|
Original Broadcast Date : Thursday, June 9, 2011
Time: 8:30 AM PT / 10:30 AM CT / 11:30 AM ET
Speakers:
Linda Rising, Independent Consultant
Cognitive scientists tell us that we are hardwired for deception—overly optimistic about outcomes. In fact, we surely wouldn't have survived without this trait. With this built-in bias as a starting point, it's no wonder that software managers and teams almost always develop poor estimates. But that doesn't mean all is lost. We must simply accept that our estimates are optimistic guesses and continually re-evaluate as we go. Linda Rising has been part of many development projects where sincere, honest people wanted to make the best estimates possible and used “scientific” approaches to make it happen—and all for naught. In many projects, because re-estimation was regarded as an admission of failure, the team spent too much time and endless meetings trying to “get it right.” Offering examples from ordinary life—especially from the way people eat and drink—Linda demonstrates how hard it is for us to see our poor estimating skills and offers practical advice on living and working with the self-deception that is hardwired in all of us.
About the Presenters:
Linda Rising, Independent Consultant
Linda Rising has a Ph.D. from Arizona State University in the field of object-based design metrics and a background that includes university teaching and industry work in telecommunications, avionics, and tactical weapons systems. An internationally known presenter on topics related to patterns, retrospectives, and the change process, Linda is the author of Design Patterns in Communications, The Pattern Almanac 2000: A Patterns Handbook, and Fearless Change: Patterns for Introducing New Ideas (with Mary Lynn Manns). Find more information about Linda at lindarising.org.
Presentation Transcript:
To do that. Again, we mentioned the Haiti partners yesterday. The silent auction closes at noon today, you can pickup your items between noon and 2:30. 2:30 is when the expo closes, so don't go later than that.
Again we're glad to welcome our virtual audience today through Agile Connect. Welcome them. And I want to start with our opening Keynote today, Linda Rising.
I've known Linda for a lot of years, more than we can count. She's an independent consultant recently of Phoenix, Arizona. But she's recently moved to Nashville, Tennessee, where she's practicing saying "Y'all". She's trying to get acculturated. She's interested in patterns, the process of change but most of all these days, she is fascinated with, with your brain and that's what she is going to talk about today.
Deception and estimating, how do we fool ourselves. Please welcome, Linda Rising.
Thank you, Lee. How you all doing? Pretty good.
Yeah.
Yeah. Are you awake?
Yeah.
I believe that every conference should have a weird talk, at least one. So, I have taken it upon myself to step in to that role. So, this will be the weird talk for this conference and there will be some things that you won't like. So, let me begin by apologizing, say I'm sorry, some of these things are gonna be upsetting.
No one likes to hear about
deficiencies in their brains.
So I've also allowed a little bit of time at the end for questions.
So, as you run into those little bumps in the road where you say but
Linda, well wait a minute, I... just write that down and we'll have a little time at the end for questions and that goes for those of you who are in the virtual audience as well because I know your brains are going to struggle with this talk.
So, I wanna also apologize for not being an expert on this topic.
My PhD is in computer science and ahead of that are a lot of other degrees of various sorts but they are all technical and in that long string of degrees,
if you looked closely at the courses that I signed up for, you would only find one psychology course, which I took because I had to.
My husband says the reason you're fascinated with psychology now is that it's more technical. You can actually slide someone into an MRI machine and you can watch as their brains light up, and you can see whether they really prefer Coca-Cola or Pepsi.
You don't have to ask, you can see it for yourself.
So, as we run into problems and things that are going to disturb you, if you're not happy with that, you can just say, well maybe, maybe Linda doesn't really know what she's talking about.
If you don't get a chance to ask a question,
during the talk or after wards, you've got my email address. I love to talk about, the brain, and I promised as soon as I get home from wherever I'm traveling, to answer. So, I like to have a formal disclaimer and also there are going to be a few of you, who are really going to be interested in what I have to say, and would like a little more information.
So, unlike many of the talks, okay, this, this is the disturbing part: it begins now. Most of the talks that you have heard or are going to hear at this conference are based on experience or opinion. In our domain, we do very few controlled experiments. So, when we are enthusiastic about agile for instance. we have no proof. They have no data. We don't do controlled experiments.
We don't do a controlled balanced experiment where we take two teams, one of which is following some other process and one of which is following an Agile process and actually measure, according to some well defined success criteria, whether or not Agile, is better, than anything else. And if you ask most Agile enthusiasts, the reason they give for being an Agile enthusiast is, I just know, I feel, I believe.
Think for a moment if the drug companies, operated the same way we do, it would be something like, I, I, I tried this little blue pill for my headache, and it really worked for me, you might, try it yourself. What do you think? Try it, try it for yourself. You'll like it. We'll get everybody on little blue pills.
