Do Judges Favor Certain Horses?

Judges mostly agree with each other. But when you turn every judge’s placings into a single comparable number per (judge, horse) pair — how much better or worse a judge ranks a horse than the horse’s average across all judges — a few pairs deviate strongly from the pack.

Across 55 judges and 3,335 (judge, horse) pairs (each pair appearing at least 5 times), the largest bias is 0.57 in place-percentile units. That’s a real signal worth a second look — but with this many comparisons screened at once, some large deviations are expected by chance alone. Treat every row here as a question worth a closer look, not a statement about any judge or horse.

Analysis refreshed Aug 2026 · all figures from the Combined Judges roll-up.

What the metric means

1. Normalise within each judge’s field. For every individual judge’s row, each horse’s placing becomes place_pct = (place − 1) / (n − 1) against the field that judge ranked — 0.0 is a win, 1.0 is last. This makes judges with different class sizes comparable.
2. Pair up judge and horse. For each (judge, horse) pair with ≥ 5 appearances: judge_avg is the mean place_pct that judge gives the horse; global_avg is the horse’s mean place_pct across all judges.
3. bias = judge_avg − global_avg. Negative bias means the judge ranks the horse better than its all-judge average; positive means worse. All values are in place-percentile units (0–1).

4. **Sorted by

   bias

   ** — the tables below show the largest deviations, both directions.

Look, don’t convict. With 55 judges × 3,335 pairs, hundreds of comparisons are screened at once. Even with zero real favoritism, some pairs will show large deviations by chance alone. The strongest pairs here rest on only 5–9 appearances.

Horses a judge ranks better than the pack

A few things jump out:

• SHOOTIN THE BREEZE appears three times in this list — different judges all rank it well below its global average, which is itself a noteworthy pattern (strong horse, or a horse that consistently places well under these judges).
• Several negative-bias pairs have a judge avg of 0.000 — that judge always won the horse. With n=6 that’s a striking record, and exactly the kind of thing that’s interesting to look at more closely.
• The largest raw value is actually positive (+0.566, Baker→The Cookie Club): that judge placed the horse a full 0.57 of the field’s worth worse than the horse’s average — a big outlier in the worse direction.

Judge leniency on maneuver scores

Judges also differ in how generously they score. The figure below is each judge’s mean maneuver score versus the global mean of 0.396 (higher scores are better in this sport’s scoring, so above the line = more lenient). Judges with the largest differences, shown with n ≥ 1,000:

Judges with means above the global line (green, positive diff) are more lenient — they award higher scores than the field average; judges below it (tan, negative) are harsher. Sample sizes vary widely (from ~1,100 to ~15,000), so the biggest-n judges are the most trustworthy comparisons. Judge leniency is about scoring level, which is separate from the per-horse bias tables above.

What’s not here

• Trainer data is too sparse for conclusions — only 7.8% of class-placing rows carry a trainer name — so no trainer table is reported.
• Owner data is available but was left out to keep this focused; the metric is identical if it’s ever wanted.

This is one of a series of data articles built on my equine-data results database — a personal AQHA show-results explorer that analyzes every class in the dataset.