Egad! It is the Most Likely Schedule!

Most Likely Schedule, what is it?

It is one of the questions asked more often in risk-based planning and scheduling. Project directors, managers, planners, and schedulers have to know what this term means.

Each time a scheduler tries to explain the concept, he easily gets lost in his own circumlocution. It might be his indirectness compounded by the fact that this particular term shares common but blurred boundaries with some other project and risk management terms.

In view of such difficulties, let us put their words and thoughts on paper so that project practitioners can better understand.

In project management, the Most Likely Estimate is usually an approximate calculation of future quantities (Related topic: Data Maturity, Quality and Project Integration). Estimating the duration of a project is a great example.

However, when you think a little deeper, it can likewise be an appraisal of existing quantities still awaiting definitive measurement. ML is a judgment made of the most likely value, number, and quantity, of something based on logic, source, and rationale.  Estimating the total weight of installed concrete in a given area or the tons of steel in a structure are good examples.

One has to remember that a schedule is a time estimate. It is not the exact actual time value. Experienced professionals try to provide a plausible estimate. Nobody really knows what would be the definitive value.

ML is a value that might be equal, over, short, or within the accepted tolerance of the final value. In reality, a relatively complex project will not execute precisely as planned because of risk (uncertainty) such as estimating flaws, other variables such as resources, site conditions, regulations, assumptions, scheduling constraints, and several others.

“Estimate is an approximation, prediction, or projection of a quantity based on experience and/or information available at the time. It recognizes that other pertinent information is unclear or unknown. An estimate is almost the same as an educated guess, and the cheapest and least accurate type of modeling (Dictionary, 2016) .”

We can describe it as time required to complete an activity based on usual conditions, normal resources, and historical information with no risk under consideration.

In statistic, it is the distribution mode, the time believed “most likely to happen.” It is a point of confusion because the Most Likely Schedule does not denote the mode.

Although the mode is a value representing the statistical most likely, it is a value subsequent to the iteration generated by SQRA. A very bad schedule ranged and iterated will also give a mode. The same is true with a ranged and iterated optimistic and a pessimistic schedule.

Since the Most Likely Schedule also points to the usual agreed-to schedule by the project experts, it is fair to consider it as the realistic most likely project timeline. This initial assessment is mainly from knowledgeable judgment and not through quantitative risk modelling.

Nevertheless, the developed schedule also qualifies as the Most Likely Schedule based on the chosen method of schedule development, expert judgment.

A good understanding of the three-point resulting statistics will improve one’s management aptitude dealing with the traditional project management triangle of cost, schedule, and scope (Wikipedia, 2016).

It is one of the three-point estimates which is based on the duration of the activities, given the resources typically assigned, their productivity, realistic expectations of availability for the activity dependencies on other participants and interruptions  (PMI, 2013).

Most likely date is not the same as expected date. It is but a component of the expected value. With the three point estimates (Triangular or Beta) available, one can calculate the expected duration of an activity, a group of activity, and the whole project’s duration. One can come up with the expected date.

Acronyms

BL               =          Baseline

ML              =          Most Likely

PERT          =          Program Evaluation and Review Technique

SBL             =          Schedule Baseline

SRA             =          Schedule Risk Analysis (also called SQRA)

SQRA          =          Schedule Quantitative Risk Analysis

Deterministic Schedule

The deterministic schedule is a schedule that has a singular duration estimate. The word “deterministic” alludes to a schedule with one overall duration and one start and end date. Both the overall schedule and the individual activity that comprise it did not have a probabilistic date range. If used as a baseline, the project targets a singular date each time.

Project estimation providing one duration number for each activity makes it deterministic. In other words, if the schedule has 100 activities, each activity has one estimate associated with it. If the development of the schedule used the attributes of the Most Likely duration, then this deterministic time estimate is also the Most Likely schedule.

All single duration (single estimate) schedules are deterministic. A baseline schedule is deterministic. A target and a most likely schedule are. There should be no confusion there.

In many cases, the project can be more accurate by applying a simple PERT (Program Evaluation and Review Technique) to the deterministic schedule risk model. The final time estimate is a result of an estimating technique that uses a weighted average of three numbers. It is when the project embarks on Schedule Quantitative Risk Analysis through three-point estimating.

Considering the organization’s risk appetite, and the probability of successfully achieving the end deliverables, the project is able to decide what schedule baseline to approve. If probability is too low, the project can choose to change some of its working assumptions and parameters in order to seek improvement. In some cases, it can decide to completely rely on the calculated schedule contingency.

Even in today’s age of Big Data, one of the challenges to estimators is still how to make projects more predictable. It is difficult for them to accept that they are like the weatherman who always gets it wrong; forecasting sun but instead catches rain and hail! Project estimators wanted to increase their batting average for good reasons.