So, this is a talk about deception and if I'm traveling in a country where English is not the first language, you have to spend some time on this, but maybe even in Las Vegas. What does it mean? What does it mean? Deception, deception. Because we're surrounded by it here. In fact, when we talk about how your brain works, the people who designed this place were masters of deception.
I'm not going to tell you how long it took me to find the registration desk today to check out, and then find my way back. I got lost several times and I think that's intentional. Not on my part, but on the part of the designers. They're trying to lead me somewhere, where I wouldn't normally go by showing me something that's a little deceptive.
Another word that could substitute for deception would be lying. Lie. So this is the message of the talk today. Because we do that. We do that to ourselves and we do that to others. All day long. I'm sorry. And there's absolutely nothing we can do about it. We're hardwired to do this and it has survival benefit.
Otherwise, we wouldn't have it. Our ancestors also deceived themselves, were overly optimistic and if they hadn't been, they wouldn't have survived and we wouldn't be here today. So, it is a good thing. You know, one of my experiments about deception has to do with something that's been repeated many times. It's a classic experiment in psychology, two groups of people that disagree violently on a subject.
You all know I just moved from Phoenix, Arizona, so then example that comes to mind is gun control. People who have decided opinions on gun control, for or against. So, two groups of people who disagree on an issue were given a paper with data. That's numbers, and the paper was exactly the same for both groups. Both groups read the same paper about this very contentious issue. And what's surprising about this experiment is both groups said the same thing. This paper supports my point of view. How is that possible? How is that possible if they disagree and it's the same paper? How could they reach that conclusion? Any ideas?
Cherry-picking the data?
Cherry-picking the data, that's one way of saying it, yes. Yeah. Or...
Yes . They. And we're using - notice the pronoun? That we've already started using? "They"? Well those people, of course, they have deficiencies, not me. I would never do that. But those people, they, somehow distorted what they were saying. So the psychologists tell us that we carry a set of filters on our brains, all of us.
And it's unique. We each have our own little set of filters that we've been working on all our lives. So think about the filter's I'm carrying around. And everything, everything we see, everything we hear. Well that would include this talk, wouldn't it? Has to pass through those filters, so anything that's objectionable or disturbing that upsets our world view may not even make it through those filters.
So when you walked in the room today you probably thought you were a rational thinker. Let's take a poll. How many of you believe that, when you walked into the... It's okay, you can talk about that other person now. Yes, you believed that. You believed that when you walked in. Now you know better, right?
Now you are not sure, are you? I'm sorry, I'm sorry. We do it all the time in everything we do. How many of you are married? Now, was that a rational decision, for you? Did you make a little decision tree? On the one hand, or on the other hand? And in the end, the conclusion was, "yes, this is the one for me." Did you do that?
I have known some software developers who actually did- a little decision table of some kind. But I think for most of us, that is not a rational decision. And so we might, alright Linda, okay, we'll give you that one and there might be a few others where it was. Okay, it had to do with the F word, feelings.
It was an emotional decision. So here's another bad part. The cognitive scientists are telling us now that all our decisions, not just so as the little fluffy ones that we might want to concede, but all of our decisions, are made below the level of consciousness, and are not rational. Are you okay?
You can't decide. You can't decide. You can't make a rational decision. This is an interesting statistic, people who study things like this. See, my good friend Michael Mar here, is an expert on studying things and collecting data. Those people know, they understand about some of these kinds of decisions that we have to make.
So, some of those people have pointed out that when we die, if we want to buy life insurance, well, we can actually, it's a little bit like Las Vegas. We place a little bet on the table saying, no, I'll bet I'm going to die sooner. And then you'll pay me a lot of money, or actually, I won't be around, will I?
Oh, but it's for my family, it will be for my family and when we look at those tables, from all the people who studied this and we look at the expectation for our own lives, all of us tend to say, oh, but I will live about ten years longer. So we all believe we're about ten years younger than we are.
Now for someone my age, that doesn't really help much, I need a lot more than ten years, so I'll do my own deception, thank you very much. The other interesting one here that had something to do perhaps with your decision to get married was, did you look at the statistics for divorce? While waiting at the alter, did you say, well 50/50, let's keep our fingers crossed.
No, of course not. you thought those statistics are about other people. My marriage on the other hand, this is forever. I have a good friend who has been married five times. I think she said that every time. Oh no, this time. So we are hardwired for this optimistic trait that we have and it's a very good thing.
We know that we have these little filters. So, we see what we want to see. We don't even have to worry about things that we don't want to see because those little filters will actually remove them from our consciousness. But, what we are very good at, not making the rational decision, but explaining after the decision has been made, why we made it.
That's called rationalization, and we are superb at doing that. So, we can explain to anyone very clearly, we can give the logical arguments. Why, something was done. Why we made a decision on our project or in our lives. We are superb at that, and of course believing the whole time, that that's those other people who are biased, or who distort information.
