Friday 26 June 2015

My thoughts on "Agile for Humans" podcast

I recently listened to a podcast on the topic of #NoEstimates on the "Agile For Humans" podcast (follow the link for the podcast and the participants).

Overall, I think it was a great podcast! Some good points were raised and the discussions were interesting. It left me with some thoughts I'd like to share here. And I took some quick notes while listening, so I'll share them here.

First of all, on the topic of estimates, I think the discussions were very developer-centered. And there's nothing wrong or strange with that, since the participants are all software developer coaches and similar. But when it comes to the need for and use of estimates, it's not just about the software developers. So, I would have liked to hear a discussion more from a "business" perspective. That's the overall feeling I got.

There was a discussion about the name/hashtag "no estimates". The name was practically praised. Sure, it was briefly mentioned that the name was "controversial", etc. But I missed hearing a discussion of all the "bad" things that come out of it, as I see it: the binary thinking, false dichotomies, Sturgeon's law, that estimates are lies, waste, etc. Things that I constantly see on the topic. Little to none of that was covered. I would have liked to hear some thoughts about how this could be addressed.

And, sure enough. JB Rainsberger soon started to talk about that we were "brainwashed" (yes, he used that word). I actually think that's what the name/hashtag does to you - words matter! From here, the rest of the podcast kind of lost a lot of credibility to me, because no one confronted this. Why was this such a big turnoff for me? I see it like this:

Why do we use the word "brainwashed" and estimates in the same sentence? That's probably the most interesting question of all. Why do we see it like that? Am I really brainwashed with estimates? I suppose if I were, I wouldn't know, right? But am I brainwashed with using seat belts as well? Tooth brushing? Using knives? Salaries? Math? Eating? Does that sound silly? Yes it does. But why does no one see it as silly when calling use of estimates "brainwashing"? I hear you say "It's not the same thing". No it isn't, but I can mention hundreds, if not thousands of times per day when I estimate. Implicitly or explicitly. It can happen in the blink of an eye. Just count the number of times per day you check what time it is. Some might claim that it isn't estimating in a "software sense". Sure. But my point here is to show how natural and ubiquitous estimates are. Just like eating, knives or math. And why do we see the use of them as "brainwashed" just because it relates to building software? I wonder.

As Geert Bollen so nicely put it in a tweet:
you have me down as "pro-estimates", supposing that even means anything. Isn't it silly what this does to us?

Further thoughts. The topic of "How late were we?" was discussed. And it was mentioned that the need to know "How late were we?" wasn't really useful. I'd say yes, it's useful. It's called: learning. If I thought that something would take 10 hours but ended up at 5 or 15 (if the scope didn't change), there's learning there. I agree, there's not much use in knowing that for the specific situation, but learning for the future. "Last time I missed taking this thing into account. I won't miss that again." Or anything. If you have estimated, not learning from the outcome is ignoring the learning opportunity.

More thoughts. "How we always have done it" was discussed. It is a common fallacy that "how we've always done it" shouldn't matter. Often it is connected with "Things aren't the same now as it were back then". Sure. But perhaps we just need to learn why we use the approaches we do? Rather than revolt and redo learning. Sure, sometimes we need to rethink some practices because times change. But things are also how they are because sometimes it's been tested and proven to be a great way to do things. We can actually learn, we don't have to repeat history.

Lot of criticism there :) As I said in the beginning, despite these things, I really enjoyed listening to the podcast and it was really helpful and learning for me. Go ahead and listen to it!

2 comments:

  1. Excellent post, Henrik. Your observation about the learning nature of examining "how late were we" neatly pinpoints one common disconnect with NoEstimates advocates. It's unfortunate that very little of this came out in the podcast discussion, which was largely an anodyne chitchat among the like-minded, as is so much of the NoEstimates debate.

    There’s so much to reject about other statements made in the podcast: For example, having no clear definition of what they’re even talking about, and then expressing pride in that, because extreme statements pull in eyeballs? The hashtag allegedly promotes the conversation? No, it doesn’t. Or rather, it promotes it only among the like-minded, while critics are typically blocked immediately. Meanwhile, a podcast panelist talks casually about how estimates “brainwash” people, and no one steps forward to say hey, that’s a seriously ridiculous and extreme statement.

    Thanks for expressing your thoughts here, Henrik.

    ReplyDelete
  2. Sorry I missed this post.
    The notion "how late are we" is called "Estimate to Complete" and is a critical factor in any closed loop control system. Only by knowing that can we determine the actions and cost for those actions to get the project back to Green

    ReplyDelete