LeanAgile
Coaching and Training
Other

 Also some Useful Tools
Note that we have some Incomplete Pages That Require Work on this site:)
Note that we have some Incomplete Pages That Require Work on this site:)
You know “collaboration over contract negotiation”, right? Metrics often drive a wedge between management/stakeholders and the team, none more so than forecasting metrics. However, when you give a probability distribution as the answer to the question, “How long will it take?” instead of a single date, an amazing transformation happens. Suddenly, the team and management/stakeholders start collaborating to manage tradeoffs and risk. So, how do you generate a probabilistic forecast?
Maybe you've heard of Monte Carlo simulation. Maybe you've seen probabilistic forecasting techniques demonstrated or even used. But you just don't understand how it works. This talk is a gradual introduction of these techniques. You need to know nothing about combinations and permutations. You don't need to know how to apply complicated formulas. You need only have the ability to understand the rules to a simple strategy game.
This talk starts off explaining the simplest form of probabilistic forecasting using throughput/velocity as an example that anyone can follow. We'll then layer on more sophistication (but no complicated math) and discuss the tradeoffs of each approach along the way. In the end you'll have everything you need to understand and make use of probabilistic forecasting.
Learning Outcomes: Easily understand Monte Carlo simulation Do whatif analysis Take explicit risks into account Utilize forecasts that are presented as a probability distribution rather than a single date Understand the advantages of using this approach
larrymaccheroneprobabilisticdecisionmaking.pdf
@lmaccherone AgileCraft
Probabilistic decision making as well
4th down conversion Often made wrong Fear  means make They should go for it Bellman equation Some coaches do it better  bill belacheck
Bias eats good decisions for breakfast Probabilistic decision making to overcome bias
Wednesday 34  how get people to change based on analysis
“Every decision is a forecast” Outcome a has better outcome for company than b, c, d
Quality of decision Alternatives considered Models used for forecast
Probabilistic models are superior
Value of alternative = probability of good thing happening x value of good thing happening
$1M to invest
Worst case (25%), Likely case 50%, Best case 25%
Strategy 1: Strategy 2:
What if you have to choose only one project. Individual motivation is conservative  so don't do risky thing
How many projects do you need to determine which which project you should do
See code
Http://jsfiddle.net/lmachherone/j3wh61r7/
Library called “luminize”
Break even between between 2 and 3. For strategy 1 is better than strategy 2
Katan  strategy game  find out about this
Emotion and bias plays part Scary negative Question the huge positive
How do you trust the qualitative inputs
Argument is about who is right Decision making is about what is right
Now can trust this
Douglas Hubbard Getting probability input you can trust
Equivalent bet calibration Even pretending to bet money works.
Http://maccherone.com/luminize.com Or luminize.com
Crossing burnup chart Scope fixed Where crossed is normal “release date” But tossing out a lot of information
Put frequency chart on top of release burnup chart
Let's start over 50% that risk delay by 7 to 8 weeks Gives bimodel curve
Then perhaps smaller risk but multiple Relates to spread out version
Later data points are more useful than Last n days are optimal
Mark off chains. Find patterns
Troy Macguiness Finite element analysis Look at each work item as it goes through the system Kanban board
Look at places where queues are really large
Eg typically not add developers, but rather testers etc
Using measurement in agile environment
Don't take measurement to the dark side Here be dragons  not 7 deadly sins anymore
Trick is to slay the dragons
Dragon 1 Manipulating 1 Once they know this is what is being measuring they will game it Need to create culture where people want to improve Use metrics to drive the improve
Dragon 5 Using Convenient metric
Measure, insight, decision, outcome
Actually want to do this Decide on outcome do you want Think what decision needs to be made Gives insight
Coaching basketball Under basket, and 3 point Optimize team result  not individual
Come Wednesday at 3:45pm session
Look at agilecraft  Portfolio and project tool over alter alternates