Lesson Report:
# Title
**Finalizing the Prediction Pipeline: Baselines, Probability Shifts, and Warning Indicators**
This shortened class focused on moving students from simply naming possible outcomes to building a structured analytic process for judging which outcome is most likely. The instructor emphasized making outcomes measurable, establishing a baseline probability, identifying factors that increase or decrease likelihood, and preparing the “warnings and indicators” section required for the final memo.
# Attendance
– **Class length:** Reduced to **50 minutes** instead of the usual 70 due to Career Day.
– **Absent students mentioned by name:** **0**
– **Number of absences explicitly identified in transcript:** **Not specified**
– **General note:** The instructor noted a **reduced class** and briefly commented on waiting to see how many students would arrive, but no formal roll call or named absences were recorded.
# Topics Covered
## 1. Opening, time adjustment, and lesson objective: finishing the prediction pipeline
– The instructor opened by noting that the class was shorter than usual because of **Career Day**, so the session would move quickly.
– The main goal for the day was to **finalize the prediction pipeline** students had been developing across prior lessons.
– The instructor framed the session around a core question: how do students move from **stating an outcome they think might happen** to **scientifically justifying how likely that outcome is**?
– The class objective was to consolidate prior work into a process students could apply in their **final memo/project**.
## 2. Review of the three outcome categories
– The instructor began by reviewing the three broad types of outcomes students had already been using:
– **Status quo remains**
– **Escalation**
– **De-escalation**
– The instructor reiterated that the **status quo** is the simplest possible outcome because it assumes current conditions persist.
– Students were prompted to recall the other categories and to refine how they should be measured:
– An **uncertain student** identified **“escalation”** as the second outcome type.
– Another **uncertain student** helped define escalation more precisely as **“more costly.”**
– An **uncertain student** identified the third category as **“de-escalation.”**
– Another **uncertain student** translated de-escalation into measurable terms as **“less costly.”**
– A further **uncertain student** confirmed that the status quo could be understood as **“same costs.”**
– The instructor emphasized that **cost** would be the operational measure that makes outcome categories analyzable:
– Escalation = **more costly**
– De-escalation = **less costly**
– Status quo = **roughly the same costs / no major material change**
## 3. Outcome quality check: specific vs. vague predictions
– The next major step was to ensure that students’ outcome statements were analytically usable.
– The instructor introduced a distinction between:
– **Specific outcomes**: tied to a concrete, observable event
– **Vague outcomes**: too broad or ambiguous to measure reliably
– Example of a **specific outcome** given in class:
– *“China will send 15 more naval carriers into the sea surrounding Taiwan within the next six months.”*
– The instructor used this to show that analysts can later verify whether the event did or did not occur.
– Example of a **vague outcome** given in class:
– *“Tensions will increase between China and Taiwan.”*
– The instructor explained that “tensions” could refer to too many different things and therefore lacks analytic precision.
– Key lesson point: if an outcome is too vague, it cannot be meaningfully tested.
## 4. Group activity: peer review of outcome statements
– Students were instructed to make sure their **three outcome descriptions** were written down “on paper or at least something that you can send to another group.”
– The class then carried out a **short peer-review circulation exercise**:
1. Each group wrote down its three outcomes.
2. Groups **rotated** their outcomes to another group.
3. The reviewing group judged each outcome as either:
– **Vague**, or
– **Specific enough**
– Initially, the instructor asked students to note **whether and why** each outcome was vague or specific, but due to time pressure later clarified that they likely **did not have time for full explanations**, so the priority became recording the judgment itself.
– Students were told to **write their judgments directly on the paper**.
– Once papers were returned to their original owners, each group was instructed to:
– Review the judgments made by their peers
– Identify any outcomes marked **vague**
– Put a **star** next to vague outcomes as items that would need to be fixed later
– The instructor noted that vague outcomes would be revisited **next week**, but students should already be aware of which ones needed revision.
## 5. Transition from naming outcomes to testing outcomes
– After the peer review, the instructor shifted from outcome formulation to **probability assessment**:
– How do you go from “here is an outcome that might happen” to “this is the outcome most likely to occur”?
– This framed the rest of the lesson as a move from outcome generation to **structured evaluation**.
## 6. Worked example: Iran and regime change
– To demonstrate the method, the instructor used **Iran** as the main example case.
– The class focused on an outcome roughly framed as:
– Within the next 12 months, will there be **actual regime change** in Iran?