We, on the other hand, see things as they are, objectively, clearly. This is even true for scientists, even scientists? Think about it. A scientist has an investment in some decision that was made at some point, has done research for years coming up with certain conclusions. What is that scientist going to do when faced with data that shows perhaps that his life's work, has been invalidated?
Is he going to jump up and down and say, oh yippee, science will make progress at my expense. I don't think so. I'm on a review committee for a lot of papers for conferences and I received a comment once that I have saved. It's a very good example of someone who was reading a scientific paper who was a scientist, who, in the review, where the conclusion was to reject the paper, said, I wouldn't believe this even if it were true.
How many of you know Ignat Summer Rise? Oh, this is one of Uncle Bob's stories. I thought perhaps after Martin jumped up and down, we'd all remember Uncle Bob's stories. Even if it symbolized what was an interesting character who observed, that somehow, when women gave birth in the hospital, attended by a physician, that the mortality rate for those unfortunate women went way up. Whereas women who delivered at home, or in a setting where they were attended by a midwife, didn't suffer from the same increase in mortality, and he wondered why. So he started collecting data and began observing, and what he saw was that the physicians were going from performing autopsies to delivering babies without washing their hands.
At the time that was accepted practice. So he began experimenting and introduced a little solution of lye and encouraged doctors to wash their hands after having had them buried deep in the innards of a dead person. Said, before you go over to deliver the babies, please wash your hands. This is an experiment I would like to measure if there is a resulting decrease in mortality.
And sure enough, as you can imagine, there was. So he collected data, objective data. And when he demonstrated to the physicians, it does make a difference. We're not sure why. They didn't understand the germ theory yet. But we can see clearly that this does make a difference. It saves lives. Wouldn't doctors care about saving lives?
Isn't that what the Hippocratic oath is all about? No one listened. No one paid any attention to his very nice data: to his experiments. Women kept on dying. Doctors kept going form performing autopsies to delivering babies for decades and decades. Ignatz died in an insane asylum. No one would listen. I like this quote from Max Flanc.
"A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die. And a new generation grows up that is familiar with it." We have to wait for the old ideas to die physically, for those people who hold those ideas to die.
How many waterfall people must pass on to the great reward before Agile takes over.
Well, if we are truly rational, then all of the experiments that have been done and here's an example, would surely show us how rational we are. So, this was an experiment that was done about hiring. In it, since it's balance and control, half of the subjects were given the positive information about the candidate, here are all the good things, followed by the things that were not so good. And then in the other group that order was reversed. No, it's the same information. It's positive first then negative. Negative first then positive. So for those of you who are creating resumes you might be paying attention at this point. Would that make a difference, do you think?
Well it must make a difference or I wouldn't be talking about it. So, what do you think? Which one? Which one had more influence? Positive first or negative first? Let's take a vote. How many vote for the positive first? What do you think? Ooh, it's going to be 50-50. Negative first. OK. How many don't care?
Two people. Thank you very much. OK. So it's interesting that it did make a difference. It's the same information. Why would it make a difference if it's the same information? If we're truly rational about our decisions, for the order in which data is presented have an effect on our ultimate decision.
And it turns out it does, because of something called framing. You can, sort of, put the brain in a mood to do something. You lead it down the pathway and that first step is critical. That's why influence strategies that are so powerful build on an initial little step. That's why, they lead you through the casino, you're already there.
So, now it's easy to just step up to that slot machine, or that poker table. So, framing sets the context for the decision and the rest, your brain just says, well, OK, well. But, the most important thing is, and has already made that decision. This experiment has been repeated many the times and what they find is that, under pressure, there is even more of a tendency to go with the positive information first, because the brain doesn't work as well under pressure.
How many of you, when asked to give estimates...Here is something like the following: "Hey take your time. No problem. If you...I know you're busy. Sometime next week, give me your best estimate." Is that how it works? Anybody have that? I want to go to that company. Well, we do have somebody. There's always somebody.
That's very good. You're an outlier. I know a company where people were locked in a room. Said you're not coming out until you can say that we can deliver this by June. No pressure. No pressure. Also no food, no bio breaks. So we know that on average, there are people who measure this. Can you imagine what kind of
field work this would be? There are three lies in ten minutes of a conversation. Let's see, how long have I been talking?
All these experiments show the same thing. A survey of high school students all thought they were above average. Let's take a little poll here.
How many of you think you are above average? Of course you are, you're at this great conference,you took the time and you invested a certain amount of energy and perhaps your own money to come to this conference. That definitely means you are above average. It turns out 80 percent of our people believe they are above average, and it doesn't matter whether they are software developers or drivers.
There's a very famous experiment that was done in, I think it was Sweden, where they asked drivers, are you above-average? And 80%, 80%, said they were above-average and that includes many of those drivers who are in hospital, having caused an accident as a result of their poor driving skills.
