If you assigned 1 story point to every story, how much slicing would you do before you were confident you could actually complete each story in a day or so? Congratulations! You just discovered #NoEstimates. Stop trying to fit the estimate to the story. Start slicing the stories.
NoEstimates retweeted
It's a well know fact that meteorology (science of weather) can scarcely forecast more than 5 days of weather ahead. Yet, software projects (which are also affected by the weather!) pretend they can plan in detail several months ahead. #noestimates
NoEstimates retweeted
"It's hard to predict what this dice will roll."
"Well, then, you need to get better at predicting."
Conventional wisdom about software development processes - that unpredictable work needs better estimates, not a more predictable way of working.
Guaranteed it won't be anyone with "manager" or "director" in their title.
A clear example why Estimates are useless at best, and extremely unfair at worst.
Guess who gets blamed for these situations?
#noestimates
A common criticism of #NoEstimates is that the term is misleading. It's not without merit. Given its emphasis on empirical evidence over expert opinion here are just a few terms that would be more accurate:
Evidence-Based Estimation
Performance-Driven Planning
Adaptive Planning
NoEstimates retweeted
Any management that wants you to give them predictions so they can plan is living in la-la land. That's the opposite of agility. Slapping agile-ish words on waterfall does not give you agility.
NoEstimates retweeted
Why do ppl assume that estimates are necessary? I don't see the need for them myself. They're solid waterfall thinking—big upfront plans, milestones, months-long delivery cycles. Not much agility. It's deployable right now. It will be deployable w/ this new capability tomorrow.
NoEstimates retweeted
If you’ve established a pretty high level of predictability, then estimates probably aren’t all that harmful. When you have a lot of variability, they can be devastating.
The #ProEstimates crowd love to point to other professions and say "See, you're just being silly. You wouldn't do X without an estimate." Ironically there's usually a mountain of evidence that X has the same or worse problems with estimation accuracy as software development.
NoEstimates retweeted
I've yet to see a product that was improved because of estimates.
I've yet to see a product that was delivered sooner because of estimates.
The requirements don't care about your deadlines. The code doesn't give a damn about your estimates. You can strong arm, cajole and coerce developers all you want and the software won't care. It will take as long as it takes.
NoEstimates retweeted
Is your refinement meeting taking too long? Are you spending most of the time in estimation discussions? Here's a tip, ask only: "can one person complete this story in about 1 day of work?" If the answer is "no", just ask the team to split the story! #NoEstimates
OH old mgr explaining budget to a new mgr...
Old: We use ROMs for the len of each project a year in advance
New: Wait, ROMs?
Old: Yes, then we calc cost as Len * std salary of assigned resources
New: std? Not actual?
Old: No.
New: Nothing about this makes sense
Old: ???
The #ProEstimates crowd insinuate we think we're somehow special.
We're not but software is. Not because it's inherently less predictable, but because it's more malleable and more divisible.
The smaller the time horizon the less estimates matter. So break down requirements 1/2