text stringlengths 0 2.35k |
|---|
**Jerod Santo:** Somehow I'm here. |
**Natalie Pistunovich:** Hi there. |
**Mat Ryer:** Hello. |
**Jerod Santo:** Hello. |
**Natalie Pistunovich:** Are you all enjoying the longest day of the year in this side of the hemisphere? |
**Mat Ryer:** Yes. I'm struggling on it a bit now... But yeah, it's been good. It's a bit too long for me. What about you, Jerod? |
**Jerod Santo:** The longest day of the year - I feel like there's a velocity metaphor in there somewhere... |
**Natalie Pistunovich:** Or the shortest day. It depends on which hemisphere you are. |
**Mat Ryer:** Yeah, that's a good point. |
**Jerod Santo:** Right. If you have sprints that last one day, this is the best day of the year to work. |
**Mat Ryer:** Yeah, you get a lot done. |
**Jerod Santo:** But what is velocity? I don't know. |
**Natalie Pistunovich:** What is the velocity? |
**Jerod Santo:** Is it speed? |
**Mat Ryer:** Yeah, that's a good question. I suppose really it's about generally how quickly as a dev team you're delivering things. But obviously, there's been attempts to measure and then use that to predict, and I think there's been some work better than others, but I think it's definitely an interesting subject, a... |
**Jerod Santo:** \[04:11\] There's lots of ways that people have tried to measure this right? And predict it even. |
**Mat Ryer:** Yeah. |
**Jerod Santo:** Or use predictions in order to determine it. You've done some velocity tracking, haven't you, Mat, with pace.dev? |
**Mat Ryer:** Yeah, well, with velocity, but like the Agile practice of velocity. That practice actually has some merit. So for anyone that doesn't know what that is, you basically try and assign some kind of abstract score to work. And the point isn't really to get it right from a time perspective or anything; we can ... |
So say you got 200 of these abstract points done in two weeks, or three weeks, then you know that in the next three weeks probably you're gonna do another 200. So then you can look at your sized work and spend those 200 points, wherever that is, selecting the work that you think is the most important. And then if you g... |
**Natalie Pistunovich:** That brings up two questions, at least. One is do you measure things like how many points you spend on writing emails, and things like that? And the second one is, how is this different from Agile? |
**Mat Ryer:** Well, to answer those in reverse, this is really the Agile practice; one of the Agile practices is that, really. This is one of the things you can do to measure this stuff. You mentioned about the email and stuff, the other work that you need, and what about tech debt, and other fixes and other meetings a... |
**Natalie Pistunovich:** So doing things like also planning your use of points in the same category as emails. |
**Mat Ryer:** Yeah, so the planning stuff is interesting, because if you've done it -- I did work on a team once where we consistently did 200 points every two weeks. And they're called points because there is -- one of the important things is that this is an abstraction. So you're not timing things; you're not saying ... |
So it's a terrible time to try and guess how long something's gonna take, isn't it? |
**Jerod Santo:** It's the worst possible time. |
**Mat Ryer:** Yeah. It's much easier at the end, when you've done it. |
**Jerod Santo:** \[07:51\] Right. So retrospectives are nice. But the pointing sessions for me is the Achilles heel of this entire concept, because it's never right. And not only is it always wrong, but it's inconsistently wrong. Now, this has been my experience. I've heard it back to me as well, so I'm not the only on... |
So I'm kind of -- I get frustrated, because I feel like the numbering is so fraught. |
**Mat Ryer:** Well, one of the practices that I've seen work is they do basically the card roulette thing. This is where you sit around and you talk about the issue or the work, anyone that's got something to say chips in, and kind of paint a picture, an approximate picture... And then you say, "Okay, 3, 2, 1 - everybo... |
And honestly, the numbers almost aren't even the valuable bit. It's when everyone on the team is like "That's a three", and then somebody will hold up a 13 and you're like "What do you mean 13?" and they'll know something that you didn't know and nobody else knew. And that's a great way to kind of just get that out and... |
And often, interests align along those same kinds of lines as well, so they're quite good for identifying who might be interested in doing particular work, or taking on/tackling particular problems. But it's definitely not perfect. |
**Jerod Santo:** I can see where that might be nice. So you average those numbers out, or you just look for the outliers and say, "Hold on a second... Let's talk more about this"? |
**Mat Ryer:** Yeah. So you get a consensus really, at the end, hopefully... And you may compromise a bit. Sometimes you say, "Okay, we'll go to a five then." Because it's possible that that person's just wrong, and everyone's like "No, that's not a problem. That's fine." Or "We've had that before and it wasn't a big de... |
**Jerod Santo:** Can you say 13? Is that a number that you've given, or you just made that up to show the drama? |
**Mat Ryer:** Oh, no, no. |
**Natalie Pistunovich:** It has to be Fibonacci, so it's fine. |
**Jerod Santo:** Why does that have to be Fibonacci? |
**Natalie Pistunovich:** We don't make the rules... |
**Jerod Santo:** \[laughs\] We just follow the rules. |
**Natalie Pistunovich:** But actually, this conversation makes me think of a book that I'm listening to right now... It's called Noise, by Daniel Kahneman. And this is the same person who -- he is, I think, a Nobel Prize winner for behavioral economics. And there's a lot of conversation biases, and he, I think, is the ... |
So Noise is a new concept that they are introducing, let's say, that - yes, there is bias, but bias is what you said, Jerod, in the beginning - it's consistent. So you said that the numbers are not always off, but they're like randomly off. And this is exactly how the book starts, with saying that there are - let's thi... |
You know that research that they did, that judges on average would give easier sentences when it's the beginning of the day and after lunch, and harsher sentences right before lunch. This is an example of external noise. Things you cannot measure, you cannot predict. It's not always the same, it's inconsistent and so o... |
\[12:21\] And it's very interesting to think of all this planning and all this velocity in the light of noise. Like, do you always do your sprint meetings at the same time? Does everybody come after lunch? Or if you have a team around the world, then you'll be right before lunch, I'll be right before dinner, dying to c... |
**Jerod Santo:** Yeah. I mean, that might be a good thing, because you won't have consistent bias, or consistent noise. You'd have people in different states of life... I think the judge example is -- |
**Natalie Pistunovich:** After lunch you'll be always mellow, like "Yeah, we can do this. It's okay. It's not too much." And I'll be always like "Let's just call it a day. Seven. Fine. Seven. Get this out of the way." |
**Jerod Santo:** Right. So you guys balance each other out. I don't know, that's interesting. How do you account for noise in these planning meetings...? |
**Mat Ryer:** Yeah. So that external noise is interesting. And actually, in these planning sessions, anything that is just outside of what we know about that particular work, what that particular item is, is kind of ignored. So in a way this is very simplistic for that reason. But it has the benefit of kind of just lik... |
But the Fibonacci is interesting, if you think about the sequence. You end up with meaningful gaps. Because if you just have a scale of 1 to 10, what would be the difference really between a six and a seven? That's a quite a difficult thing. Instinctively, I think it's a no. And the idea is you want everyone to be some... |
So update a URL - that's definitely a one, because that's the simplest task you can do, is probably go and hopefully find the URL and update it. Famous last words, of course, because that's exactly the point, is then you realize, "Oh, the URL is made up from--" |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.