– The instructor defined this more precisely as a situation in which the **Iranian Revolutionary Guard Corps** would **no longer be in control of major government institutions**.
– Before analyzing the specific case, the instructor asked students to identify the current regime:
– An **uncertain student** referenced the **Ayatollahs and the Revolutionary Guard Corps** as the actors in control.
– The instructor then introduced the first major analytic concept: **the baseline**.
## 7. Establishing the baseline
– The instructor defined establishing the baseline as asking:
– **How often does this type of outcome occur throughout history on an international stage?**
– Using the Iran example, the class asked:
– **How often does major regime change actually happen in the world?**
– Students were invited to answer from their accumulated knowledge in international relations rather than formal research for the moment.
– An **uncertain student** answered that it happens **“almost never.”**
– The instructor described this as a **rudimentary version of baseline-setting** and summarized the conclusion as:
– For now, regime change should be treated as **very rare**
– Important clarification:
– The instructor acknowledged that the term **“very rare”** was not yet quantified rigorously enough for professional analysis, but it was sufficient for the day’s exercise.
– Students were then directed to establish a baseline for their own topics using simple language such as:
– very rare
– not very often
– kind of often
– often
## 8. Second worked example: nuclear bomb use and why baseline is not enough
– The instructor introduced a second example:
– *“In the next 12 months will the United States drop a nuclear bomb somewhere in Iranian territory?”*
– Again, the first step was baseline-setting:
– **How often does one country drop a nuclear bomb on another?**
– Student contributions included:
– An **uncertain student** observed that this has happened **“only twice.”**
– Another **uncertain student** added: **“But it was the US.”**
– The instructor used these remarks to make an important analytic point:
– The historical baseline for nuclear weapon use is **extremely rare**
– Yet people were still worried recently that it might happen in the Iran crisis
– The instructor then asked why concern existed despite the rarity.
– An **uncertain student** replied: **“It’s Trump.”**
– The instructor unpacked this by connecting current context to changing probability:
– Recent **patterns of escalation**
– The role of the **United States as the only historical user of nuclear weapons**
– The possibility of **other nuclear powers becoming involved**
– The lesson point was that **baseline alone is insufficient** because present conditions can shift an event’s likelihood upward or downward.
## 9. Probability spectrum: from baseline to dynamic assessment
– The instructor then introduced a visual/analytic model:
– A probability line with **0%** on the left and **100%** on the right
– Students were told to imagine placing an outcome’s baseline somewhere on this line.
– For the nuclear bomb example, the instructor informally suggested something like **5%** as a rough placement, while explicitly noting that:
– there is a **more scientific way** to do this,
– but they did not have time for formal research in this class.
– The instructor stressed that once the baseline is placed, the next task is to identify factors that move the probability **right** or **left**.
## 10. Costly signals vs. prohibitive costs
– The instructor operationalized probability movement using two types of factors:
– **Costly signals**: indicators suggesting the outcome is becoming more likely and should move **to the right**
– **Prohibitive costs**: constraints or disincentives that should move the probability **to the left**
– Students were instructed to draw the **0–100% spectrum line** on paper for their own outcomes and begin placing their baselines.
– For each outcome, they were then asked to identify:
– **Three costly signals** that might increase the likelihood
– **At least two prohibitive costs** that might decrease the likelihood
– The instructor explained that this process is meant to help students determine which factors are capable of **materially shifting** the probability of an outcome.
– Because of time, the class did **not** complete the next step:
– **weighting** these signals and costs
– The instructor explicitly said that the **weighting part**—figuring out which factors matter more and by how much—would be saved for **Tuesday’s class**.
## 11. Applying the framework to the final memo/project
– The instructor connected the day’s concepts directly to the final written assignment.
– Students were told that, for the final memo, they will need to compare **three outcomes**:
– **Status quo**
– **Escalation**
– **De-escalation**
– The purpose of the method introduced today is to assign a probability estimate to each outcome, for example:
– Status quo: 80%
– Escalation: 40%
– De-escalation: 20%
– The instructor noted that in a real professional setting analysts would likely compare **many more outcomes**—perhaps 12 to 20 rather than only 3—but for this class, three outcomes are sufficient.
– Key assignment connection:
– Students should use the baseline + signals + costs framework to judge **which of the three outcomes is most likely**.
