Learn : Postpone can be used to shift outstanding repetitions to a later time, esp. in cases of a severe material overload. This options is particularly useful in heavily overloaded incremental reading.

Introduction: More and more learning

In the last millennium (i.e. before the advent of incremental reading in SuperMemo), the world of learning provided two extreme alternatives:

1. traditional learning with high volume of the learning material and dismal retention. This learning can be called High In - High Out (HIHO) as high volumes of knowledge processed on the input are accompanied by equally high volumes of knowledge on the "output" (i.e. lost out to forgetting)
2. SuperMemo with a small volume of strategic material remembered for years with a programmed level of knowledge retention (LILO - Low knowledge input with low forgetting outflow)

Neither alternative provides a complete solution to the problem of forgetting. Combining the two was left to the student. The optimum strategy was to use traditional learning to sift the strategic material, and then to memorize the strategic material with SuperMemo. 

Weaknesses of traditional SuperMemo were a powerful inhibitory force against its universal adoption:

1. The process of selecting the strategic knowledge was not automated
2. The student had to type in most of the learned knowledge to the program
3. The student had to master knowledge structuring skills to make the memorized knowledge truly applicable
4. Passive review was discounted as ineffective and was not used as a knowledge prioritization tool
5. Retention was programmable solely via the forgetting index (within 3-20% forgetting rate)
6. Accumulation of outstanding material after a short break was listed as the number one cause of users dropping SuperMemo for good

Incremental reading bridges the worlds of traditional learning and that of SuperMemo by providing a fluent transition between all priority levels starting from quick review, to repeated passive review, to active review, and to active recall at all level of the forgetting index down to 3% (i.e. nearly 99% recall rate). Incremental reading also includes an anti-overload option: Postpone that blurs the concept of the outstanding material that levied heavy toll on SuperMemo usership.

However, until SuperMemo 2002, incremental reading did not automate the process of "prioritizing by overflow". In other words, high volumes of learning would force the student to manually reschedule learning in branches or subsets of lower priority. Reading lists were used as an imperfectly prioritizing stopcock against the material overflow. Manual rescheduling of the material was necessary to protect core knowledge from a dramatic drop in retention. Substantially reduced retention slows down the learning process not only via the forgetting index vs. acquisition rate relationship, but also through unraveling of the incremental reading process where the new knowledge is built on a shaky foundation of what has been learned earlier. In addition, the flow of thought in incrementally read articles becomes disrupted if the core knowledge retention is not protected.

SuperMemo 2002 introduces a Postpone tool that makes it possible to reschedule lower priority material with a keystroke. The student can painstakingly work out priorities for individual subsets or branches of knowledge and then apply these easily on a daily basis at a negligible cost of time.

To fully understand the importance of Postpone, you need to mater the concept of incremental reading first.

Postpone dialog

The Postpone dialog box presented below includes three tabs: Scope, Parameters, and Adjust. The Scope tab settings are mostly done automatically for you depending on the context in which you use Postpone. The Parameters tab makes it possible to define the degree to which items and/or topics are postponed in the selected subset, and which items and/or topics are exempt from rescheduling. Once you define Parameters you can save the new setting with Scope : Save. Later you can restore the setting by using Settings name combo box.

Parameters tab

You need to understand the Parameters tab if you want to customize the degree of delay in rescheduling.

Delay factor determines how much elements should be delayed. For example, if you choose the delay of 1.1 (10%) on an element with the interval of 100 days, it will be delayed by ten days, i.e. rescheduled to the interval of 110 days. 

Maximum interval puts a ceiling on the length of the delay interval. For example, if you choose the delay of 1.1 on an element with the interval of 200 days, and the maximum interval is 5 days, the element will be delayed only by 5 days (instead of the 20 days produced by multiplying the original interval by the delay factor).

Minimum interval puts a floor on the length of the delay interval. For example, if you set it to 3, all delayed elements will be delayed by no less than 3 days.

Skip conditions determine which elements will be exempt from delay.