Three-Point Estimates

Historical records showed that in the ‘50s, the US Navy’s Special Projects Office created a program evaluation and review technique (PERT) for complex projects .

“PERT is a project scheduling technique to analyze and represent the tasks involved in completing a given project. It incorporates activity duration variability and relies on similar concepts as the critical path method (Center, 2016).”

It measured distributed values for estimates called optimistic, most likely, and pessimistic. This is where one would trace the project management term “Most Likely” or “ML.”

For example, given the following durations:

Optimistic (O)          = 100 days

Most likely (ML)       = 160 days

Pessimistic (P)           = 250 days

 “These relative values are the subject of discussion when facilitating a risk session or when running schedule risk analysis. Default Most Likely (ML) value of each activity considered is equal to the activity’s remaining duration. The default was borne out of the basic principle of schedule development that calls for the schedule to represent the “most likely.”

It will be beneficial for anyone performing a three-point estimate to review the information and data quality of the following documents: 1) Basis of estimate, 2) Detailed Estimate, and 3) Estimate Summary.

This provides a good understanding on how the duration of an activity or work package came about. It is mandatory that the facilitator is familiar with the estimate, because crosschecking the resource inputs is essential (Frago, R., 2014.Risk-based Management in the World of Threats and Opportunities.Schedule Quantitative Risk Analysis.p99.ISBN 978-0-9947608-0-7, Canada).”

Depending on the assumed distribution, the below equation will apply.

  1. For Triangular distribution (Figure 1)

Expected duration         =          (Optimistic + ML + Pessimistic)/3

=          [(100+160+250)/3]

=          170 days

Figure 1 – Triangle Distribution Curve

       2. For Beta distribution (Figure 2)

Expected duration         =          (Optimistic + 4ML + Pessimistic)/6

=          [(100+4(160) +250)/6]

=          165 days

Figure 2 – Beta Distribution Curve

Developing the Most Likely Schedule

How the project comes up with the Most Likely Schedule depends on the project objectives, perspectives, and approaches. With the organization aligned to common objectives, with the approved criteria uniformly applied, the ML Schedule should not be too difficult to develop. Enumerated below are recommended approaches on how to develop the Most Likely Schedule.

a) The most likely schedule must satisfy the quality requirement of schedule risk analysis.

b) At any time, the most likely schedule must be the best schedule baseline candidate (Frago, R., 2016.Project Baseline Top Ten Prerequisites.Wordpress.Your World, Our Risk Universe).

It should be the most reliable and economical schedule available.

  • a result of interactive planning
  • Schedule scope shall equal plan scope
  • uses critical path method
  • uses more recent information as foundation
  • Several others (read BL Top 10 prerequisites)

c) Avoid built-in risk in the schedule

  • Assume NO restrictions on all resources
  • Assume high performance teams

Reminder:

Developing the Most Likely Schedule and doing the Schedule Risk Analysis are two distinct processes. The project shall consider risks during schedule risk analysis only to avoid double dipping; i.e. to avoid taking the same risk into account twice (Frago, R., 2014.Brief Introduction to SQRA.Future Sate Analysis.Slideshare.net)

Approaches and Philosophies

In addition to the foregoing, the project can use the following additional approaches and philosophies as guide in the creation, development, and management of their schedules.

  1. Use the most likely deterministic schedule as model in SRA
  2. The deterministic schedule used in SRA is a good BL schedule candidate. This statement depends on the quantified probabilities.
  3. Remove built in (or buffer float) in the schedule risk model.
  4. ‎Unless an effective part of the model, remove dummy activities (or buffer durations) in the schedule risk model.
  5. ‎A most likely schedule is composed of activities with most likely duration. Most likely durations are estimated based on any of the following method:
  • Most current estimate or much better, the approved frozen estimate
  • Expert ‎judgment (in-house or third party experts)
  • Accepted schedule benchmarks (internal and/or external benchmarks)
  • ‎Using historical records (internal and/or external history)
  • Using the P50 schedule calculated from a recently concluded SRA (Frago, R., 2015.Project Schedule:P50, Anyone?.LinkedIn Pulse)

P50 as the Most Likely Schedule

When data population increases, despite the randomness of available variables, the end-result is normal distribution. It is with this concept in mind that a P50 risk-based baseline holds a possible solution against the chance of overrun.

Preventing unachievable schedule is the main concern. In order to develop the P50 baseline, each of the activity making up the sub-projects has to be at a P50. Each of the sub-projects making up the projects portfolio has to be at a P50 as well.

