research paper on book store

• Harvard Business School →
• Faculty & Research →
• Working Paper
• HBS Working Paper Series

Reinventing Retail: The Novel Resurgence of Independent Bookstores

• Format: Print
• | Language: English
• | Pages: 29

About The Author

Ryan L. Raffaelli

More from the author.

• February 2023 (Revised July 2023)
• Faculty Research

Moleskine Foundation: Can Creativity Change the World?

Institutional emplacement and the novel resurgence of independent bookstores.

• November 2022

Leader Action Orientations

• Moleskine Foundation: Can Creativity Change the World?  By: Ryan Raffaelli, Alexandra C. Feldberg and Sarah Gulick
• Institutional Emplacement and the Novel Resurgence of Independent Bookstores  By: Ryan Raffaelli and Ryann Noe
• Leader Action Orientations  By: Ryan Raffaelli, Akshaya Varghese and Laura Weimer

online bookstores Recently Published Documents

Total documents.

• Latest Documents
• Most Cited Documents
• Contributed Authors
• Related Sources
• Related Keywords

Patronage of Online Bookstores by Postgraduate Students at Nigerian Universities

The preference for online bookstores is increasing among students; however, research that focuses on online bookstores has not been given attention in the Nigerian university context. This study examined, through empirical analysis, the patronage of online bookstores by postgraduate students at Nigerian universities. A survey design was adopted using a questionnaire to collect data from 969 postgraduate students selected from twelve universities from the South-west and South-south regions. Six research questions were developed. The findings reveal that Amazon is the online bookstore most often patronised by Nigerian postgraduate students. The books ordered most often are textbooks, fiction, and reference books. The factors driving patronage of online bookstores are convenience, the wide range of products available, the opportunity to purchase anytime and anywhere, the ability to save time, and cost advantages. Culture and censorship do not have an impact on the patronage of online bookstores. Overall, postgraduate students are satisfied with their experience of online bookstores, although the challenges include internet connectivity, long waiting times for delivery of ordered books, inadequate information and communication technology (ICT) skills, limited access to computers with internet access, loss of ordered books during transit, and the risks inherent in paying for books ordered online. Since only one bookstore in this study, Ikasuwa, is based in Nigeria, it is recommended that Nigerian bookstores and booksellers consider creating online platforms to reach out to customers.

Quality Demand for Online Bookstores

A comparison of korean and japanese reader reviews on the『the first errand of dew』in online bookstores, collectiondevelopment.com: using amazon.com and other online bookstores for collection development, an analysis infant picture books recommended by online bookstores and expert picture book websites, evaluating the attributes of online bookstores: empirical evidence from young consumers in vietnam, game development to introduce indonesian traditional weapons using mda framework.

It is unfortunate that more and more youths are less concerned and are not interested in acquainting and studying Indonesian traditional culture, especially concerning on traditional Indonesian weapons. This fact is noticed from data as collected from several bookstores including online bookstores which the number of books on traditional weapons originating from abroad as published in Indonesia has a significant comparison compared to books on traditional weapons from Indonesia, in a stagerring 1:7 comparisson. Therefore, it is necessary to develop a media that can generate interest in couriosly learning about Indonesian culture, especially when introducing traditional Indonesian weapons. This research was conducted to develop a media in the form of android games that can be applied for the introduction of Indonesian culture, especially traditional weapons. The employed method of game development is the MDA Framework. Evaluation was carried out by using the Playtesting Evaluation and Gameflow methods. The evaluation results related to the concentration of players during game playing presents that the game has given a lot of encouragement to players to play games in several aspects of: game graphics, game sound, game stories, gameplay, simple interfaces, and completion of missions and main objectives regarding traditional Indonesian weapons, trigerring the player interest in playing this game and to maintain their focus throughout the game. In overal, the evaluation results using playtesting evaluation and Gameflow test depicts that a rating above average with an overall rating of 3.61 (72.18%) and the game Mystical Weapon of Nusantara can be accommodated as a medium for the introduction of traditional Indonesian weapons as indicated by the probability of a t-test of 1,169x10-19 which is far below the value of α = 0.05. Thus, it is concluded that there is an improvement in user knowledge, as well as the value of respondent answer showing above the average, which is 3.83 (76.67%).

A study on Korean literature introduced in Chinese online bookstores

Examination of the accuracy and applicability of information in popular books on dog training.

