Explore Decision Intelligence
Decision Intelligence becomes more useful when its principles are tested against real decisions.
Research & Insights is where Pascal Institute examines evidence, forecasts, decisions, and outcomes in practice. We use data, case studies, experiments, and real-world examples to explore what improves decision quality, and what can quietly undermine it.
The objective is not to prove that every Pascal idea is correct.
It is to test assumptions, measure results, and learn from what the evidence shows.
Research in Practice
Pascal case studies begin with real outcomes, but the outcome is rarely the end of the investigation. We reconstruct what was knowable at the time, examine the underlying evidence, and distinguish what the data establishes from what it merely makes plausible.
Cal Raleigh: When Regression Becomes Something More
Cal Raleigh followed one of the greatest offensive seasons ever produced by a catcher with a dramatic decline in 2026. Regression was predictable. The magnitude and nature of the decline were not.
Using Baseball Savant data, preseason projections, contemporaneous reporting, swing and contact measurements, and academic research on confidence and skilled performance, we examine the interaction between skill, favorable variance, expectations, injury, timing, and performance feedback.
The evidence suggests Raleigh genuinely became a better hitter in 2025, while favorable variance helped amplify that improvement into an historic result. In 2026, his bat speed remained remarkably stable even as his contact deteriorated. Later, his average contact point moved approximately 8.6 inches farther out in front of home plate before a three-home-run game made his resurgence obvious.
The case raises a broader decision-science question: Can luck affect future performance not only through the outcome it produces, but through the expectations and feedback that outcome helps create?
Explore the Cal Raleigh case study →
Brandon Pfaadt: When the Outcome Hides the Process
Brandon Pfaadt allowed no earned runs over 6⅓ innings against the San Diego Padres. The box score suggests a strong performance. But a pitch-by-pitch review of the first inning tells a more complicated story.
Using MLB Film Room video, Baseball Savant data, pitch location, and the catcher’s apparent targets, we examine what happens when execution and outcome disagree. A poorly located pitch can produce an out. A well-executed pitch can produce a ball. Even a strikeout can hide a significant miss.
The case study illustrates a central principle of decision science: outcomes provide evidence, but they do not tell us everything about the quality of the process that produced them.
Explore the Brandon Pfaadt case study →
From Principles to Evidence
A framework can tell us how decisions should be made.
Research helps determine whether those ideas actually work.
Here we explore questions such as:
How accurately do our predicted probabilities match actual outcomes?
Does a measurable forecasting edge persist over time?
When does additional information improve a decision, and when does it simply create noise?
How much does human judgment improve, or weaken, a statistical model?
How should capital allocation change as uncertainty increases?
When should new evidence cause us to update a forecast?
When is abandoning a previous decision better than continuing to invest in it?
These questions rarely have simple answers.
That is precisely why they are worth studying.
What We Explore
Forecasting & Probability
We examine how forecasts perform against reality, including calibration, probability estimation, base rates, scoring methods, and the difference between confidence and accuracy.
A forecast should eventually face the evidence.
Decisions & Outcomes
Good outcomes can follow poor decisions. Bad outcomes can follow good ones.
Decision reviews examine what was known when a choice was made, what alternatives existed, what assumptions influenced the decision, and what can reasonably be learned from the result.
Models & Analytics
Models simplify reality. That makes them useful—and imperfect.
We explore model development, validation, expected value, market comparisons, data quality, human judgment, and the circumstances in which analytical tools improve decisions.
Behavior & Judgment
Not every decision problem is mathematical.
Loss aversion, overconfidence, recency bias, sunk costs, incentives, group dynamics, and other behavioral forces influence how evidence is interpreted and how decisions are made.
Understanding the decision-maker can be as important as understanding the data.
The Pascal Research Approach
Pascal research should follow a few basic principles:
Start with a question, not a conclusion.
Research should investigate an idea rather than search for evidence that confirms what we already believe.
Record expectations before outcomes when possible.
Predictions become more useful when they cannot be rewritten after the result is known.
Measure uncertainty.
We prefer probabilities, ranges, and explicit assumptions to unsupported declarations of certainty.
Separate process from outcome.
The result matters, but it does not tell us everything about the quality of the decision that preceded it.
Report inconvenient results.
Evidence that contradicts an assumption can be more valuable than evidence that confirms it.
Update when the evidence changes.
Changing a conclusion in response to better information is part of the process, not a failure of it.
The Laboratory: Sports
Sports provide an unusually useful environment for studying decisions under uncertainty.
Events occur frequently. Outcomes are observable. Historical data is abundant. Markets publish probabilities through prices. Forecasts can be recorded before outcomes occur. Results arrive quickly enough to create repeated feedback.
Most importantly, the environment is unforgiving.
If we estimate a team has a 55% probability of winning, reality eventually gives us enough observations to evaluate whether forecasts like it actually occur about 55% of the time.
That makes sports more than an application of Decision Intelligence.
It makes sports a laboratory for studying it.
The principles developed there can then be examined in investing, business, forecasting, resource allocation, and other environments where feedback is slower or less obvious.
What Happens When We’re Wrong?
Research is most valuable when reality disagrees with us.
A model that underperforms may reveal a missing variable.
A forecast that was too confident may expose poor calibration.
A profitable strategy may turn out to have benefited from randomness.
A decision that initially appeared unsuccessful may prove reasonable when evaluated against the information available at the time.
Instead of hiding those cases, Pascal Institute intends to examine them.
The question is not simply:
Did it work?
It is:
What did we expect, what actually happened, why were they different, and what should change next time?
That turns outcomes into feedback.
And feedback creates the opportunity for compounding improvement.