Probability and Ensembles

Relates probabilities, ensembles, and statistical averages in physical systems.

Conceptual context

An $E$ is an ideal collection of systems prepared under the same $C$, while individual copies may occupy different $i$. Instead of tracking one copy, statistical thermodynamics asks what fraction of copies occupies each state and how that population produces observable quantities. The $p_i$ is not a vague guess: it represents the part of the $E$ assigned to one compatible state. With that idea, many microscopic possibilities become one controlled macroscopic prediction.

🟢 Essential level

An $E$ lets us imagine many copies of the same physical system, all prepared under the same macroscopic preparation conditions, but distributed among different $i$. The $p_i$ tells what fraction of those copies belongs to each microscopic possibility. If many copies have the same $A_i$, that value contributes more strongly to the $A_{prom}$. If the values are widely spread, the $\sigma_A$ warns that the average does not describe the whole microscopic variety.

The key idea is to separate preparation from outcome: preparation fixes the allowed set, while probabilities describe how that set is populated. Thus the $E$ does not replace the real system; it offers an orderly way to translate microscopic uncertainty into macroscopic physical prediction. This perspective avoids a common classroom error: believing that probability belongs to one isolated copy. It belongs to the full distribution of the ideal set, the statistical parliament of the model. This notion of ensemble is what later makes it possible to define the partition function and the different statistical collectives.

🔵 Formal level

Basic formulas

Formally, the $E$ contains $N_{ens}$ ideal copies subject to the same $C$. The $p_i$ of an $i$ can be read as an ideal frequency when the $n_i$ is compared with the total number of copies.

Probability as ensemble frequency

Because every copy must occupy some compatible state, probabilities must be normalized.

Ensemble normalization

For an observable quantity, each state carries a $A_i$. The $A_{prom}$ is obtained by weighting each value with its $p_i$, not by taking a simple average when states are not equally probable.

Ensemble average of a quantity

Real student questions

Does the $E$ physically exist? Not as a box full of real copies. It is an ideal representation of many equivalent preparations.

Why not use only the real state? Usually the exact $i$ is unknown, and macroscopic predictions depend on populations.

Is $p_i$ a frequency? It can be interpreted as an ideal frequency when $N_{ens}$ is very large and copies are equivalent.

Is an average always enough? No. The $A_{prom}$ may match for two very different distributions. The $\sigma_A$ says whether individual values are close to or far from the average.

Final synthesis

An $E$ turns many compatible microscopic possibilities into a probability distribution. With $p_i$, $A_{prom}$, and $\sigma_A$, physics moves from uncertain individual states to controlled macroscopic predictions.