There is a wealth of popular literature available on dog behavior and training; sourcing reliable and trustworthy advice is important to achieving successful training. The aim of this study was to select five best-selling (at that time) dog training books, and review their general content and references to basic learning theory and human communicative cues. An Internet search was performed on three online bookstores’ websites for “best selling” “dog training” books. The books were by Millan and Peltier (2006), Fennell (2002), Stilwell (2005), Pryor (1999), and Monks of New Skete (2002). The results showed marked differences across all books, including inconsistencies in the depth of information provided, and some starkly contrasting training methods were advocated. Overall, these books were not all considered to function as instructional manuals. The persistent popularity of these books suggests that they have likely contributed appreciably to the type of information accessed by dog guardians.

Lost Identity – Metadata Presence in Online Bookstores

Export citation format, share document.

IEEE Account

• Change Username/Password
• Update Address

Purchase Details

• Payment Options
• Order History
• View Purchased Documents

Profile Information

• Communications Preferences
• Profession and Education
• Technical Interests
• US & Canada: +1 800 678 4333
• Worldwide: +1 732 981 0060
• Contact & Support
• About IEEE Xplore
• Accessibility
• Terms of Use
• Nondiscrimination Policy
• Privacy & Opting Out of Cookies

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity. © Copyright 2024 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

Enter the email address you signed up with and we'll email you a reset link.

• We're Hiring!
• Help Center

Online Book Store

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

Moscow Rules: A Quantitative Exposé

• Conference paper
• Open Access
• First Online: 09 June 2022
• Cite this conference paper

You have full access to this open access conference paper

• Eduardo Miranda   ORCID: orcid.org/0000-0001-8195-7506 10  

Part of the book series: Lecture Notes in Business Information Processing ((LNBIP,volume 445))

Included in the following conference series:

• International Conference on Agile Software Development

8402 Accesses

1 Citations

This article analyzes the performance of the MoSCoW method to deliver all features in each of its categories: Must Have, Should Have and Could Have using Monte Carlo simulation. The analysis shows that under MoSCoW rules, a team ought to be able to deliver all Must Have features for underestimations of up to 100% with very high probability. The conclusions reached are important for developers as well as for project sponsors to know how much faith to put on any commitments made.

You have full access to this open access chapter,  Download conference paper PDF

Similar content being viewed by others

Toward Impact Generation and Future Research

Friedenspädagogik

• Agile planning
• Release planning
• Requirements prioritization
• Feature buffers
• MosCoW method

1 Introduction

MoSCoW rules [ 1 ], also known as feature buffers [ 2 ], is a popular method to give predictability to projects with incremental deliveries. The method does this by establishing four categories of features: M ust Have, S hould Have, C ould Have and W on’t Have, from where the MoSCoW acronym is coined. Each of the first three categories is allocated a fraction of the development budget, typically 60, 20 and 20 percent, and features assigned to them according to the preferences Footnote 1 of the product owner until the allocated budgets are exhausted by subtracting from them, the development effort estimated for each feature assigned to the category. By not starting work in a lower preference category until all the work in the more preferred ones have been completed, the method effectively creates a buffer or management reserve of 40% for the Must Have features, and of 20% for those in the Should Have category. These buffers increase the confidence that all features in those categories will be delivered by the project completion date. As all the development budget is allocated by the method, there are no white spaces in the plan, which together with incentive contracts, makes the method palatable to sponsors and management.

Knowing how much confidence to place in the delivery of features in a given category is an important concern for developers and sponsors alike. For developers it helps in formulating plans consistent with the organization’s risk appetite, making promises they can keep, and in calculating the price of incentives in contracts as well as the risk of incurring penalties, should these exist. For sponsors, it informs them the likelihood the features promised will be delivered, so they, in turn, can make realistic plans based on it. To this purpose, the article will explore:

The probabilities of delivering all the features in each of the categories: Must Have, Should Have and Could Have, under varying levels of under and overestimation of the features’ development efforts

The impact of features’ sizes, dominance, number of features, and correlation between development efforts in said probabilities

The effect of budget allocations other than the customary 60/20/20 on them.

To calculate the probabilities of delivery (PoDs) we need to make suitable assumptions about the distribution of the efforts required to develop each feature since the single point estimate used in the MoSCoW method are insufficient to characterize them.

In this article, those assumptions are derived from two scenarios: a low confidence estimates scenario used to establish worst case Footnote 2 PoDs and a typical estimates scenario used to calculate less conservative PoDs.

The potential efforts required and the corresponding PoDs, are calculated using Monte Carlo simulations [ 3 , 4 ] to stochastically add the efforts consumed by each feature to be developed.

