Glossary
Definitions for every metric, term, and concept used across the Equine Intelligence platform — from performance analytics to bloodline scoring.
Adjusted clear %
A horse's or rider's clear-round percentage after it has been corrected for how few rounds it is based on. A raw 100% off 3 rounds becomes a more realistic figure once the small sample is accounted for. See Shrinkage. The raw percentage is always still shown alongside it.
Age curve
The typical pattern of how clear rate changes with a horse's age across the whole population. Used to flag horses performing ahead of, or behind, what's normal for their age.
Clear Round %
The percentage of competition rounds in which a horse or rider completed the course without any jumping or time faults. One of the primary performance indicators on the platform.
Clear streak
The number of consecutive clear rounds a horse or rider has jumped. We show both the current streak and the longest streak on record.
Confidence grade
A simple label — Thin, Medium or High — showing how much data a stat is based on. Thin means very few rounds (treat with caution); High means enough rounds to rely on. Thin = under 5 rounds, Medium = 5–14, High = 15+.
Confidence interval
The range a true percentage is likely to sit within, given the sample size. A horse shown as "75% (41–93%)" has a best estimate of 75%, but with few rounds the real figure could plausibly be anywhere from 41% to 93%. A wider range means less certainty.
Consistency Score
A derived metric that rewards stable, repeatable performance over time — distinguishing horses that regularly go clear from those with volatile results.
Difficulty index
A 0–100 rating of how hard a class or course was, based on how few of the field went clear. A class where only 14% went clear scores far higher than one where 56% did. Used to give more credit for clears achieved against tough courses.
Difficulty-adjusted clear %
A clear-round percentage that gives more weight to clears jumped in harder classes (see Difficulty index), so a horse competing at the top level isn't out-ranked by one clearing easy tracks.
Empirical Bayes
The statistical method behind Shrinkage. In plain terms, it blends each horse's own record with the league average, leaning more on the league average when the horse has few rounds. It is standard practice in sports and ratings systems.
Fault Average
The average number of faults accumulated per round across all scored starts. Lower is better; used alongside clear round % to assess overall accuracy.
Form curve
A rolling view of a horse's or rider's recent clear rate over their last several rounds, with an arrow showing whether they are trending up or down (momentum).
Jumping faults
Faults from knocking down a rail or a refusal (4 faults each), as distinct from Time faults. Splitting the two shows whether a horse is careless or simply slow.
Momentum
Whether a horse or rider is improving or declining recently, based on the change in their rolling clear rate (or rating) over a chosen time window.
Non-Completion Rate
The percentage of starts where a horse or rider did not complete the course — including refusals (R), eliminations (EL), withdrawals (WD), and did-not-starts (DNS). A key indicator of reliability.
Percentile
Where a horse or rider ranks relative to a comparison group, as a 0–100 position. The 90th percentile means better than 90% of the group (e.g. of the same age, height or region).
Regression to the mean
The tendency for extreme results from small samples to drift back toward average as more data comes in. A horse that's gone clear 4 from 4 will almost always settle to a lower rate over a full season. This is why Shrinkage is applied.
Reliability score
A 0–100 measure of how dependably a horse delivers, combining its clear rate with how consistent it is across conditions.
Rounds Completed
The total number of competition rounds a horse or rider has data for on the platform. Higher round counts increase the reliability of all derived metrics.
Sample Size
The number of data points (rounds) underpinning a metric. Metrics with low sample sizes are flagged as less reliable and shown with a reduced confidence indicator.
Shrinkage
A correction that pulls a horse's or rider's percentage toward the league average by an amount that depends on how many rounds back it up. Few rounds are pulled hard toward average; many rounds barely move. It stops lightly-tested horses with a perfect record from topping the rankings ahead of proven performers. Think of it as every horse jumping a set of "average warm-up rounds" before its own record is fully trusted: adjusted % = (clears + 0.31 × warm-up rounds) ÷ (rounds + warm-up rounds).
Style profile
A horse's or rider's tendency split into carefulness (rails down) versus pace (time faults), placing them into tactical types such as "careful and quick" or "fast but careless".
Suitability score
How well-matched a horse is to particular conditions — height band and arena surface — based on where it has produced its best results.
Time faults
Faults for exceeding the time allowed (typically 1 fault each), as distinct from Jumping faults. A horse collecting time faults is careful but slow.
Top-3 Rate
The percentage of starts where a combination finished in the top three placings. Used as a proxy for competitive excellence within the field.