Adjusting the probability of each sub-project schedule is simple. Changing final activity dates to reflect the P50 date is equally easy.

In the past, identifying the right dates and duration to increase achievability is just a wish. This time it is different. The available risk analysis tools these days are available to convert all project activities to a P50 (or if you want, any other probability value, even greater than P50).

The idea of P50 activities supporting other P50 activities to come up with a P50 schedule makes sense. It can potentially make execution less stressful because the target is sound. The target becomes a lot more achievable (Frago, R., 2014.Risk-based Management in the World of Threats and Opportunities).

Let us say it together, “Egad! It is the Most Likely Schedule!”

Rufran C. Frago – Author (31-Aug-16)

Rufran is the author of the book Risk-based Management in the World of Threats and Opportunities: A Project Controls Perspective.

Those who are interested, join Rufran at the following sites:

Related articles authored by Rufran Frago.

  1. Phantom Schedules
  2. Man is the Center of the Risk Universe
  3. Project Schedule Baseline Top 10 Prerequisites
  4. Setting Critical Path
  5. Schedule Critical path
  6. Primer to Good Schedule Integration
  7. Project Schedule: P50, Anyone?
  8. Schedule Baseline Dilemma Part 1
  9. Schedule Baseline Dilemma Part 2
  10. 4D Scheduling Part 1: What is it about?
  11. 4D Scheduling Part 2
  12. 4D Scheduling Part 3
  13. Mega-Projects Schedule Management and Integration
  14. Scaffolding Hours: What are they? Part 1
  15. Scaffolding Hours: What are they? Part 2
  16. Your World, Our Risk Universe
  17. Rufran Frago in the Global Risk Community Site
Advertisements

About rcfrago

Rufran C. Frago is a practicing Professional Engineer (APEGA), a PMP (PMI), a CCP (AACE) and a RMP (PMI). He has published more than 100 articles. He is the author of the book Risk-based Management in the World of Threats and Opportunities: A Project Controls Perspective He studied at Batangas State University and University of Batangas graduating with a Diploma in Petroleum Refinery Maintenance Technician (1979), Bachelor of Science in Mechanical Engineering (1984), and Bachelor of Science in Management Engineering in 1987 respectively. He was in his senior year taking up Bachelor of Science in Electrical Engineering, needing only one semester to complete, when he took a break to concentrate on married life. Rufran has never stopped academic learning after getting his degrees in the University. He continues his education by taking up some MBA courses under the University of the Philippines-PBMIT Consortium (1987-1988). He completed Computer Technician Program at International Correspondence School, Pennsylvania, USA in 1994, Applied Project Management Certificate program at Southern Alberta Institute of Technology in 2009, and Professional Management Certificate program specializing in Construction Management in 2014. He is now completing the Professional Management Certificate program specializing in Risk Management. He was a recipient of the Gerry Roxas Leadership Award (1976) and the American Field Service (AFS) Scholarship in 1976-77, studying in America for a year. Upon his return in 1977, California-Texas Philippines (Caltex Philippines Inc.), one of Asia’s biggest oil and gas refineries at the time, awards him with a two-year national college scholarship, specializing in Petroleum Refinery Maintenance. He went on extensive training in various maintenance disciplines for the next two years. Caltex hired him upon his graduation in 1979. He has spent more than 38 years of his life working in the Oil & Gas, Petrochemicals, Oleo-chemicals, Sugar Refining/Manufacturing, Consultancy, High School and University Education industries in Asia, Middle East, Canada, and North Africa). Rufran has worked with Caltex, Uniman, Unichem (now Cocochem), ARAMCO-KSA, Central Azucarera de Tarlac, Arabian Gulf Oil Company-Libya, Batangas State University, St. Bridget’s College, JG Summit Petrochemicals, Halliburton-Kellogg, Brown and Root, and OPTI Canada. He works with Suncor Energy Inc at present. He has wide range of expertise that includes problem solving, project management, training and mentoring, programs and projects planning and scheduling, cost management, risk-based management, construction management, project review and auditing, estimating, engineering and design, fabrication and module management, maintenance, operation, material selection, warehousing, EH&S and reliability engineering (predictive and preventive maintenance). He wants to share his knowledge and leave behind some form of legacy to all specially his wife, children and grandchildren, Eva and Mia.
This entry was posted in Analysis, Baseline Management, Business, Construction Management, Critical Path, Data Assessment, Decision Making, Deterministic Schedule, Economics, Integrated Schedule, Most Likely Schedule, Planning and Scheduling, Program Management, Program Schedule, Project Management, Risk Assessment and Treatment, Risk Universe, Risk-based Management, rufran frago, Rufran's Blogs and tagged , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s