Since there are so many - any online dating participants here? There's such a large population now that they can study it. And it's an interesting collection of people. It's younger than the normal population. Isn't that interesting? And it's also full of women who are not as heavy as the women in a normal population.
And men who are taller. And they are all well educated and they love extreme sports. I wonder why that is?
We even do this to our children. Now, tell Aunt Mary that you like the Christmas present that she made for you, and give her a big hug, and tell her you love her. Is that a bad thing?
So those of you who raised your hands when I said are you married, has your spouse ever come up to you having purchased some expensive new garment, and asks, "Do you think this makes me look fat?"
And, what did you say?
You say, oh yeah, I'll tell you tomorrow.
I want a divorce.
I want a divorce?
Don't you think there are times when a little deception is a good thing?
Perhaps that's why we inherited that tendency.
But there are times when it's okay not to be brutally honest.
"Yeah, I think you could stand to lose at least fifteen pounds. And I'll tell you where that fifteen pounds should come from." That's probably not a good idea. So, it's a good thing, this, it helps keep us together, a little deception. It's good for our social structure. Not to always be brutally honest.
That may be a... keeps us from unraveling. We do this with everything we eat and drink. Lots of studies to show that we're not very good at estimating how much we've had to drink. Well maybe that's not a surprise to you. We have a bias for the vertical. Even professional bartenders in experiments have shown this same bias.
These two glasses hold the same amount of liquid now. But we tend to believe that the one that's taller holds more. They actually have done experiments with children to get them to drink less soda, less carbonated sugary drinks, by simply changing the size of the glass. They won't drink as much if it's in a taller glass, they believe they have the same amount.
So we over estimate the vertical. We think the vertical line here is longer than the horizontal line. So, if anybody wants to come up, and actually measure that, you're welcome to. And that explanation that has been proposed for this bias is that if we were walking across the Savanna, we'd want to have a pretty good handle on any change in height, and so we tend to look at that as though it were taller than it actually is.
We know that we eat more from things that are in large containers, and things that are closer.
So, anyone here on a diet, one of the things that you can do that is very easy that doesn't even involve a lot of change is simply buy dinnerware with smaller plates or smaller bowls and don't serve from the table .
Put those serving dishes at least three feet away
so that you're forced to stand up, and take a few steps to get a second helping.
There are lots of studies that show simply increasing the distance between you and food means that you will eat less.
And of course, your appetite detector is not very good, and it's slow.
So you won't even notice the difference.
One thing that we do know about, well, most the people in this room, I think... do you spend most of your time sitting? In front of a computer, or sitting in a meeting room, or sitting, listening to somebody talk. We spend most of our time, sitting. That we were designed to walk, we're walkers, we should be standing. we should be moving.
I have a whole talk on that. So, I don't have time to give it today, but as a result of all this sitting, we have destroyed the relationship between the hormone that regulates appetite. We get, we begin to believe we're hungry even though we haven't used up sufficient calories. Sitting has altered our hormonal interactions.
We know that names influence, how much we eat. If we order something from the menu and it's called, "Decadent Chocolate Volcano Cake to Die For", that we will not only pay more for that and we will eat more of it. But, when interviewed at the end, we will say, we enjoyed it more than if it had been called, chocolate cake.
Well, that's other people, of course.
That's not you.
Oh, what happened?
What happened on 911?
I remember being in an airport. I never got anywhere. I couldn't understand what was happening and nobody was telling anybody anything.
And we were running around and we, we could see pictures, airplanes flying into buildings.
What was that?
What was, What was going on?
Finally they said, you know, we're not going to be having any flights today.
You'd better all go home.
Go home and what we did there was watch airplanes flying into buildings Had a profound effect on how we viewed the world.
I finally got on a plane on Saturday.
Going from St. Louis to Minneapolis was a 767, which is a pretty big plane.
It 's the only time I've ever boarded and have the flight attendant say,
"there are only nine people on this flight.
Sit anywhere you want.'
Wasn't any business class, unfortunately.
Where do you suppose the nine people on that flight sat?
Where did we sit? Yes, we all sat together. Window aisle, window aisle as far forward as possible. It's the only time I've ever seen people not spread out on a flight. It's also the only time I've ever seen people talk to each other. And the question was, "Well where were you and where have you been since Tuesday?" And to a person.
The other people on that flight all said the same thing, which was, "I am never flying again." They thought this was about airplanes. "I am never getting on an airplane again."
Well what did Americans do? If they said, "I'm not getting on an airplane again", did they stop? having vacations? Driving down to the lake over the weekend?
Of course not. They just altered their transportation mechanism. Instead of an airplane, they chose to drive. They had altered their view of the risk of those modes of transportation. They now saw airplanes and - most people are willing to do that anyway, because most people are afraid to fly. So any little excuse will push that decision in the direction of "I'm not getting on a plane again."