The rest of the paper is organized as follows: Sect.  2 provides an introduction to the MoSCoW method, Sect.  3 introduces the Monte Carlo simulation technique and describes the calculations used for the interested reader, Sect.  4 discusses the two scenarios used in the calculations, Sect.  5 analyzes the main factors affecting the method’s performance, Sect.  6 discuss the method’s effectiveness in each of the scenarios and Sect.  7 summarizes the results obtained.

2 The MoSCoW Method

The MoSCoW acronym was coined by D. Clegg and R. Baker [ 5 ], who in 1994 proposed the classification of requirements into Must Have, Should Have, Could Have and Won’t Have. The classification was made on the basis of the requirements’ own value and was unconstrained, i.e. all the requirements meeting the criteria for “Must Have” could be classified as such. In 2002, the SPID method [ 6 ] used a probabilistic backcasting approach to define the scope of three software increments roughly corresponding to the Must Have, Should Have and Could Have categories, but constraining the number of Must Have to those that could be completed within budget at a level of certainty chosen by the organization. In 2006, the DSDM Consortium, now the Agile Business Consortium, published the DSDM Public Version 4.2 [ 7 ] establishing the 60/20/20% recommendation although this, was probably used before by Consortium’s members on their own practices. The current formulation of the MoSCoW prioritization rules is documented in the DSDM Agile Project Framework [ 1 ].

During the project planning phase, see Fig.  1 .a, features are allocated to one of four sets: Must Have, Should Have, Could Have, and Won’t Have on the basis of customer preferences and dependencies until the respective budgets are exhausted.

MoSCoW rules at play: a) During planning, b) in execution

During execution, Fig.  1 .b, features in the Must Have category are developed first, those in the Should Have second, and those in the Could Have, in third place. If at any time the work in any category requires more effort than planned, work on them will continue at the expense of those in the lower preference categories which will be pushed out of scope in the same amount as the extra effort required. The advantage for the project sponsor is that, whatever happens, he or she can rest assured of getting a working product with an agreed subset of the total functionality by the end of the project.

For the MoSCoW method to be accepted by the developer as well as by the sponsor of a project, the risk of partial deliveries must be shared between both of them through incentive contracts since approaches like firm fixed price or time and materials, that offloads most of the risk on only one of the parties could be either, prohibitive or unacceptable to the other. Contractually, the concept of agreed partial deliveries might adopt different forms. For example, the contract could establish a base price for the Must Have set, with increasingly higher bonuses or rewards for the Should Have and Could Have releases. Conversely the contract could propose a price for all deliverables and include penalties or discounts if the lower priority releases are not delivered. This way the incentives and disincentives will prevent the developer from charging a premium price to protect itself from not delivering all features while the sponsor, is assured the developer will do its best, in order to win the rewards.

3 The Monte Carlo Simulation

The Monte Carlo method is a random sampling technique used to calculate probability distributions for aggregated random variables from elementary distributions. The technique is best applied to problems not amenable to closed form solutions derived by algebraic methods.

The Monte Carlo method involves the generation of random samples from known or assumed elementary probability distributions, the aggregation or combination of the sample values according to the logic of the model been simulated and the recording of the calculated values for the purpose of conducting an ex-post statistical analysis.

The technique is widely used [ 3 , 4 ] in probabilistic cost, schedule and risk assessments and numerous tools Footnote 3 exist to support the computations needed.

The results presented in the paper were calculated using @Risk 7.5. As these are the product of simulation runs, they might slightly differ from one run to another, or when using a different number of iterations or platforms.

The rest of the section explains the model used to generate the cumulative probability curves and calculate the PoD for each MoSCoW category: Must Have (MH), Should Have (SH) and Could Have (CH), with the purpose of allowing interested readers replicate the studies or develop their own simulations. Those not so inclined might skip it, with little or no loss in understanding the paper. The name of the parameters should make them self-explanatory however, conceptual definitions about its meaning and usage will be provided throughout the paper.

The probability of completing all features in a given category in, or under, an \(x\) amount of effort is defined as:

The cumulative distribution functions: \(F_{MH} \left( x \right), \,F_{SH} \left( x \right)\, {\text{and}}\, F_{CH} \left( x \right)\) , are built by repeatedly sampling and aggregating the effort required by the features included in each category.

similarly, for features j and k, and:

subject to the maximum allocation of effort for each category:

The Probability of Delivery (PoD) of each category is defined as:

All quantities are normalized for presentation purposes by dividing them by the \(DevelopmentBudget\) .

