These are my notes on curi's analysis and commentary on Chapters 1-2 of Goldratt's book "Theory of Constraints". The video will be available in the future.
@09:00 number of constraints in a system / boy scout analogy
Goldratt has talked elsewhere about how if a system had no constraints it would grow without limit. We don't see that, so there must be at least 1 constraint.
Theory of Constraints, Goldratt says that real-life systems has only a few (like 2 or 3?) constraints. The boy scout analogy doesn't really show that the limit is 2 or 3, IMO.
@10:21 Focusing Step 2: Exploit
curi points out the word "exploit" is unclear. I agree. curi's suggested replacement, "optimize", is an improvement.
Elsewhere Goldratt or one of his people stated that the "exploit" step means: take whatever steps allow you to get more goal units out of a constrained resource *without requiring additional investment*. Here the avoidance of additional *investment* is to be contrasted with additional *expense*, which is OK in this step. And note that TOC classifies employee salaries as expense, and buying a new machine as investment. So adding an additional shift on a bottleneck machine would be an appropriate action for the Exploit step, but buying a second machine would not, even if we assume that either option would enable the system to double the machine's output.
I'm not sure this is said anywhere, but I think Exploit also means local optimizations only. If an improvement involves changing other parts of the system, then it's not part of the Exploit step.
@11:30 Focusing Step 3: Subordinate
curi points out that "subordinate" isn't great wording. Again, I agree, but I'm not sure of a better one here. It's telling that English doesn't yet have a good word for this important concept.
Elsewhere, Goldratt or one of his people said that "subordinate" means: update the policies and priorities of *every other part* of the system in order to maximize (in terms of the goal) the output of the constrained resource.
@12:11 Focusing Step 4: Elevate
curi again points out that this is not a clear word, and I agree, though once again, I'm not sure of a better one.
The meaning is to make investments to directly improve the performance of the constrained resource itself.
@13:53 Moving constraints
> You don't want the constraint to move around a lot, because it takes significant effort to re-organize around wherever it is, so, to some extent you want to figure out like, where is a good place for the constraint to be, because there has to be a weakest link, so it's like, ok, what do you want the weakest link to be? Figure out a good place to put it and then just leave it there and organize around that.
Yeah. The ideal location of the constraint is a function of the underlying structure and economics of the business. For example, for a retail shop, I think Goldratt has said that the fundamental constraint is the amount of walk in traffic the shop gets from qualified customers. That's not going away, so it makes sense to organize the business around that as a constraint.
@14:37 Inertia
Inertia is a huge issue. Even aside from its importance as something to be avoided in the fifth focusing step, Goldratt talks in Beyond the Goal about how inertia keeps people from realizing the benefits of technology. (Discussed at
http://curi.us/2094-expanding-our-limits)
In general, People just think that what they have is good enough and carry on. No one is interested enough to look for big improvements in the way things are done.
@15:35 "policy constraints"
Goldratt mentions policy constraints in the book here, but elsewhere he said that he no longer considers policies to be *constraints* as such. I can't find a quote from Goldratt on the topic right now, but I did find a related quote from an informative essay on TOC from 2004:
https://web.archive.org/web/20040803165000/http://www.dbrmfg.co.nz/Bottom%20Line%20Process%20of%20Change.htm
> Although the word “policy constraint” has been more recently disowned, maybe “policy issue” is more apt; there are still many more than one constraint per system.
@16:45 describing focusing steps in another way
> It sounds like he wants to describe the same five steps -- these are called the "focusing steps" -- in some other way... I don't think I've read a second version of these five steps that's like, presented from a different perspective.
Where did curi see this? I read the passage and couldn't find it. Later (1:27:00), Goldratt refers to "[t]he steps which are equivalent to the above five steps, but are expressed in the terminology of the improvement process itself", so curi is right, but what did he see?
Skipping ahead to @1:27:30:
> I've seen this before in terms of like trying to find the right change: what to change/what do you change it to/how do you make it happen, but I didn't know it was meant to be equivalent to the five focusing steps. I thought of it as a different thing, rather than as a different version of the same thing.
Did curi mis-remember what he was thinking 45 minutes ago? At 16:45, he referred to it as a "second version of these five steps", but at 1:27:30 he said he didn't think of it as a "different version" of the five focusing steps.