So they began to drive. And in driving had more accidents. And in those accidents, killed more people. In fact, at the end of a year, 800 extra people died. That's three times the number killed in the four planes on 911. I mean, for those with a macabre sense of humor you might say, "Perhaps that's what the terrorists had in mind." Can you imagine what that meeting was like?
"I've got a great idea, we'll fly some planes into some buildings, and then the Americans will be afraid to fly, and they will start driving more and kill each other!" Brilliant. Brilliant. It altered our perception of risk. We did that. We did that. Those 800 lives. We did that.
So we know we're not very good at risk estimation. Ask smokers. I'm not gonna take a poll. Smokers believe that it 's OK for me, I know, I read those statistics. They're more afraid of flying than they are of dying from a very bad habit. When clearly the data show the average life expectancy reduction for smoking is five years compared to flying - a day.
When Lee asked, "What should I do with my money?" I was wondering, "Had he looked at the odds?" I won't ask you to raise your hands, how many of you play the lottery? And especially the big lotteries, you know, the ones where the giant jackpots and - look at the odds for there, I can't even how many zeros that is. We all know we're not going to die of a heart attack, so course I'll have a double cheeseburger with bacon on the side.
And a lot of that has to do with the same effect that we observed in 911. When someone wins that big lottery, they interview them, "Well tell us, how was that winning the big lottery?" "Oh, it was great, I'm so excited, I can't imagine, finally, me, I won the lottery." I think they should interview the losers.
Well, how was that losing the lottery? Well, it wasn't too great. You know, I play every week and I haven't won yet. So, I'm getting a little discouraged . Well, thank you very much. If they did that for every loser, it would take about 9 years to get through everyone. Would that make an impression on your brain?
Say, look at that, there are a lot more losers than winners. It would be a graphic demonstration. So some of this we can't avoid. Our ancestors were hardwired to fear certain things, and that's what we still fear. Anytime we see something on the television, a plane crash, that makes a big impression.
I would definitely advise going to hear of Lee's talk. I'm sure there are lessons learned there, because we do pay attention when planes go down. Whereas, an automobile accident, that doesn't even make the news. I've driven by lots of accidents and wondered when I got home, would there be something on the news about that terrible accident?
And most of the time, no, it's an ordinary occurrence. It doesn't make an impression. We don't really pay any attention. So, when this is all have to do with software estimates estimates? Well, we all tend to believe that we're better than we are. I don't know if you watch Prairie Home Companion where all the women are strong, all the men are good-looking, and all the children--way above average.
That's us. That's us. We tend to believe that we're just a little bit better than we are. In fact, I have often thought, to myself, that if I really understood what you were thinking right now, about me and this talk, I'd probably want to run back to my room and hide with the covers over my head.
But no, I don't think that at all. I believe that you all love me and you're just sitting there eating up every word I'm saying. And that's what keeps me going. We all do that. We all do that.
Found this interesting quote. A strong tendency to ignore previous feedback and persist in overconfidence in their own estimates. Persist. Persist. Well surely, I mean, I've got Michael Mall here on the front row. I was a mathematician before I became a computer scientist. Of course this was a long time ago. So that's how we're going to stay friends with Michael. This was a long time ago and I was working for a company that said, "You know, we're going to buy one of these estimation tools." As a mathematician, I had a chance to go to some training and look under the hood of that estimation tool and I want to tell you, I was blown away.
Look at the powerful mathematics we've got here. Wow! I loved it. And not only formulas, but they had data. It's okay if you don't have you're own data. It's okay because we have data. We've got data for embedded systems, we've got data for telecom, we've got data for avionics, we've got any kind of data you want. Just plug that data in and by golly out comes the estimate.
Of course it's going to cost you, but we thought it was worth it.
So, we bought one of those estimation tools and I had all kinds of fun tweaking those parameters. Doing interviews about information on the project. Cranking out the numbers. And it wasn't any better than that. You know this one? That's wet finger in the wind, 3.2. Yeah, that feels about right. Of course that was a long time ago. I think probably, I'm sure those tools are better now. I was talking to somebody recently after I had given this talk and he said,
"You know, I'm looking for some formulas. I need to, I need to have some way of justifying to my boss. I can't, I can't just do that. I got to have some justification." And I thought, "Well OK, there is a good reason for having some data or tools or something behind your guess". He said, "I just can't make it up." I said, "Why not?
That's really what we're doing. It's an estimate."
I think maybe the time has come to throw that away. Let's maybe have another word because somehow we're now enamored of that word and we think it has real meaning. It's just a guess. We hope it's our best guess, but it's just a guess.
So, I promised that if you were interested in any of these experiments or topics that I could send you information about it. So, here's an article that appeared in Harvard Business Review which I have and I will send you. Shh! Don't tell Harvard Business Review that I said that.
This was an experiment that was done with managers. Oh! How many managers in the room? Okay, good. So, there was an experiment done with managers and they were managing a little project, that was sort of a scenario driven thing. Where they said okay: we're going to see how well you do! The managers were managing the project and all of a sudden things started to go wrong.