4 Low and Typical Confidence Scenarios

Figure  2 contrasts the two scenarios mentioned in the introduction. The low confidence scenario is characterized by the uniform distribution of the potential efforts required to realize each feature, with the lower limit of each distribution corresponding to the team’s estimated effort for the feature and their upper to increments of 50, 100 and 200% above them, to express increasing levels of uncertainty. Since all values in the interval have equal probability, this scenario corresponds to a maximum uncertainty state [ 8 ]. This situation, however unrealistic it might seem, is useful to calculate a worst case for the PoD of each category. In the typical confidence scenario, the potential efforts are characterized by a right skewed triangular distributions, in which the team’s estimates correspond to the most likely value of the distribution, meaning the realization of many features will take about what was estimated, some will take some more and a few could take less.

Probability distributions for the effort required by each feature in the low (uniform distributions) and typical (triangular distributions) confidence scenarios

The right skewness of the typical estimate distributions is predicated on our tendency to estimate based on imagining success [ 9 ], behaviors like Parkinson’s Law Footnote 4 and the Student Syndrome Footnote 5 , which limit the potential for completing development with less effort usage than estimated, and the fact that the number of things that can go wrong is practically unlimited [ 10 , 11 ]. Although many distributions fit this pattern, e.g. PERT, lognormal, etc., the triangular one was chosen for its simplicity and because its mass is not concentrated around the most likely point [ 12 ], thus yielding a more conservative estimate than the other distributions mentioned.

As before, the right extreme of the distribution takes values corresponding to 50, 100 and 200 percent underestimation levels. For the lower limit however, the 80 percent of the most likely value was chosen for the reasons explained above.

Considering this second scenario is important, because although having a worst case for the PoDs is valuable as they tell the lowest the probabilities could be, relying on them for decision making may lead to lost opportunities because of overcautious behaviors.

5 Level of Underestimation, Correlation, Number of Features in a Category, Feature Dominance and Non-traditional Budget Allocations

Before calculating the PoDs for each MoSCoW category under the two scenarios, the impact of different factors on the PoD is explored with the purpose of developing an appreciation for how they affect the results shown, i.e. what makes the PoDs go up or down. Understanding this is important for those wanting to translate the conclusions drawn here to other contexts.

Although the analysis will be conducted only for the low confidence estimates for reasons of space, the same conclusions applies to the typical estimates scenario, with the curves slightly shifted to the left.

Figure  3 shows the impact of underestimation levels of up to 50, 100 and 200% of the features’ individual estimates on the PoD of a Must Have category comprising 15 equal sized features, whose development efforts are independent from each other.

Independent, as used here, means the efforts required by any two features will not deviate from its estimates conjointly due to a common factor such as the maturity of the technology, the capability of the individual developing it or the consistent over optimism of an estimator. When this occurs, the efforts are correlated rather than independent. Having a common factor does not automatically mean the actual efforts are correlated. For example, a feature could take longer because it includes setting up a new technology, but once this is done, it doesn’t mean other features using the same technology would take longer since the it is already deployed. On the other hand, the use of an immature open source library could affect the testing and debugging of all the features in which it is included.

The higher the number of correlated features and the stronger the correlation between them, the more individual features’ efforts would tend to vary in the same direction, either requiring less or more of it, which would translate into higher variability at the total development effort level. This is shown by curves “r = 0.2”, “r = 0.6” and “r = 0.8” in Fig.  4 , becoming flatter as the correlation (r) increases.

Correlation brings good and bad news. If things go well, the good auspices will apply to many features, increasing the probability of completing all of them on budget. Conversely, if things do not go as well as envisioned, all affected features will require more effort, and the buffers would not provide enough slack to complete all of them.

Estimating the level of correlation between estimates is not an easy task, it requires assessing the influence one or more common factors could have on the items affected by them, a task harder than producing the effort estimates themselves. So while correlation cannot be ignored at risk of under or over estimating the safety provided by the method, the cost of estimating it, would be prohibitive for most projects. Based on simulation studies, Garvey et al. [ 13 ] recommend using a coefficient of correlation of 0.2 across all the estimated elements to solve the dilemma, while Kujawski et al. [ 14 ], propose to use a coefficient of 0.6 for elements belonging to the same subsystem, as these would tend to exhibit high commonality since in general, the technology used and the people building it would be the same, and 0.3 for elements on different subsystems, because of the lower commonality.

Cumulative completion probabilities under increasing levels of underestimation. The simulation shows a PoD for the Must Have features of 100% for an underestimation level of up to 50%, of 98.9% at up to 100%, and of 1.3% for an underestimation in which each feature can require up to 200% of the estimated budget.

The PoDs are also affected by the number of features in the category as well as by the existence of dominant features, which are features whose realization requires a significative part of the budget allocated to the category. See Figs.  5 and 6 .