• if you check Skip items or Skip topics, items or topics (respectively) will not be rescheduled 
• Interval beyond will exempt elements with intervals longer than a selected value. For example, if you set Interval beyond to 365 days, no elements with interval longer than one year will be rescheduled. This can help you avoid infinite postpones. Once the element is postponed beyond a given interval, it may automatically enter the core of elements that you will have to review to zero the outstanding elements queue
• Forgetting index below makes it possible to determine the level of the forgetting index that qualifies the learning materials as the core knowledge. For example, if you set Forgetting index below to 5%, items with forgetting index 3% and 4% will not be rescheduled. This way you can make sure that the most important items in your collection never get postponed
• A-Factor below is analogous to Forgetting index below. The only difference is that you will use it with topics instead of items. Set A-Factor of most important articles to a very low value (e.g. 1.01). This way you will be able to filter them out at Postpone and make sure they never get rescheduled
• Delay above makes it possible to avoid rescheduling of elements whose delay reached a too high level (in interval percent). The default value is 50%

Adjust tab

The Adjust tab can be used for fine tuning delays on subsets with a complex priority structure of individual branches of knowledge. Sub-branch postpones determines how the rescheduling procedure interprets branches included in the subset that have their own postpone setting defined. Here are the possible options:

• Respect setting - respect the setting defined for individual branches and ignore the currently used global setting of the Parameters tab. For a comprehensive postpone towards the end of your learning day (i.e. at times when you want to focus on your core knowledge), postpone all less important branches and be more lenient on branches that include most vital pieces of knowledge
• Ignore setting - use the Parameters tab setting for all elements in the postponed subset. Ignore all branch postpone definitions. For example, if you want to postpone all topics in your collection, you can ignore the settings for individual branches, which may have Skip topics selected. This option is the fastest as it does not need to scan branches for individual postpone definitions
• Always choose most conservative setting - check the branch setting and postpone as little as possible (i.e. choose the definition that is least drastic)
• Always choose most liberal setting - check the branch setting and postpone as much as possible (i.e. choose the definition that provides for the highest delay). For example, minutes before you finish your learning day, postpone whatever can be postpone and focus only on the cream of the crop of your knowledge

Advanced settings on the Adjust tab include: Include elements that are not outstanding (switching between Postpone and Dilute), Modify item delay in proportion to the forgetting index (for delaying items in proportion to the forgetting index) and Modify topic delay in proportion to A-Factor (for delaying topics in proportion to their A-Factor).

Simulate, Postpone and Dilute

Once you define the settings on the Parameters tab and, optionally, fine tune them on the Adjust tab, save them with Save on the Scope tab. You can now test the postpone procedure by clicking Simulate. If you are not satisfied with the results, redefine Parameters and Save the setting again. Once you are happy with the result of the simulation, choose Postpone. 

The option Dilute should not normally be used (e.g. see the subset processing menu). It is analogous to Postpone but it also delays elements that are not outstanding. Occasionally, it may happen that you learn a lot of material in a branch that is no longer important to you. Instead of deleting such a branch, you may decide to incrementally sift it of the less relevant material. To reduce the emphasis on this branch in your daily review, use Dilute and spread the material over a longer period of time.

Using Postpone in various contexts

• globally - to postpone elements globally, choose options from the Learn : Postpone submenu on the main menu
• browser - to postpone elements listed in a browser, use Learning : Postpone on the subset processing menu or press Ctrl+Alt+P
• contents window - to postpone elements in a given branch, select it in the contents window, and use Learning : Postpone on the subset processing menu or press Ctrl+Alt+P
• element window - to postpone elements belonging to any ancestor branch of the current element (e.g. during repetitions), choose Learning : Postpone branch on the element menu. Respond with No until you reach the desired ancestor branch
• Ancestors window (Alt+P) - right-click over any ancestor branch and choose Postpone branch

Exemplary strategy

Few people are able to focus efficiently on learning with SuperMemo for longer than 40-60 minutes. However, you should be able to do far more if you divide your learning day into portions. You will get best results when learning at times of highest alertness (see: Good sleep for good learning to find out the two main alertness points in a healthy circadian rhythm). Here is an exemplary strategy:

• Main block, 45 min, morning, randomized repetitions, incremental reading, item repetitions, etc.
• Main repetitions block, 30 min, early evening, randomized repetitions after executing a postpone on topics with the exception of the highest priority articles
• Core knowledge block, 20 min, randomized repetitions after executing a postpone on lower priority items and the remaining topics. Adjust : Sub-branch postpones : Respect setting can be used here to define separate postpone rules for individual branches and categories
• Vital knowledge block, 20 min, before sleep, randomized repetition after executing a postpone on all but the most important pieces of knowledge. Ideally, you should be able to fine tune postpone parameters in a way that you are able to complete all repetitions on your vital knowledge material. This condition is necessary for ensuring the programmed level of knowledge retention guaranteed by SuperMemo

Postpone or Mercy?

Postpone should drastically reduce the role of Mercy in SuperMemo. However, you may still need Mercy in the following circumstances:

1. Randomizing repetitions (Mercy : Criteria : Randomize)
2. Rescheduling the outstanding material after a longer break (with chosen Mercy : Criteria)
3. Advancing future repetitions to an earlier date (e.g. before a vacation)