Managers had to make some decisions, what to do when things go wrong. So, they did and some of those decisions were not very good.
Hard to imagine, I know. They got to the end, and people who were running the experiment said "Well, let's do a little retrospective. Let's look back on what happened here, and let's examine those decisions and let's see whether or not they were the best possible decisions given what you know." Were those the best things to do?
And the managers agreed; well, probably not. I think we could have done a better job. Now, that we look back on it, well, that's what we like, isn't it? At the end of each interval, whether it's a short two week delivery span, or whether it's a long release schedule, at the end, we look back and we do a retrospective.
I'm not a fan of retrospective, and we learned from that. Well, that's the intent. So what happened here? Did they debrief with the managers. Oh yeah, we could have done a better job. Let's examine those decisions and then another scenario with the same set of managers. Things began to go south on the project and what did the managers do?
made the same mistakes. Well, that's managers. Right? Oh, those managers, those managers. Not you. So an estimate is just a guess. It's based on our experience. And we should be learning from our experience. We should be doing the best job we can of learning from our experience. But, I know this is really difficult, especially for those of you who are doing Agile, and you know that we are better at shorter estimates than we are at longer ones.
But your estimates, I don't care what tool, what method, whether you stay together as a team, I don't care what you're doing. Your estimates will never be perfect. I'm sorry.
Even if your tracking them and you say, "Oh, we're getting better over time." That's a good thing.
And you should be learning but there's no guarantee that the next interval release project will be close enough to what you've just done so that your estimate will be perfect.
How many of you have heard the 'Black Swan'? It's an interesting book, it really has one point. So, I'll just tell you what it is, and I'll save you a lot of time. You can't predict the future, there it is, there are no facts about the future. There is a wonderful story in that book thought it's about the Thanksgiving turkey.
So, the turkey has plenty of data in a way not actually collecting them the numbers, but the turkey knows that the farmer comes in every day at the same time and brings food. What a nice guy? So, based on that, my estimate for tomorrow is that the farmer will come in at the same time and bring in the same delicious food.
Uh, oh, tomorrow is Thanksgiving day. The farmer comes in about the same time, but he is not bringing in any food, oh, oh. I guess we couldn't predict that. Can't do that prediction thing on in a project either. You don't know. You know, there's some belief that if we were all doing the same thing all the time, you know, for projects that are similar and the team sort of stayed together, that our estimates would be better.
It seems intuitive, doesn't it?
My husband is a developer and he works for a small company and they have exactly that. They've had essentially the same core of people now for years. And they do the same kinds of projects over and over again. And every time it's the same. He comes home at the end of the day and he says, "Well we've got another project, but we've done something like it before so I don't think it will be a problem estimating it." And he writes up the estimate and what they find in the middle of the project, is they run into surprises. Little surprises or big surprises and their estimates are always off.
So, even with a small team doing essentially the same thing over and over again. It's really not the same thing It's a myth, I think.
The other interesting thing about your brain is that when you estimate, that is the creation of something. You're making it up. You may have some tools behind you, but you're essentially making it up. And that part of your brain that's involved is different. From the part of your brain that's involved in calculation. So there's a fine line here. Not only do we need a different name for estimation because it's just a guess, but we need a different process for handling it because; once there is a 4 on the table and we know that managers love numbers.
So you give them a four, and they go crazy. And they start to do inappropriate things with that number four. They believe that since it looks like a four, it must be a real four. When in reality it's just a guess. There must be some way of either writing it in purple ink or something to say, "This is really not a four it looks like a four but it's not.
It's a guess and it's kind of like a four but not really."
Because for those of you who understand a little bit about number theory, you realize that if you take that four that is kind of a squishy four and you start doing inappropriate things to it like adding, subtracting, multiplying, dividing. Those are inappropriate operations on a squishy floor. It just doesn't work. It's a lot like, well, everybody's seen a sports game where the guys on the field had numbers on their backs.
Suppose we said, ah, well look at this team. I wonder what the average number is for that team. Oh, I see it's 46. Well, they must be pretty good. No, those numbers don't mean anything. They're just ways of identifying something. Actually, I have no idea where those numbers come from. How did they decide who is a 14 for instance, and is that a good thing, to be a 14?
So, they are just labels. They're squishy. They move if you try to add, subtract, multiply, or divide. And so we also need to be realistic about how we do these, they're different parts of the brain, they're different kinds of things. And the name for it should be different. Not only for the thing itself but for the process.
We have a lot of problems here. I see our best hope is relying on what we can see just ahead. We know that intervals now and agile development of any flavor are getting smaller and I think that is a good thing. Because trying to do an estimate for two weeks, we are going to be better. And we're going to know more and we're going to be closer to what's really going to happen.