As in the case of correlation, a small number of features and the presence of dominant features result in an increase in the variability of the estimates. Dominant features, contribute to this increase because it is very unlikely that deviations on their effort requirements could be counterbalanced by the independent deviations of the remaining features in the category. As for the increase of variability with a diminishing number of features, the reason is that with a fewer independent features, the probability of them going all in one direction, is higher than with many features.

The model in Fig.  7 challenges the premise of allocating 60% of the development budget to the Must Have category and explores alternative assignments of 50, 70 and 80% of the total budget. Reducing the budget allocation from 60 to 50% increases the protection the method affords at the expense of reducing the number of features a team can commit to. Increasing the budget allocation for the Must Have allows developers to promise more, but as will be shown, this is done at the expense of reducing the certainty of delivering it. For the 50% allocation level, there is a 100% chance of delivering the Must Have for underestimations of up to 100%, and of 68.2% for underestimations of up to 200%. At the 70% allocation level, the simulation shows that the PoD for the Must Have, when the possibility of underestimation is up to 50% still is 100%, but that it drops sharply to 34% when the underestimation level rises to up to 100%. For the 80% allocation level, the PoD for the Must Have falls to 49.7% for the up to 50% underestimation level and to 0 for the other two. The rest of the paper will then use the customary 60, 20 & 20% allocation scheme.

Probability of completing all features in the Must Have category under a given percent of the budget when the underestimation level is up to 100% and the efforts are correlated (r > 0)

Influence of the number of features on the PoD for a Must Have set containing the number of equally sized independent features indicated by the legend on the chart, with an underestimation level of up to 100%. The PoD offered by the method drops sharply when the set contains less than 5 features

Influence of a dominant feature on the PoD. Each set, with the exception of the dominant at 100%, contained 15 features, with the dominant feature assigned the bulk of the effort as per the legend in the chart with the remaining budget equally distributed among the other 14 features. The safety offered by the method drops sharply when a feature takes more than 25% of the budgeted effort for the category. Underestimation of up to 100% and independent efforts

Probability of delivering all Must Have features for Must Have budget allocations of 50, 60, 70 and 80% under different underestimation conditions. The respective number of Must Have features for each budget allocation were 12, 15, 17, and 20.

6 Probabilities of Delivery for Each MoSCoW Category

This section discusses the PoDs for each MoSCoW category: Must Have, Should Have and Could Have under the following conditions:

Low confidence estimation, independent efforts

Low confidence estimation, correlated efforts

Typical estimation, independent efforts

Typical estimation, correlated efforts

In all cases, the underestimations considered are of up to 50, 100 and 200% of the estimated effort, a 60/20/20 effort allocation scheme and a Must Have category comprising 15 equal sized features with Should and Could Have categories comprising 5 equal sized features each. These assumptions are consistent with the precedent analysis and with the small criteria in the INVEST [ 15 ] list of desirable properties for user stories. For the correlated efforts cases, the article follows Kujaswki’s recommendation, of using an r = 0.6, as many of the attributes of an agile development project: dedicated small teams, exploratory work and refactoring, tend to affect all features equally.

6.1 Low Confidence, Independent Efforts

Figure  8 shows the PoDs for all MoSCoW categories for the low confidence, uncorrelated features, r = 0, model. At up to 50% underestimation, the probability of delivering all Must Have is 100%, as expected, and the probability of delivering all Should Have is 50.2%. At up to 100% underestimation, the probability of delivering all the Must Have still high, 98.9% but the probability of completing all the Should Have drops to 0. At up to 200% the probability of delivering all the Must Haves is pretty low, at 1.3%. In no case it was possible to complete the Could Have within budget.

6.2 Low Confidence, Correlated Efforts

As shown by Fig.  9 , in this case the variability of the aggregated efforts increases, with the outermost points of the distribution becoming more extreme as all the efforts tend to move in unison in one or another direction. Comparing the PoDs for this case with those of the previous one, it seems paradoxical, that while the PoD for the Must Have at 100% underestimation level goes down from 98.9 to 74.0, the PoD for the same category at 200% underestimation level goes up from 1.3 to 26.9%! This is what was meant when it was said that correlation brought good and bad news.

