In this talk, the speaker introduces "Restraint" as the second R, arguing that adding more components to a system does not improve performance because systems operate multiplicatively, not additively. Using a light‑switch example, he shows how each component’s less‑than‑perfect reliability compounds, dramatically lowering overall system reliability. He identifies the human operator as the least reliable element in most systems and suggests that improving human reliability—by defining essential, non‑negotiable elements, setting clear boundaries, and taking micro‑steps—raises the floor of the system’s performance, yielding far greater returns than trying to push the ceiling of already efficient components. The speaker emphasizes focusing on the weakest link, establishing certainty about the future, and prioritizing efforts that lift the reliability floor for exponential gains.
Chapters
Second R is Restraint. This is the idea that more does not equal closer. And I've written about this a bunch. We are programmed for more. We are additive creatures. We want, you know, this is why we have big houses. This is why we have storage units. This is why we have closets full of clothes that we never wear because we are just programmed more is better. Actually, that's not true. In terms of system reliability, if you are, for instance, if you have a business and you have a, um, sales success system, a revenue-generating system built, the more things that you add to that system, the lower that system performs because Systems are not additive. Systems are the multiplication of percentages because nothing is 100% guaranteed. Nothing works 100% of the time. Best example, just a quick example that has nothing to do with business. If you, your light in your office or living room operates as part of a system, there's a light switch, there's wires that lead to the light, there's the light fixture, there's a light bulb, there's the, um, there's the, the electric bill. I mean, all of these things have to be at 100%, and none of them are at 100%, 100% of the time. And so if your light goes on 90% of the time and the light switch is working 90% of the time, your system reliability is 81%. 0.9 times 0.9. If the human only remembers to pay the phone bill 90% of the time, you just drove down the efficiency to 72%, and you only got 3 components and they're all 90% effective. So just think about that when you're thinking about any kind of system, and everything is a system. And every system— what is it that Deming said? Every system is perfectly designed to get the result it gets. So, and in any system, anybody want to take a guess at which is the least reliable piece? Scott or Jeff? I would have to say the human. The human pushing the buttons is always the least reliable. This is why a lot of what we talk about is about working on yourself, becoming, um, seeing, stepping into, and staying in your power to take control of your destiny. Uh, so the more reliable that we can make the human, that is the, the, the piece of the system that if we can raise the floor on the human's reliability, we raise the ceiling on the overall efficiency of the system. So what we want to do in this restraint piece is we want to define what's essential, what's non-negotiable. And we want to set up the boundaries about who we let in, who we keep out, what we let in, what we keep out, what behavior we allow ourselves to engage in and what we're where we don't, and we want to micro-step our way into a more reliable system. We, we're actually increasing what Dan Nicholson would call certainty about our future by paying attention to the essential components of the system and improving the least efficient piece each time. Each, each, each with each iteration. And the, the kind of the, the pro tip here is our instinct is to raise the ceiling on the most efficient. The problem is once you get to about 80%, any effort and attention and time that you put into raising the highest efficient piece, pushing the ceiling, the return on investment that is extraordinarily low. Whereas if you push the floor of the least effective component, the return on investment of time, attention, money, et cetera, is exponential. That's why we're always raising the floor.