@18:05 process of ongoing improvement
Goldratt's concept of a "process of ongoing improvement" (POOGI) is like the BoI idea of traditions that enable continual knowledge creation.
@19:00 Figuring out what to change to
In Theory of Constraints, Goldratt wrote:
> We must first of all find - WHAT TO CHANGE... We must [then] clarify to ourselves - TO WHAT TO CHANGE TO...
Yeah. One of my programming mentors once wrote, that, when you are given a programming problem to solve: DON'T IMMEDIATELY START CODING - first, analyze the old requirements and the new requirements.
@20:48 Simple solutions
> People are always trying to look at complicated things instead of just understand[ing] the basic things in a really clear and thorough way and then build[ing] on that. So they try to build up all this complexity, but it's too shoddy. Their foundation isn't solid enough and then they put the next layer [on] and it's not solid enough... and they keep trying to build up all this complexity. They think they need 10 layers to make this advanced, sophisticated thing. But it'd be better if they just had 3 layers and they got them right with no cracks, no holes. Or at least -- it's never going to be perfect, but you can definitely build to a higher standard of no known errors or no known significant errors.
This reminds me of learning math. I enjoy doing stuff that's technically pre-algebra (like Art of Problem Solving's "Prealgebra" book [
https://artofproblemsolving.com/store/item/prealgebra]. It's about using middle-school math, but not in a rote way, but in a creative way to solve middle school math contest problems.
There are lots of areas in my life, for instance regarding discussions, in which I try to do complex stuff without understanding the simple stuff well first. For example, I still don't understand basic grammar thoroughly. I've never read through a grammar book and tried to understand it well the way I have with math books.
> The concept of significant errors is interesting. What you have to do with each error is actually understand what the consequences are and then decide if its OK. And if you decide its OK it's not really actually an error, it's just a design feature that's within the tolerances allowable.
Interesting concept from curi that the standard for judging if a decision is an error is to look at its consequences. (Probably its reasonably expected consequences.) If the consequences of a decision are OK, then the decision itself is not an error.
Aside: there's a typo in the book that curi didn't comment on. Goldratt wrote:
> Complicated solutions don't work, simple one's might.
The plural of "one" is "ones", with no apostrophe.
22:30
> If the first two questions of WHAT TO CHANGE? and TO WHAT TO CHANGE TO? are considered to be technical questions, then the last one, HOW TO CAUSE THE CHANGE? is definitely a psychological one.
I think the context Goldratt has in mind here is improving the performance of a system that involves people. Goldratt's advice might apply to the case of self-improvement in that, if you're the typical broken person, changing will involve psychology -- your own.
@24:00 office politics
Static memes are similar between organizations. If you are familiar with the office politics etc of other organizations you have a head start in understanding a new organization's psychology.
@27:00 Any change is a perceived threat to security.
Maybe this applies to self-improvement as well. For the typical person trying to change themselves, there will generally be some part of them that is threatened by the change. If there wasn't this kind of resistance, they might well have made the change already.
@36:00 Being behind on an exponential growth curve
When you're making exponential progress, being behind a competitor by just one year means a continually increasing gap between you and your competitor in absolute terms. Say you both improve at 20% per year, but you waited a year. After 15 years you're behind by 3 units, and at 20 years you're behind by 8 units (
http://www.wolframalpha.com/input/?i=graph%201.2%5E(x%2B1)%20-%201.2%5Ex,%200%3Cx%3C20)
@44:28 Odd phrasing: "Come with a change"
Goldratt writes:
> Almost all of us have come, at one time or another, with an idea of how to improve something.
The phrase "come with an idea" is non-standard, but it has 6.5 million results on Google:
https://www.google.com/search?q=%22come+with+an+idea%22
That's too common to be a typo. Maybe "come with a <foo>" is from a dialect of English like Indian English or Israeli English.
@44:57 resisting change
> By the way, Goldratt himself has talked elsewhere -- he said that people don't exactly *resist* change, because there's some changes that they embrace. Like, getting married is a big change, or having a kid is a big change, but people often want those changes and seek out that change. So, just saying "people resist change, they're conservative" is not a good explanation. They resist *some* changes and they embrace some other changes. They like some changes and they don't want some changes. So, there's actually a judgment call involved, rather than just "they hate change itself". No one hates change itself. No one's against every single change.