Probability of delivering all features in a category in the case of low confidence estimates under different levels of underestimation when the efforts required by each feature are independent (r = 0)

To understand what is happening, it suffices to look at Fig.  10 . Figure  10 .a shows histograms of the Must Have aggregated independent efforts for uncertainty levels of 50, 100 and 200%. Because of the relatively lower upper limit and the tightness of the distribution spread afforded by the sum of independent efforts, the 100% uncertainty distribution fits almost entirely to the left of the total budget, scoring this way a high PoD. A similar argument could be made for the 200% uncertainty level, except that this time, the distribution is almost entirely to the right of the total budget, thus yielding a very low PoD. As could be seen in Fig.  10 .b, when the efforts are correlated, the distributions spread more widely, making part of the 100% distribution fall to the right of the total budget line, reducing its PoD, and conversely, part of the 200% distribution might fall to the left of the line, thus increasing its PoD, which is what happened with this particular choice of parameter values.

Probability of delivering all features in a category in the case of low confidence estimates under different levels of underestimation when the efforts required by each feature are highly correlated (r = 0.6)

Histograms for Must Have features’ effort (a) left – independent efforts, (b) right – correlated efforts

6.3 Typical Estimates

Figures  11 and 12 show the typical estimates’ PoDs for uncorrelated and correlated efforts respectively. As expected, all the PoDs in this scenario are higher than in the case of the low confidence estimates. In the case of independent efforts, at up to 50% underestimation, the PoDs for the Must Have and the Should Have are 100%. At up to 100% underestimation, the PoD for the Must Have is 100% with the PoD for Should Have dropping to 39.7%. At up to 200% the probability of delivering all the Must Haves still high, at 70.5%, but there is no chance of delivering the Should Have. In no case, any Could Have were completed. For the correlated efforts case, the respective probabilities at 50% underestimation are: 100% for the Must Have, 88.7% for the Should Have and 20.6% for the Could Have. At 100% underestimation: 96.4, 50.3 and 8.6% respectively and at 200% underestimation: 59.8, 20.5 and 3%.

Probability of delivering all features in a category in the case of typical estimates under different levels of underestimation when the efforts required by each feature are independent (r = 0)

Probability of delivering all features in a category in the case of typical estimates under different levels of underestimation when the efforts required by each feature are highly correlated (r = 0.6).

This article sought to quantitatively answer the following questions:

What are the probabilities of delivering all the features in each of the categories: Must Have, Should Have and Could Have, under varying levels of under and overestimation of the features’ development efforts?

What is the influence of features’ sizes, feature dominance, number of features, and correlation between development efforts in said probabilities?

What is the effect of budget allocations other than the customary 60/20/20 on them?

To answer question 1, it is necessary to look at Table 1 which summarizes the results for the low confidence and typical estimates scenarios, for the three levels of underestimation studied: 50, 100 and 200%.

Not surprisingly, the results indicate that the method consistently yields a high PoD for the Must Have features. What is noteworthy, is its resilience in face of up to 100% underestimation of individual features in the category. For the Should Have, the results are robust for up to 50% of underestimation and with regards to the Could Have, they should only be expected if destiny is smiling upon the project.

Question 2 is important for practitioners preparing release plans. For the method to offer these levels of certainty, the number of features included in each category should be at least 5 with none of them requiring more than 25% of the effort allocated to the category. If these conditions are not met, the safety offered by the method drops sharply. Correlation, as mentioned before, is a mixed blessing. Depending on which direction things go, it can bring the only possibility of completing all the features in the project. Notice that in Table 1 , all the Could Have can only be completed when the efforts are highly correlated since all of them must be low. Under the independence assumption, when some could be low and others high, there is no chance of completing them on or under budget.

With regards to question 3, the 60, 20, 20% allocation seems to be the “Goldilocks” solution, balancing predictability with level of ambition. As shown by Fig.  7 , changing the allocation from 60 to 70%, has a dramatic impact on the safety margin which, at the up to 100% underestimation level, drops from 98.5 to 34%.

Finally, it is worth making clear, that the analysis refers to variations in execution times of planned work and not changes in project scope, which should be addressed differently.

The author gratefully acknowledges the helpful comments of Hakan Erdogmus. Diego Fontdevila and Alejandro Bianchi on earlier versions of this paper.

These preferences might induce dependencies that need to be addressed by the team, either by incorporating lower preference features in the higher categories or by doing additional work to mock the missing capabilities.

Worst case, means that if some of the assumptions associated with the scenario were to change, the probability of delivering within budget would increase.

@Risk by Palisade, Crystal Ball by Oracle, ModelRisk by Vose and Argo by Booz Allen among others.

Parkinson’s Law, the 1955 assertion by British economist Cyril Northcote Parkinson, that “Work expands so as to fill the time available for its completion”, regardless of what was strictly necessary.