Top-5 Rate
The percentage of starts where a combination finished in the top five placings. A broader measure of competitive relevance within a class.
Win Rate
The percentage of starts resulting in a first-place finish. Particularly relevant for jump-off and speed class analysis.
Arena Type
Classification of the competition surface — Grass, Artificial Indoor, or Artificial Outdoor. Performance can vary significantly by arena type, and filters allow isolation of each.
Elite points
The weighted score used in the Elite Rankings points table (not EI Rating). Calculated per result as: height × (placing + clear bonus − faults) × field size × competition strength, then weighted by recency. Recent results count more heavily (decay ~37% per 365 days), rewarding current form. Accumulates over 12 months without shrinkage — this is a points total, not an ability estimate.
Elite Rankings
A 12-month points ranking of the top tier only (1.40m+ classes, first round only). Accumulates Elite Points across the period to reward both results and campaigning volume at the highest level. Recent results are weighted more heavily than older ones (recency decay). DISTINCT from EI Rating: Elite Rankings answer 'Who has won the most at the top level over the last 12 months?', while EI Rating estimates true per-start ability across all levels. A horse can have a high EI Rating but lower Elite points if it doesn't campaign much, or vice versa.
EI Rating
An overall strength rating for a horse or rider, based on points earned from each competition round. Each round's points come from placing, clear-round bonus and faults, then are weighted by four factors: the class's Difficulty Index (how hard the course was), Field Strength (how strong the entries were), field size, and the class height. Height is an explicit factor in two ways: (1) placing and clear-round bonuses are scaled by class height so a win at 1.50m is worth roughly 2.5× a win at 1.00m, and (2) a height multiplier (0.42–1.80 by band) further weights the overall score. This ensures performing at higher levels decisively outranks dominating low classes — a strong result at 1.40m+ will outrank a perfect record at 1.10–1.20m. Ratings are shrunk toward the population mean using empirical-Bayes shrinkage (prior weight 25, stronger for thin samples), so provisional horses with few rounds cannot sit above settled horses with extensive records. A minimum of 5 scoring rounds is required to appear in the main ranking. A height ceiling caps the maximum achievable rating by the highest level a competitor has genuinely competed at — a horse that has only ever jumped 1.10m cannot reach the top rating band, while 1.40m+ competitors can reach the full range. Higher is better. Each rating is shown with a confidence badge — Settled or Provisional (see Confidence) — so you can see how established it is. DISTINCT from Elite Rankings: EI Rating estimates true per-start ability across all competition levels and time (not just 12 months, not just 1.40m+), while Elite Rankings is a 12-month points total rewarding campaigning at the top level. The method builds on established sports-rating techniques such as Glicko (widely used in chess), combined with empirical-Bayes shrinkage to keep lightly-tested combinations from being over- or under-rated.
Height Band
A grouping of classes by jump height (e.g. 1.10m, 1.20m, 1.30m, 1.40m+). Rankings and analytics can be filtered to show results within a specific height band or 'height and above'.
Height Mode
Controls whether a height filter shows only an exact height (Exact) or that height and all higher classes (Height+). The default is Height+ to capture a horse's full upper-level record.
Horse Rankings
Multi-metric leaderboards ranking horses by performance across a selected height band, arena type, and time period. Sortable by clear round %, fault average, consistency, and more.
Min Rounds
A filter that sets the minimum number of competition rounds required for a horse or rider to appear in a ranking list. Prevents low-sample outliers from distorting leaderboards.
New Entrant
A horse that has appeared in the rankings for the first time within the selected tracking period — either due to a debut or returning after a gap with sufficient new rounds.
Confidence (Settled / Provisional)
The badge shown next to an EI Rating. "Settled" means the rating is based on enough rounds to rely on; "Provisional" means few rounds have been recorded and the rating may move significantly as more results come in. A ± figure (labelled "conf") may also appear — a smaller figure means higher confidence.
Rider Rankings
Leaderboards ranking riders across all horse combinations, aggregating performance by height, region, and season.
Riser / Faller
A horse that has moved significantly up (Riser) or down (Faller) in the rankings within the selected period. The platform surfaces the biggest movers to highlight in-form and out-of-form horses.
Season
The New Zealand show jumping season, typically running from approximately August to May. The platform segments data by season to allow year-on-year performance comparison.
