Ranking Teams: Build a Power Ranking with rank()

What happens when two teams tie? Sorting just shrugs and picks an order; ranking has to answer it explicitly — and the tie-breaking option people reach for first is usually the wrong one. We'll build a power ranking of MLB teams by run differential with pandas' rank(), walk through which tie-breaker actually does what you mean, and finish with a lollipop chart — a cleaner alternative to bars when the ordering is the entire point.

This builds on Summary Statistics and Distributions. The data is the bundled sample_standings.csv (real 2023 MLB standings), so it runs offline.

1. Turn a column into a ranking

   rank() assigns each row its position. By default it ranks ascending (smallest = 1), so for a power ranking we rank descending — the biggest run differential is number one.

   import pandas as pd
   
   df = pd.read_csv("sample_standings.csv")
   df["Rank"] = df["RunDiff"].rank(ascending=False, method="min").astype(int)
   print(df.sort_values("Rank")[["Rank", "Team", "W", "RunDiff"]].head(8).to_string())

   Power ranking by run differential

       Rank     Team    W  RunDiff
   0      1   Braves  104      231
   2      2  Dodgers  100      207
   3      3     Rays   99      195
   7      4  Rangers   90      165
   1      5  Orioles  101      129
   5      5   Astros   90      129
   10     7    Twins   87      119
   14     8   Padres   82      104

   The Braves top the board at +231, and the order tracks quality closely. Look at the tie: the Orioles and Astros both finished at +129, so both get rank 5 — and the next team is rank 7, not 6. That's the method="min" rule at work, and it's the convention sports standings use.

2. Ties: pick the right method

   How rank() breaks ties is the setting people forget. The main options:

   • method="min" — tied rows share the lowest rank (two 5ths, then a 7th). The standings convention.
   • method="dense" — like min, but the next rank isn't skipped (two 5ths, then a 6th).
   • method="average" — the default; tied rows get the average of the ranks they span (two would-be 5th/6th rows both become 5.5).

   None is "correct" in the abstract — they answer slightly different questions — but choosing without thinking is how rankings end up subtly wrong.

3. Draw a lollipop chart

   A lollipop chart is a bar chart on a diet: a thin stem from zero to each value, capped by a dot. When the message is the ranking rather than precise magnitudes, it's less ink and easier to read than fat bars.

   import matplotlib.pyplot as plt
   
   top = df.sort_values("RunDiff", ascending=False).head(12).sort_values("RunDiff").reset_index(drop=True)
   y = range(len(top))
   fig, ax = plt.subplots(figsize=(8, 6))
   ax.hlines(y=list(y), xmin=0, xmax=top["RunDiff"], color="#C2B7A1", lw=2)   # the stems
   ax.plot(top["RunDiff"], list(y), "o", markersize=9)                        # the candy
   ax.set_yticks(list(y)); ax.set_yticklabels(top["Abbr"])
   ax.set_xlabel("run differential")
   fig.savefig("power_lollipop.png", dpi=144, bbox_inches="tight")

   

   Two layers make the lollipop: hlines draws the stems from zero out to each value, and plot(..., "o") drops the dots on the ends. Sorting the slice ascending puts the best team on top, and the eye runs straight down the dots to read the ranking.

Troubleshooting