Student Syndrome, a term introduced by Eliyahu M. Goldratt in his 1997 novel Critical Chain to describe the planned procrastination of tasks by analogy with a student leaving working in an assignment until the last day before its due date.

Agile Business Consortium: Chapter 10 MoSCoW Prioritization, January 2014. https://www.agilebusiness.org/page/ProjectFramework_10_MoSCoWPrioritisation . Accessed 10 Oct 2021

Cohn, M.: Agile Estimationg and Planning. Prentice Hall, Upper Saddle River (2006)

Google Scholar  

Garvey, P.: Elements of Cost Risk Analysis. MITRE Corporation (2014)

Magennis, T.: Managing software development risk using modeling and Monte Carlo simulation. In: Lean Software and Systems Consortium Conference, Boston (2012)

Clegg, D., Baker, R.: CASE Method Fast-track: A RAD Approach. Addison-Wesley, Reading (1994)

Miranda, E.: Planning and Executing Time Bound Projects. IEEE Comput. (2002)

DSDM Consortium: Not longer accessible. http://www.dsdm.org/version4/2/public/default.asp

Shu-Cherng, F., Tsao, H.: Entropy optimization: shannon measure of entropy and its properties. In: Encyclopedia of Optimization, Springer, Boston (2001). https://doi.org/10.1007/0-306-48332-7_119

Newby Clark, I., Buehler, R., Koehler, D., Griffin, D.: People focus on optimistic scenarios and disregard pessimistic scenarios while predicting task completion times. J. Exp. Psychol. Appl. (2000)

Halkjelsvik, T., Jørgensen, M.: Time Predictions: Understanding and Avoiding Unrealism in Project Planning and Everyday Life, Springer, Cham (2018). https://doi.org/10.1007/978-3-319-74953-2

Kitchenham, B., Linkman, S.: Estimates, uncertainty and risk. IEEE Softw. (1997)

Hulett, D.: Practical Schedule Risk Analysis, Gower (2009)

Garvey, P., Book, S.C.R.: Probability Methods for Cost Uncertainty Analysis, 2nd. CRC Press (2016)

Kujawski, E., Edwards, W.R., Alvaro, M.L.: Incorporating psychological influences in probabilistic cost analysis. Syst. Eng. 7 (3) (2004)

Wake, B.: INVEST in Good Stories, and SMART Tasks. XP123, 17 8 (2003). https://xp123.com/articles/invest-in-good-stories-and-smart-tasks/ . Accessed 19 Nov 2021

Download references

Author information

Authors and affiliations.

Carnegie Mellon University, Pittsburgh, PA, 15213, USA

Eduardo Miranda

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Eduardo Miranda .

Editor information

Editors and affiliations.

University of Oslo, Oslo, Norway

Viktoria Stray

University College Cork, Cork, Ireland

Klaas-Jan Stol

LUT University, Lahti, Finland

Maria Paasivaara

University of British Columbia, Vancouver, BC, Canada

Philippe Kruchten

Rights and permissions

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/ ), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Reprints and permissions

Copyright information

© 2022 The Author(s)

About this paper

Cite this paper.

Miranda, E. (2022). Moscow Rules: A Quantitative Exposé. In: Stray, V., Stol, KJ., Paasivaara, M., Kruchten, P. (eds) Agile Processes in Software Engineering and Extreme Programming. XP 2022. Lecture Notes in Business Information Processing, vol 445. Springer, Cham. https://doi.org/10.1007/978-3-031-08169-9_2

Download citation

DOI : https://doi.org/10.1007/978-3-031-08169-9_2

Published : 09 June 2022

Publisher Name : Springer, Cham

Print ISBN : 978-3-031-08168-2

Online ISBN : 978-3-031-08169-9

eBook Packages : Computer Science Computer Science (R0)

Share this paper

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

The Penn Libraries has built a new, user-focused library catalog, with a simpler look and easier search features. Explore it at find.library.upenn.edu .

Tools for Research and Writing

Penn Libraries makes a variety of tools available that facilitate the process of writing research papers at any level. Platforms like NoodleTools, PowerNotes, and LaTeX can help you organize your notes, create outlines, and more.

Browse research and writing tools

Featured tools

Need more research and writing help? Contact your subject librarian ! 

Noodle Tools

Latex and overleaf.

• Use technology and software
• Get Expert Help
• Faculty/Instructors
• Undergraduates
• Graduate Students

BookPeople of Moscow

• Browsing: »

Please visit our new website: www.bookpeopleofmoscow.com

Our official store website is now online at http://www.bookpeopleofmoscow.com . Please visit us there!

