Certified ScrumMaster (CSM) Planning, Estimation, and Tracking Progress Syllabus

What is Planning Poker?

A consensus-based, relative estimation technique in which team members simultaneously reveal cards representing their estimate for a backlog item, then discuss divergences and re-vote until they converge. It harnesses group wisdom while reducing anchoring bias.

Why do Planning Poker participants reveal their cards simultaneously?

To prevent anchoring and bandwagon bias, no one is influenced by hearing or seeing others' estimates first, so each person's independent judgment surfaces and disagreements can be discussed.

What card sequence is most commonly used in Planning Poker, and why?

A modified Fibonacci-like sequence (e.g., 0, 1, 2, 3, 5, 8, 13, 20, 40, 100). The widening gaps reflect that larger items carry greater uncertainty, discouraging false precision for big stories.

In Planning Poker, what happens when estimates diverge widely (e.g., one person says 3, another says 13)?

The highest and lowest estimators explain their reasoning to surface hidden assumptions or missing information, then the team re-votes. This discussion is often the most valuable output, not the number itself.

What special Planning Poker cards exist besides numbers, and what do they mean?

A '?' card (the estimator lacks enough understanding to estimate), a '0' (trivial/already done), an infinity/large card (too big, must be split), and sometimes a coffee-cup card (request a break).

Who provides the estimates in Planning Poker, and what are the roles of the PO and Scrum Master?

The Developers (the people who will do the work) provide the estimates. The Product Owner clarifies items but does not estimate, and the Scrum Master facilitates but does not impose estimates.

What is Affinity Estimation (affinity grouping)?

A fast estimation method where the team silently sorts many backlog items along a spectrum from small to large (or into size buckets) by similarity, then assigns point/size values to the groups. It is well suited to estimating a large number of items quickly.

What is the 'bucket system' of estimation?

A scalable technique where predefined buckets (e.g., 0,1,2,3,5,8,13,20,...) are laid out and items are placed into the appropriate bucket through divide-and-conquer discussion, allowing hundreds of items to be sized rapidly.

What is dot voting as used in estimation or prioritization?

A technique where participants distribute a limited number of dots/votes across items to express relative size or priority; totals reveal group consensus. It is quick but coarse.

What is the difference between an estimate and a forecast?

An estimate is an approximation of the size or effort of a specific item or piece of work; a forecast is a probabilistic projection of future outcomes (e.g., what will be done by a date) derived from estimates plus empirical data like velocity. Estimates are inputs; forecasts are outputs.

Why does Scrum favor the word 'forecast' over 'commitment' for sprint scope?

Because work is done under uncertainty; the Sprint Backlog is the Developers' best prediction of what they can complete, not a guaranteed contract. The 2011 Scrum Guide deliberately changed 'commitment' to 'forecast' to reflect empirical reality.

What does it mean that Agile estimates are probabilistic, not deterministic?

They express likelihood rather than certainty; completion can be stated as a range or confidence level (e.g., '80% chance done by sprint 6') because variability and unknowns make any single exact date unreliable.

What is the cone of uncertainty in estimation?

A model showing that estimates are least accurate at a project's start and progressively narrow (become more accurate) as work proceeds and knowledge increases. It justifies re-forecasting frequently rather than locking early estimates.

What is velocity in Scrum?

The amount of work, measured in story points (or items) of completed, 'Done' Product Backlog items, that a team finishes per sprint. It is an empirical, team-specific measure used to forecast future capacity.

How is velocity calculated for a sprint?

Velocity = sum of the estimated points of all backlog items that fully meet the Definition of Done within the sprint. Partially completed items count zero; only fully finished work is included.

How is average velocity typically determined, and why use an average?

By averaging the velocities of the last several completed sprints (often the last 3 or more). Averaging smooths sprint-to-sprint variation, giving a more reliable basis for forecasting than any single sprint.

Why should velocity NOT be used to compare two different teams?

Because story points are relative and calibrated independently by each team, a 5-point story for one team is not the same as a 5 for another. Velocity is meaningful only for forecasting within the same team.

Why is velocity considered a planning/forecasting tool rather than a performance metric?

It measures a team's empirical delivery rate to forecast capacity, not how 'good' the team is. Using it as a performance target invites point inflation and gaming, undermining its forecasting value.

Should partially completed work be counted in velocity? Why or why not?

No. Only items meeting the Definition of Done count, because Scrum values working, potentially releasable increments. Counting partial work would inflate velocity and produce unreliable forecasts.

What is a Sprint Burndown chart and what does it track?

A chart that plots the remaining work (typically Sprint Backlog effort in hours or points) on the vertical axis against time/days in the sprint on the horizontal axis, trending downward toward zero as work is completed. It tracks progress within a single sprint.

On a burndown chart, what do the axes and the ideal line represent?

X-axis = time (days of the sprint or sprints of a release); Y-axis = remaining work (hours or story points). The ideal line is a straight diagonal from total work to zero; the actual line shows real remaining work, and being above the ideal line signals being behind.

What is a Release Burndown chart?

A chart tracking remaining work (story points) across multiple sprints toward a release, with the Y-axis as remaining backlog points and X-axis as sprints. It shows whether the team is on pace to finish the release scope by the target sprint.

What is a Burnup chart and how does it differ from a burndown chart?

A burnup chart plots completed work rising over time toward a total-scope line. Unlike a burndown (which shows only remaining work), a burnup separates two lines, work completed and total scope, so scope changes are visible distinctly from progress.

What key advantage does a burnup chart have over a burndown chart?

It makes scope changes explicit: because the total-scope line is drawn separately, added or removed work is clearly visible, whereas on a burndown a scope increase can be mistaken for lack of progress (a flat or rising remaining line).

On a burndown chart, what does a flat (horizontal) line indicate?

No net reduction in remaining work, either no work was completed, or completed work was offset by newly added/discovered scope. It signals a stall or scope growth that should be investigated.

What does it mean if the actual line on a sprint burndown is below the ideal line?

The team is ahead of schedule, less work remains than the ideal trajectory predicts at that point, indicating it may finish early or could take on more work.

How can a trend line on a release burndown/burnup be used to forecast completion?

By extending the slope of completed work (or remaining-work reduction) as a trend line, you can project the sprint/date at which work will reach zero (burndown) or meet the scope line (burnup), giving a data-driven completion forecast.

What is a Cumulative Flow Diagram (CFD) and what does it reveal?

A stacked area chart showing the number of work items in each workflow state over time (e.g., To Do, In Progress, Done). The band widths reveal work-in-progress levels and bottlenecks, and the horizontal gap between bands approximates lead/cycle time.

State Little's Law as a forecasting tool and what it relates.

Little's Law: Average Cycle Time = Average Work In Progress (WIP) divided by Average Throughput. It relates how long items take to how much work is in progress and the completion rate, implying that reducing WIP shortens cycle time.

What is empirical progress measurement in Scrum, and what counts as real progress?

Measuring progress based on observed, actual outcomes, specifically working, 'Done' increments delivered, rather than on plans, percent-complete estimates, or effort spent. Empiricism rests on transparency, inspection, and adaptation, so only validated working software signals true progress.