## 12. Final memo requirement: warnings and indicators
– The last major concept introduced was the **“warnings and indicators”** section required in the final memo, which the instructor said appears at the end of the syllabus requirements.
– The instructor defined warnings and indicators as:
– **Three conditions that must be met in order for an outcome to occur**
– Important qualification:
– Students **do not need** to establish these three conditions for the **status quo**, because the status quo simply means conditions largely remain the same.
– They **do need** to establish them for:
– **Escalation**
– **De-escalation**
– The instructor described these conditions as **tripwires** or **dominoes**:
– If these things happen, the likelihood of the outcome would rise toward near certainty.
– To make the concept concrete, the instructor used a hypothetical example:
– Imagine waking up in the future to the headline: **“Nuclear bomb dropped on Tehran.”**
– Students should then ask: **what three headlines must have appeared in the previous two or three weeks for that to have happened?**
– Example conditions the instructor suggested:
– **Iran strikes a major target in Europe or the United States**
– **Iran unilaterally rejects all American negotiations for a settlement**
– This was presented as the basis for students’ own warnings-and-indicators sections.
## 13. Closing guidance and next steps
– Because the class was ending, the instructor did **not assign formal weekend homework**.
– However, students were **strongly encouraged** to begin planning the structure of their paper by thinking through:
– their three outcomes,
– the specific signals and costs they would measure,
– and the conditions that must be met for escalation or de-escalation to occur.
– The instructor closed by reminding students that the next class on **Tuesday** would continue with:
– how to **weight** the identified signals and costs
# Student Tracker
> No student names were spoken clearly enough in the transcript to match participation confidently to the roster. Contributions below are therefore recorded as **uncertain**.
– **Uncertain student 1** — Identified **escalation** as one of the three main outcome categories.
– **Uncertain student 2** — Defined escalation in measurable terms as **“more costly.”**
– **Uncertain student 3** — Identified **de-escalation** as the third outcome category.
– **Uncertain student 4** — Defined de-escalation as **“less costly.”**
– **Uncertain student 5** — Characterized the **status quo** as involving **the same costs**.
– **Uncertain student 6** — Contributed to the discussion of regime change by stating that such major changes happen **“almost never.”**
– **Uncertain student 7** — Identified the **Ayatollahs / Revolutionary Guard Corps** as the current ruling power structure in Iran.
– **Uncertain student 8** — Noted that nuclear weapons have been used **only twice**, helping establish the baseline for nuclear bomb use as extremely rare.
– **Uncertain student 9** — Added that those two historical cases involved **the United States**, helping explain why U.S. involvement affects present-day concern.
– **Uncertain student 10** — Contributed the point **“It’s Trump,”** prompting discussion about how current political context can shift probability away from the historical baseline.
– **Several unidentified students (group work)** — Participated in writing outcomes, judging whether peer outcomes were vague or specific, and asking clarification questions about how to apply the probability-spectrum activity to their own cases.
# Actionable Items
## High Priority
– **Revisit vague outcomes next class** and revise them into concrete, measurable statements.
– **Teach weighting of signals and costs on Tuesday**, since this was explicitly deferred due to time.
– Ensure students can clearly distinguish:
– **baseline probability**
– **costly signals**
– **prohibitive costs**
– **warnings/indicators**
## Medium Priority
– Check whether students’ current three outcomes for the memo are:
– aligned to **status quo / escalation / de-escalation**
– expressed in **specific, testable language**
– Reinforce that **status quo does not need a tripwire list**, but escalation and de-escalation do.
– Confirm that students understand that **cost** is the main operational measure being used for escalation/de-escalation.
## Assignment / Project Preparation
– Students should begin organizing the **final memo** around:
– three outcomes,
– a baseline for each,
– three costly signals,
– two prohibitive costs,
– and three warning/indicator conditions for escalation and de-escalation
– Students were **not assigned formal weekend homework**, but were **strongly encouraged** to start structuring the paper before Tuesday.
## Administrative / Context Notes
– Class was shortened because of **Career Day**, which limited time for deeper practice.
– Reduced attendance may have affected pace and group dynamics; no official attendance list was recorded in the transcript.
Homework Instructions:
NO HOMEWORK
The professor explicitly said, “I’m not going to assign you guys a homework assignment for over the weekend,” and the only follow-up was a non-mandatory recommendation to “strongly recommend that for your given cases you begin thinking about how you’re going to structure your paper.”