Planning Poker is an estimation technique in Scrum that promotes consensus- based effort estimations. Team members make estimates by playing numbered cards face-down, instead of speaking them out loud, to avoid cognitive bias. Once everybody has chosen a card, the cards are revealed and the estimates discussed, with high and low estimators explaining their reasoning. The process is repeated until a consensus is reached. This method is highly collaborative and every perspective is considered, resulting in more accurate estimations.

Affinity Estimating is a technique where a large amount of user stories are roughly estimated in a short time. The team sorts the user stories into categories based on their relative size. They can, for instance, be sorted as small, medium, large, or as per T-Shirt sizes (S, M, L). The discussions that emerge during the categorization process bring up different perspectives and lead to a mutual understanding of the stories.

T-Shirt Sizes is a technique where user stories are estimated relative to one another, in terms of Small (S), Medium (M), Large (L) and so on. It is an easy and quick way to estimate large backlogs, minimizing the time spent on the estimation process. Although this method is less precise, its speed can be very beneficial for long-term road mapping and project perspective.

Ideal Time is an estimation technique that measures how long a task would take to complete if there were no interruptions or distractions. This measure helps the team forecast the work that can be taken into an iteration. However, it's necessary to remember that in the real world, a multitude of factors can impact work, hence it's only an 'ideal' and not the actual completion time.

The Bucket System is another approach where the items (typically user stories) are sorted by size and placed in 'buckets'. These buckets represent different ranges of sizes or levels of effort. Each bucket is given a numerical value according to the size or level of effort it represents. This method can be fairly quick and allow a high level overview of the overall size of the product backlog.

In a typical Scrum setup, the development team works in iterations known as 'sprints'. The 'velocity' of a team refers to the amount of work they can complete in a single sprint. After a few sprints, an average velocity can be calculated to estimate future performance. However, an important caveat is to consider that velocity isn't a perfect indicator because empirical evidence might suggest that conditions change over time (e.g., team's composition or morale). At best, it serves as one of the many useful tools in a Scrum Master's toolbelt that can provide a kind of 'weather report' for the future.

The Wideband Delphi technique encourages anonymous, independent estimation by each team member. The process typically involves iteration until a consensus has been reached. Each estimator independently estimates then all estimates are revealed at once. Discussion follows where information is exchanged about each estimate. The facilitator might provide an average for guidance but doesn’t provide the ‘right’ answer. The iterations can continue until consensus is achieved, though it isn't enforced. This can be incredibly valuable as it manages cognitive biases like anchoring, where early estimates can unduly influence future ones.

Relative Mass Valuation is a technique to decide on story point scores for user stories. The team sorts user stories into piles based on their relative sizes, with no two stories in one pile. After sorting, story-points are assigned to each user story according to the pile in which it lives. Caution must be exercised to not treat the points as concrete values, but rather as a conversation starter around complexity and uncertainty.

Named after the renowned gambling mecca, this technique uses statistical modelling to simulate potential outcomes based on the variability of a task's estimation. It’s used when historical data is available and forecasted using random sampling. Although more complex than some other Scrum estimation techniques, it allows us to answer 'When will it be done?' Under certain conditions, or rather 'When will all of it be done?' given the complexity and variability of software development work.

Parametric estimating is a technique that uses statistical modelling to develop a cost estimate. It uses a set of parameters - factors that are likely to impact the estimate, such as the size of the product, the complexity of the work, the skills of the team and so forth - and uses these to develop a mathematical model. From this, a range of estimates can be generated. The strength of the parametric estimating technique is that it both recognizes and quantifies the inherent uncertainty in any estimate.