I can remember estimates that went out for years working with the government, working with Boeing on the triple seven airplane. How did that happen? Two guys sitting over a beer and the guy in Boeing is promising the guy at United, "Oh sure, we'll have those planes ready by", and out pops a date. Where did that come from?
And then they got filtered down to those of us who were digging the transcript Yes, we say, "OK, it's right. June of... Yeah, OK, we'll do the best we can."
So, two weeks. Wow! Enormous. Enormous improvement. An incredible benefit. To not only have the shorter interval, but to have with you at that point, the stakeholders. To have transparency. To say: "Given what we just saw, with this team, and here is our back log, we think we can promise."
It's so much better than the... It was like a dance. What we used to do in the software development. It was the dance of deception. Yes, we believe, we'll have it in two years. And the person we were dancing with, of course, knows that we're not going to have it then. But they say, 'oh, thank you, we'll be happy to see that in two years.' We were both deceiving ourselves.
We were both pretending that we were going to deliver on time. So, now when someone who still follows that model said, well you know, if you're only looking two weeks ahead, how can you promise that you're going to deliver? At some point in the future and I said, well how did we do it in the past?
With the dance of deception. Oh yes we can deliver. And of course, we didn't. We were just deceiving ourselves. We lean on each other on an Agile team and we think, well, that's good, because I'll tell you what I think and you'll tell me what we think. And we'll put on those gummy bears in the middle of the table, and we'll shake it up, and that will, that will be pretty good.
So, experiments show that when you're with a group, your ability to estimate goes down. You kind of lean on the group, in a way, and say, "Well, I'm not sure how long this is will take but hey, here, Fred is going to help out and Joe will be there. So, yeah." You'll become more optimistic about your capabilities than when you are by yourself.
And what we know is that you can't just buy into a little piece of Agile. It's got to come as a package. You need to do retrospectives. You need to look back and try to do the best job you can and learning from your experience, you need to be transparent. Everyone needs to be honest about what's happening and honest about the fact that an estimate is just a guess.
Would be a good place to start.
I love this quote from Ron Jeffries, "The main thing to remember is... this is not a fact. Well, what happens isn't just what you laid out there. It'll be a lot like that. But it will be somewhat better or somewhat worse. And, once in a while, you'll get lucky and you'll do lots better. And sometimes, you might not even come close.
However it goes, the good news is that you don't have to wait until the end. In the iteration plan, every couple of weeks, you have a chance to reassess, where you are and how fast you are going. So you will have every opportunity to steer your project to success.? That's the Agile method, I could not say it any better.
It's not about perfection. You know, there are some people though, who, they just want an answer. So, I had a very good friend of mine who sent me this picture after he heard this talk. And he said, okay, Linda you have my permission to pass this on. So, if you really want a number, just ask the estimation goat.
Go to estimategoat .com and the goat will tell you. If you want to know whether you should marry that nice person you've been dating for six months, you can also ask the goat that. In fact, the goat can answer anything, anything you want to know. Just ask the goat. Now, whether or not you want to cite that source of information in your estimates or not, I leave that up to you.
But I know some of you want certainty. We all want certainty. It's a chaotic world. We don't know how to make the good decisions. We know know that for a fact. In fact I will recommend that to Lee. If he is not sure what to do with his money, he can just ask the estimation goat.
So, I know it's been a little discouraging. And you might be thinking at some point, well, I'm really not too happy with this. And Linda you've been telling me that I just don't know how to make good decisions and that my brain was kind of a hack. Suppose someone offered you a chance, said: I'll solve all your problems, you take this magic pill and all of that will go away. All that deception will disappear and you will be perfectly rational about everything.
Not only your estimates, but your religion, your politics, gun control, any issue, you will be perfectly rational. Would you take that pill? What do you think? Yes. You want the pill. Okay. Let's have a look at the very small segment of the population that is perfectly rational. They see things clearly.
Much more clearly than the rest of us. Unfortunately, most of those people are in mental institutions on serious, serious medication. They are clinically depressed. Clinically depressed people are rational. That's why they're depressed. So what you have with a deceptive brain is the ability to be happy in the face of a reality that would be depressing.
So its definitely a good thing. So hang on to your brain. Take care of it. Because its doing the best it can. Kent Beck said perfect is a verb, it's not a goal. I think that definitely applies to estimation. So I know it's been disturbing. I hope it hasn't been too bad. And I consciously tried to save about ten minutes for any questions that you might have or the virtual audience might have and I think we have people with microphones who are going to run around, so if you have questions, this would be a good time.
Any questions? We have a question down here? Okay? I can repeat it?
Where can we get the reading list that you mentioned?
Oh. Did I say reading list? No, I can recommend some books. Yeah, I can recommend some books. Actually the best book, if you just wanted to start with a book, is called "Stranger to Ourselves". It's a book by Timothy Wilson, who is a Cognitive Scientist at the University of Virginia, and it's a very good introduction to the whole area of how we make decisions, how we think, "Stranger to Ourselves".
