And, you see that this is leading to the next problem. But in the firefighting activities where we then start. But in reality, we are quite not completely comparable with the bond with the trains because of the type of capacities that we have, like flexibility. However, this is also only limited because there’s no unlimited overtime or unlimited peak that we can work on and everyone is expecting some peak. So it’s nothing surprising that what we are talking about today is that everyone is doing this on a day-to-day business like working on the things that may come up with a higher Peak than expected or maybe a peak in general. So as you said, it’s nice that mathematicians prove something at some point and they might be right, but we all know it already.
Alexander: We all know it already. But the important point here is that we get into trouble if we plan from these kinds of deadlines only. Yeah, here are the different deadlines and then we plan our capacity in such a way that we will be able to be just in time to deliver to all these different deadlines. if you think about the Eisenhower metrics of Important versus Urgent. We have these four different quadrants, Important but not urgent. Urgent but not important, Urgent and important and so on. Of course, you don’t want to never work if this is not urgent, not important things. But, if you work only on the things that are important and urgent, then you’ll get into trouble. So I think you need to have the capacity to work on important, not yet urgent things. And if you have that area then you are in a really nice spot. And so have an overall plan that works like this and that also allows you to have this buffer because if you have this buffer basically in these Urgent/Important, Not yet Urgent things.
Benjamin: I think you also have to convince people, or even companies, to think about this system. Because the argument of saying, I plan in the buffer for things that we are not expecting is usually being pushed back and saying, well, I mean, plan better and plan the unexpected because we cannot afford to have a plan with a gap of a day. Everyone is saying, why didn’t we do it in a day? We will do what we would like to have one day earlier. I know it’s too important to wait for another day. Just in case. And so, come back to the example of the method of Mathematicians. It is one of the key roles of a statistician in general to convince non-mathematician of problems that they see and the risks and I think that is something we definitely need to pick up at some point. And demonstrate and give the example of why we need to have this extra day, or it is just the experience that, we’ve been through hundreds of studies and trials and deliveries and whatsoever and none of them you can compare with another none of them. So it’s all different, all different challenges. I mean, they overlap obviously but it’s in general. Whatever is pushing you back in the delivery. It’s something different and you’ve never experienced before. Otherwise, we would have thought otherwise we would have planned it in.
Alexander: And I don’t know if that’s the nature of our work environment. You always have new people joining. You have new study design features, you have new updated processes or all these kinds of other things that can happen. Like Covid. So, just plan for these unexpected things. Then if you see that you still have time you can work on the important not yet urgent things. Yeah, and that always will help you to get this stress out of the system and will make sure that you are satisfied with your work and you are less heroic. “Oh, we work to complete a weekend to meet a deadline on Monday”. We shouldn’t actually honor that, this kind of business. That’s actually bad management, if you need to do that on a continuous basis then it’s actually bad management.
Benjamin: And just a tip for proposing this, don’t do this in an extra single day to give extra as a buffer so plan it every step of the time so it’s not very obvious. Maybe a day longer and if you plan this too offensive or too open with the buffer, this will be falling back on the study team saying, you are very unorganized if you plan in an extra day so you should be better organized and plan. So that’s why be realistic in a way. And that is something that I love about people is that they’re thinking very positively and that’s including myself. It is really that we think, “well, yes, that should be manageable and we’ll manage it, don’t worry”. So it’s being convinced that you do it.
Alexander: That is a clear bias. And it’s well documented. We always overestimate how fast we can be.
Benjamin: How great we are.
Alexander: How great we are. In the short term and the long term we usually underestimate. So we overestimate what we can do in a day, or in a week. Or things like that. We underestimate what we can do in three years or five years. But as we plan in these days and weeks time frames, we usually overestimate. Look at your to-do list that you’ve written this morning. Do you always complete it? No, and that’s why I went back to complete three really main important things to call this day a success. Because otherwise you’re always depressed. So, Clam in your buffer depending on the system that you work with. I also worked in systems like critical chain management where you actually have buffered and unbuffered timelines, if you need to send back your timelines, do it. Because that will help you. And then you can front load your work and see where you can better manage things around. In most of the cases, you don’t need to go into this heroic last minute fighting stuff.
Benjamin: And also, involve your supervisor. In some cases, especially in our business, we don’t often get pushback on suggested timelines, for whatever reason. So they’re unrealistic, but they’re sometimes just not understood correctly. And if you need support, talk to your supervisor, get people involved in that act. I wouldn’t even call it escalation, but just the first point of escalation outside the study team. Where somebody more objective can give examples or Justify timelines or just talk to the client on a different level and that will help often,not necessarily. But that is something at least where you can work with and get better. Don’t let the pressure of individuals or these clients or study teams destroy your plans if you’re planning a buffer. So if you plan, plan realistically, I know it’s difficult as we are so time driven and all around the pharmaceutical industry with getting medication in a market or a publication out or whatsoever. It’s all driven by timelines. So there’s pressure, no doubt, but still, it’s better to have an extra day and make it that extra day, rather than having It a week late because of unexpected things and then know people are not being available etc and having a vacation.
Alexander: I think that was a really important episode and because it’s coming up on a continuous basis. It’s nothing that is going to change it, you will always need to fight for it. It’s not like you once implemented and then it’s done. It’s something that you need to continuously work on in mind because the pressure will be always there to kind of shorten things. And whenever you plan, you might fall into this cognitive bias of planning for everything to move smoothly, which, of course, doesn’t. So talk to you next time. Have a nice week.
Alexander: This show was created in association with PSI. Thanks to Reine who helps with a show in the background and thank you for listening. Please tell others about the show if you like it, if you found this episode to be something you want to share with others, so that they also plan a buffer into their Project, into their Daily actions, into their Monthly actions. Then share this episode with others, I would love them to benefit from this learning for as many people as possible. Reach a potential Lead Great Science and serve patients, Just Be an Effective statistician.
Never miss an episode of The Effective Statistician
Join hundreds of your peers and subscribe to get our latest updates by email!