Right. There's also a phrase for this flawed concept: "change-averse". But as you say, people don't resist change, they resist changes they don't think are worth it.
> So if someone's against YOUR proposed change, there's a reason for it other than the fact that it's a change. There's something about it that they think might be bad. They might not have thought it through super-well, it might not be the most logical resistance, but it's not just straight "I hate change". There's more to it than that.
Good point - needs to be said. I find it especially tempting to apply the "change-averse" label to others who resist not abstract changes, but changes *I* proposed. But curi's reasoning above still applies.
@56:04 disliking criticism
Goldratt writes:
>> Most of us don't like to be criticized. Vicious criticism, we like even less. But there is one thing we simply cannot tolerate--constructive criticism. Constructive criticism means that there is something correct in the criticism. It's under the belt. It's unfair. We are certainly going to hold it against the one who criticized us for a long time.
curi:
> There are a lot of people for whom, at some point in their life, there was a criticism they liked receiving. I've had discussions with people. I've had someone say to me, "I don't like criticism.", and I say back to them, "But I said several criticisms of you like half an hour ago, earlier in this discussion." And I go copy/paste them and quote the criticisms to the person. "Here, I said this criticism, I said this criticism. And your reaction was positive. You didn't say `fuck you'. You didn't get mad. You totally liked those comments. And now you're telling me you don't like criticism." And their response was basically, "Oh, I didn't think those counted because I liked them." People like criticism reasonably often. They just don't associate the criticisms that they liked, that they found reasonable, with the criticisms they don't like, the ones they found mean or threatening or unreasonable.
> So, I think most people, they like some criticisms some times. There's different factors that go into it. It's not just how reasonable is the criticism. It's also how reasonable is the person who's receiving the criticism about this particular thing. Because people, they're unreasonable at some things, there's some things where you could give a really nice criticism that's really useful, it's explained really well, and they're not going to like it just because they're so unreasonable about this particular topic. There's some thought processes in their mind about this topic that are not rational and things go wrong and even though the criticism is really good, it doesn't work for them.
> But the same person, if it's a totally different topic, there's something they're good at in life, and that they could take some criticism about. Often, it's whatever they're good at is what they can take criticism about, because that's how they got good. It's the area of life that they were willing to listen to some criticism; that's how they got better at it. So now what happens later on is you see the thing they're good at is the thing they take criticism about; it goes together. One causes the other.
> So you get like a good sports player and, maybe they're not the most rational guy most of the time, but then you talk to them about sports strategy and they get really rational. Or you talk to them about practice routines and they get really rational or something. Because that's how they ended up being good at sports, is because they were actually pretty rational about it. It doesn't work every time but there's some themes there.
> And for the less rational sports players, that's part of what coaches are for, is the coach has to figure out a way to talk to this guy so that they can actually take criticism and make some changes so they can fit the team's strategy better. And they do have to be rational in some way, like there's some way to reach them. They don't just hate criticism no matter what. There's something the coach can say that will get through to them. Otherwise they wouldn't actually be able to be on the team. There has to be some way to talk to them and get them to make some changes, for them to actually be valuable to a sports team.
> So, for some of the people it's easier. The coach doesn't have to try so far, and they're pretty rational, and that's why they're good at sports. And then there's some people where talking to them can be pretty hard, but the coach does have a way of doing it. And they don't just sit there going, "Oh, coach gave me constructive criticism, I'm going to hold a grudge for a long time, it's so unfair." There has to be some sort of better attitude than that, where things can go more smoothly than that.
> So I think the world's not as grim as Goldratt is talking about.
Yeah. Goldratt is making the same mistake about people's reaction to criticism that he warned people not to make about people's reaction to proposed changes.
@1:24:00 The Goal as literature
Goldratt wrote:
> The Goal is certainly not an exceptionally good piece of literature.
curi gives a nice way of evaluating this claim: Would you read the goal if it didn't have the business content? Just for the story and the characters? Probably not. As curi says, you'd rather read Brandon Sanderson.