But I have others and I would be happy to email you some other places where you can go to learn about the brain. Other questions? Got one. We need to get the mic so everyone is in. Everyone at home can hear.
So, that most effective method of helping our situation, my company where we have people who are opposing technical change is to wait for them to die of natural causes or otherwise? Is that method, sort of the number one method. Do you have any ideas what could be, perhaps, alternative number two method?
Yes . Uh, that's a good question. Actually, I teach a little class on influence, strategies, because I find that technical people. ..well I'm a technical person, and that when I wrote the book Fearless Change, someone of our reviewers said, you know why these patterns work so well, is they sit on a body of knowledge in social psychology called influence.
And that to me was an unknown area. I mean, I have heard of influence but I didn't know there was a field called, Social Psychology and I didn't know there were people who exclusively did research in that area. So, that changed my life and what I tried to do after that was, put that material together in such a way that I could hand it to you, because I'm assuming that most of you don't know, all of those things about your brain and all those things about ways to move people in a direction.
If, if Ignatz had known some influence strategies, he wouldn't have died alone in an insane asylum. He would have been able to move those physicians to be more open to his data. So, I recommend, there is a book called, Influence, it's in fact, it's in the bookstore. It's by Robert Cialdini and it's an excellent, excellent place to start.
Yeah, that would be better than just waiting for them to die. Yeah, you're right. Other questions.
Linda, I have one from the virtual audience.
Oh, the virtual.virtual
Back here. Yes, way back in this corner.
Okay, yes.
This person asks: "given that your premise is the is that the original estimate is optimistic to begin with. What's your suggestion when the project leads reject the original estimate as having too much time or resources?"
You know, I definitely believe in transparency. And I think it's always a good idea if all the stakeholders are involved in deciding what the estimate should be. And it can only ever be about the best it can be. So, your best estimate might involve more resources than you have. That's a fact of, I mean, I don't know anybody who's ever worked on a project who's said, 'hey we've got plenty of money, we've got so many people we don't know what to do with them and we've got a lot of time to deliver this functionality.
The customer is happy to wait until we are ready to deliver'. I don't know anybody, unless, what about you sir? That. We have our one out liar here. Normally there are constraints. There are always going to be constraints. There's no answer for that, no way to magically make extra people appear, sorry.
Other questions. Oh, OK. Sorry. Yep. There must be a better way to do this, okay.
Linda, we have a couple of questions from the online audience.
Okay.
Do you want to take more form here?
Sure. Okay. You say group estimating is actually worse. What about estimating techniques like using poker cards and discussions. I like poker. I think you should use whatever works. All right. And why not? And why not?
Have another one.
Sure.
Are estimates and Agile projects or bi-Agile teams better than others?
When I started introducing patterns into the company where I was working in Phoenix, Arizona, I got this question all the time about patterns. They said, well Linda, if we start using these design patterns, are they going to make our designs better? I didn't really have an answer for that. And finally I came up with one, which was, well, how good are you now?
On a scale of 1 to 10, are you a seven or a five? So, that's the unfortunate state of our industry. We have no way of measuring where we are now, let alone whether some new approach or new tool or new planning poker or gummy bears or t-shirts. Will that make us better? You have to know where you're starting in order to have a comparison to show an improvement and we don't.
So, we make decisions all the time, about what we will do and what tools we will buy and what process we will follow on essentially the word of mouth of somebody at some conference who said "Hey, Agile is great! Do it."
Did I answer that question? Or was I... Will that make us better? Yes. We don't know. It's pretty dismal, isn't it? The state of our industry. And it is not any better now than when I started decades ago. We're still going through the same. I go to presentations know and I say, "Wait, I heard this same presentation, take out that programming language or that methodology and substitute some other for it." It's the same discussion.
We are not moving forward and the only way that will happen is if we do controlled experiments and we can clearly show, we can start collecting data that say, "Hey, we have this".
Maybe Michael will lead us. Hey Michael, we need to start doing a better job of collecting data so that we know where we are now so that we can make those decisions. Not based on: Oh! that looks exciting or that new and cool or hey, I heard that or here's a famous person who wrote a book about it.
No, we'll have some clear decision making criteria because we know our brain, needs all the help it can get.
Other questions? You have been absolutely wonderful. Thank you for staying with me.
Linda, thanks so much. She has convinced me of two things. First of all, none of you knew what you were talking about when you gave me good advice, what you thought was good advice. So, while I was sitting in the back there, listening to Linda, I did go to estimategoat.com. It says bet it all. So, that's what I'm going to do. The, we have a quick break now. We're going to reconvene at 10:15.
That's when class session starts, that's when we have another keynote here. Have a good day. See you later.
Trackback(0)
Comments 
Write comment
 |
TrackBack URI for this entry
Show/hide comment form