Share this:

Bookpeople’s top 7 list.

Thanks to everyone for being such great customers and supporters of BookPeople of Moscow. Our first year has been going great thanks to all of you; please know that we don’t take you for granted. We look forward to each and every person who comes through the door and we are really loving getting to know everyone.

We’ll call this the  BookPeople Top 7 List  this time since I can’t think of 10 things to include. Too much election night loss of sleep I guess!

7.  We believe our community is only as strong as our investment in it. Help keep Moscow fabulous! Please look first to unique, locally-owned businesses (like ours!) and non-profits for all your holiday gift needs.

6.  BookPeople will kick off our “Thanks for shopping indie” campaign on Friday, November 23rd (the day after Thanksgiving). The store will open early at 8 am and be open until 8 pm (our usual closing time on Fridays.) The campaign features some great book choices – some might already be on your wish list!

5.  Our holiday book catalog is now available to help inspire gift ideas for family and friends – pick one up the next time you’re in the store. Our goal is to have every book in the catalog available in the store throughout the holiday season.

4.  Our website is getting closer to launching and we look forward to everything that comes with having a good website: the ability for our customers to order almost any book from us online and either have it shipped or pick it up in the store, the ability to purchase ebooks from us, an easy to use online calendar showing all the upcoming store readings and events, and more. When our website is live we’ll send you the link so you can check it out for yourselves!

And, along with selling ebooks, we’ll have Kobo ereaders available for sale very shortly – the great indie alternative to other, more proprietary ereaders like the Kindle. Kobo’s motto: Read Freely. Our sentiments exactly.

3.  Starting the day after Thanksgiving, BookPeople will sponsor a “giving tree” in conjunction with “Christmas for Kids.” On the tree will be tags with the ages and genders of children who would love to receive books for Christmas but whose families can’t afford it. We encourage everyone to purchase a book to put under the giving tree; we will wrap them and be sure they get to the children on time. It can be a meaningful family project to choose books for the giving tree together. And, a huge thank you to the girl scout troop under the leadership of Amy Ball who decorated over a hundred lovely, glittery tags for the giving tree.

2.  Did you know? Proceeds from our used book sales go into the new BookPeople Community Fund, a source of money to assist teachers in getting books for their classrooms, to support local arts, to fund literary events, and to foster literacy. Please consider donating your used books to BookPeople, and help the Community Fund grow! The used books, by the way, are in the back of the store, just past the children’s area.

1.  Last but not least, you deserve a break. We want to make your holidays easier. So, on certain days, besides our usual super staff, we will have the benefit of incredibly knowledgeable children’s librarians and teachers volunteering in the store to help you find just the right books. We gift wrap for free (as always), and will also package and ship your gifts to their recipients, so you don’t have to! Let us stand in the post office line on your behalf – we’ll only charge you for the actual cost of the postage.

Meet Nick of Black Cypress at BookPeople!

Please join us at BookPeople on Sunday evening at 5 pm to sample some delicious food by Nick Pitsilionis of Black Cypress restaurant, and pick up a signed copy of the Washington Artisans cookbook, which features one of Nick’s recipes!

Duff Brenna reading and signing, Thursday, Oct. 25, 7:30 pm

Bette Lynch Husted and Pamela Steele, Saturday at 5 pm

Alan Heathcock reading and signing Thursday, Oct. 18, 7:30 pm

Literary evening for adventure lovers, Wed. Oct. 17

Learn more about Gloria Steinem before she visits Moscow

Saroyan Prize Winner Daniel Orozco at BookPeople Wed. Sept. 26, 7:30 pm

Ebook alternative and new website coming soon to BookPeople

Good news! The American Booksellers Association announced this week a new contract with Kobo, which will take effect in a few months and replace the ABA’s old agreement with Google ebooks that expires at the end of this year. Independent bookstores all over the country and the world will soon be working with Kobo through the ABA to offer an affordable and more flexible alternative to another more, ahem, proprietary ereader and ebook provider who shall not be named. Kobo offers applications that allow their books to be read on other devices such as Nooks and iPads. BookPeople will be launching its official website by the end of the year and with it, Kobo ebooks and perhaps Kobo e-readers. Sales proceeds from Kobo ebooks through our website will be shared between us and Kobo – making it possible for customers anywhere to support their favorite local bookstore and still enjoy digital books.

We will keep you posted here about this issue. Watch for more news later in September.

Post navigation

• Already have a WordPress.com account? Log in now.
• Subscribe Subscribed
• Report this content
• View site in Reader
• Manage subscriptions
• Collapse this bar