Strength of schedule
A summary of how tough the classes and fields a horse or rider has faced have been overall. See Difficulty index and Field strength.
Brier Score
A calibration metric used internally to evaluate model accuracy. Lower Brier Scores indicate better probability calibration — predictions that closely match actual outcomes.
Calibration
A check of how accurate our predictions have been. If we say a group of combinations each have a 70% chance of going clear, about 70% of them should actually go clear. Good calibration means the model can be trusted.
Class Strength Score
A composite score (0–100) assessing how competitive a class was, factoring in field size, clear round rate relative to the platform average at the same height, and the presence of elite-level combinations.
Confidence Score
A model-level indicator of how much data underpins a given prediction. Higher confidence indicates more historical rounds for the combination at or near the class conditions.
Difficulty Delta
The difference between the platform-wide average clear round rate at a given height and the actual clear rate of a specific class. A positive delta means the class was harder than average.
Fair Odds
A decimal odds representation derived from the model's predicted win probability (e.g. 20% win probability = 5.0 fair odds). Not betting odds — a normalised comparison tool.
Is Favourite
A flag indicating the entry with the highest predicted win probability in a Performance Market. Analogous to the market favourite in a race.
Model Version
Each prediction is tagged with the version of the underlying statistical model used to generate it. This allows the platform to track accuracy improvements over time.
Performance Market
A model-generated forecast for an upcoming class, presented in a market-style format. Each entry receives probability estimates for winning, placing, and going clear.
Predicted Clear %
The model's estimated probability that a combination will complete the course without any jumping faults.
Predicted Top-3 %
The model's estimated probability that a combination will finish in the top three placings.
Predicted Win %
The model's estimated probability that a given combination will win the class, expressed as a percentage.
Suitability Score
An estimate of how well-matched a horse and rider combination is to the specific class conditions — height, arena type, and venue — based on historical performance patterns.
Upset alert
A flag on a pre-event projection highlighting a combination that could beat much higher-rated rivals.
Bloodline Alias
A mapping between a variant or raw imported name and its canonical counterpart. Aliases are reviewed and approved by platform administrators to ensure data integrity.
Breeder
The individual or stud responsible for breeding the horse. Breeder analytics aggregate performance across all horses from a given breeder to assess production quality.
Canonical Name
The standardised, authoritative spelling of a sire, dam, or breeder name used throughout the platform for consistent grouping and analytics — resolving variant spellings from source data.
Cross Predictor
An experimental tool that projects estimated performance for a theoretical breeding cross, blending the statistical profiles of a nominated sire and dam sire with platform-wide baselines.
Dam
The mother of a horse. Dam analytics track performance across a mare's offspring to identify maternal line contributions to competitive success.
Dam Sire
The sire of the dam (the horse's maternal grandfather). Dam sire analytics are widely used in bloodstock evaluation to assess the influence of the female line.
Elite Offspring
Offspring from a bloodline that have competed in 1.40m+ classes on the platform — a measure of the bloodline's production of top-level jumpers.
HHI (Herfindahl–Hirschman Index)
A statistical concentration measure used in the Population Profile to quantify how diverse or dominated a bloodline population is. A high HHI signals concentration among a few bloodlines.
Nick
A sire and dam-sire combination whose offspring perform unusually well together — a bloodline "match". Surfaced by the Cross Predictor. Only shown where there is enough data to be meaningful.
Offspring Count
The number of horses from a given sire, dam, or breeder that have recorded results on the platform.
Population Profile
A demographic breakdown of the horse database by sire, dam sire, and breeder representation — showing market share, concentration, and genetic diversity indicators.
Progeny performance index
A rating of a sire, dam or breeder based on how their offspring perform, corrected for small samples so a single good result doesn't outrank a proven producer. See Shrinkage.
Sire
The father of a horse. Sire analytics aggregate the competition performance of all offspring on the platform to assess a bloodline's influence on jumping ability.
Sire specialty
Whether a sire's offspring tend to excel at scope and big tracks (1.40m+) or at carefulness and rideability at lower heights.
Field Strength
A measure of the competitive quality of entries in a class or event, based on the historical performance ratings of participating combinations.
Class Difficulty
A metric derived from the clear round rate across a class. Lower clear rates indicate a more demanding course relative to the field.
Standout Performance
A result that significantly outperforms a combination's historical baseline — identified algorithmically and highlighted in event analytics.
Standout Delta
The numerical difference between a horse's predicted rank and its actual finishing position. A large positive delta (e.g. predicted 8th, finished 1st) flags an exceptional outperformance.
Elite Density
A measure of how many elite-level (1.40m+) combinations competed in a class or event. Higher elite density indicates a higher-quality field.
Track Pressure
An indicator of how demanding a course was relative to typical results at the same height and venue. High track pressure suggests a technically challenging or tight-time set course.
Home vs away
A comparison of a horse's or rider's results in their home region against results at away venues, showing who "travels well".
Venue
The physical competition location. Venue-level analytics track how performance patterns differ across locations, accounting for arena surface, course design tendencies, and typical field composition.
Comparison Tool
A side-by-side benchmarking interface that allows direct statistical comparison of two horses or two riders across any combination of height, arena type, and season filters.
Shortlist
A user-created collection of horses or riders saved for ongoing monitoring — useful for buyers evaluating multiple prospects or trainers tracking a group of horses.
Watchlist
A personal list of horses or riders to follow closely, with optional notes. Distinct from a Shortlist in that it is a lighter-weight tracking tool without structured evaluation.
Private Notes
User-created annotations attached to a horse, rider, or shortlist item. Private Notes are only visible to the account that created them.
My Stable
A personal dashboard showing all horses and riders a user has added to their favourites, with quick access to their latest performance data.
Saved Comparison
A stored comparison workspace that preserves the selected entities and filter settings so a user can return to the same analysis without reconfiguring it.
Match Up Tool
An admin-facing head-to-head comparison panel for any two horses or riders. Calculates a statistical verdict and highlights advantages — designed for commentary and social content creation.
Performance Archetype
A descriptive label automatically assigned to a horse based on its result patterns — e.g. 'Elite Contender', 'Consistent Clear-Round Type', 'Grass Specialist', 'Rising Step-Up Horse'. Archetypes summarise a horse's competitive identity in plain language.
Horse Suitability Score
A score estimating how well-matched a horse is to specific competition conditions (height, arena type) based on its historical performance profile at those conditions.
Career Arc
A season-by-season timeline of a horse's competitive history, showing clear round rate, maximum height competed, wins, and key milestones such as debut, first elite start, and first clear at 1.40m+.
Head-to-head
A direct record of how two horses or riders have performed when they met in the same class — who placed higher, and how often.
Milestone
A notable achievement automatically detected from results, such as a first 1.40m clear, a longest clear streak, a season-best, or a major upset.
Partnership chemistry
How much better (or worse) a particular horse-and-rider combination performs than you'd expect from each partner's individual record. A positive figure means the pairing brings out the best in each other.
Surface specialist
A horse that performs notably better on a particular arena surface, such as grass or artificial outdoor.
Entitlement
A permission flag tied to a subscription tier that controls access to specific platform features. For example, 'can_view_elite_rankings' is granted to Pro and Elite subscribers.
Published Result
A result record that has been validated and made visible in rankings and analytics. Imported results are held in review before publishing.
Import Batch
A set of results imported from a source file (typically Evo Events data) in a single operation. Batch imports are reviewed and published by platform administrators.
Predictions Game
A monthly pick-em game where subscribers allocate virtual credits across Performance Markets. Picks are settled after results and the monthly leaderboard ranks players by credit balance.
Intelligence Report
A formatted PDF document generated by the platform, containing curated analytics for a horse, rider, or event. Reports can be client-safe or include internal-only data.
Social Tile
A branded 1080×1080 image generated from live platform data for use across social media channels.
NZ Show Jumping
New Zealand show jumping competition as structured under Equestrian Sports New Zealand (ESNZ). The platform draws data from NZ competition results, primarily sourced via Evo Events.
Launch Window
An admin-only page for generating screenshot-ready social media tiles (1080×1080 PNG) to promote platform features and statistics ahead of or during the public launch.
Ranking Movement Tracker
An Elite feature page showing which horses are rising, falling, or newly entering the rankings over a selected period (7, 14, or 30 days), filterable by height band.
Game Market
A Predictions Game instance linked to a specific Performance Market forecast. Pro subscribers place virtual credit picks on game markets before they close.
Credit Wallet
Each Predictions Game participant starts each month with 100 virtual credits. Winnings from correct picks are added to the wallet; losing stakes are deducted. The monthly leaderboard is ranked by final credit balance.
This glossary is maintained by the Equine Intelligence team. Terms and definitions are updated as the platform evolves.