Odds to Finals

The odds calculation estimates your chances of making finals at each event. It blends two signals: how dancers at your strength level have historically performed at that event, and how hard the event is overall.

The formula

adjusted_odds = calibrated_rate × weight + base_rate × (1 - weight)

calibrated_rate — Among dancers in your strength tier at this event, what percentage historically made finals?
base_rate — Out of everyone who entered this event, what percentage made finals?
weight — How much to trust your individual strength vs the overall average (varies by level)

Step by step: a worked example

Let's walk through a real calculation for an intermediate follower at an event with 50 entrants and typically 6 finalists.

Step 1: Compute strength percentile

Your strength score is compared to every other dancer who has entered this event at your level and role in past years. You're ranked using a percentile:

0.0 = weakest historical entrant
0.50 = right in the middle
1.0 = strongest historical entrant

Say your strength score puts you at the 76th percentile — stronger than 76% of historical intermediate followers at this event.

Step 2: Assign a strength tier

Dancers are grouped into 10 tiers based on their percentile:

Tier	Percentile range	Meaning
Top 10%	90th–100th	Strongest entrants
80–90%	80th–90th	Very strong
70–80%	70th–80th	Strong (you're here at 76th)
60–70%	60th–70th	Above average
50–60%	50th–60th	Middle of the pack
40–50%	40th–50th	Below average
30–40%	30th–40th	Weaker
20–30%	20th–30th	Weak
10–20%	10th–20th	Very weak
Bottom 10%	0–10th	Weakest entrants

At the 76th percentile, you land in the 70–80% tier.

Step 3: Look up the calibrated rate

For each tier, we know how often dancers in that tier have historically made finals at this level. This is computed across all events and years for your level and role:

Tier	Entrants seen	Made finals	Historical finals rate
Top 10%	245	189	77.1%
80–90%	287	153	53.3%
70–80%	312	94	30.1%
60–70%	289	45	15.6%
50–60%	301	28	9.3%
...	...	...	lower

Your tier's calibrated rate is 30.1% — historically, 30.1% of intermediate followers in the 70–80th percentile made finals.

Step 4: Compute the base rate

The base rate is the overall finals rate for this event, averaged across all years it's been held:

base_rate = average(finalists / entrants) across all years

For our example event: 6 finalists / 50 entrants = 12.0% on average. Recent years (2024+) are weighted more heavily in case the event has grown or shrunk.

Step 5: Blend with the strength weight

The strength weight for intermediate followers is 0.61 — meaning 61% of your odds come from your personal strength tier, and 39% from the event average.

adjusted_odds = 30.1% × 0.61 + 12.0% × 0.39
             = 18.4% + 4.7%
             = 23.1%

Your estimated chance of making finals at this event is 23.1%.

How much does your strength matter?

The strength weight varies by level and role. At higher levels, individual strength is a better predictor. At lower levels, there's more randomness.

Level	Leaders	Followers	What this means
Newcomer	53%	40%	Roughly half strength, half average
Novice	45%	40%	Similar — lots of variance at this level
Intermediate	53%	61%	Strength starts to dominate for followers
Advanced	76%	93%	Strength is the main predictor
All-Star	64%	84%	Strong predictive power
Champion	70%	100%	For champion followers, it's entirely about strength

At the champion follower level, the base rate is ignored completely — your odds are 100% determined by how you rank among historical competitors.

How these weights were chosen

The weights aren't arbitrary. They were tuned using a technique called Brier score minimization — a standard method for calibrating probability predictions.

The process:

Take all historical competition data and split it into training (pre-2024) and test (2024–2025) sets
For each possible weight from 0.00 to 1.00 (in steps of 0.01), calculate what the predictions would have been
Measure prediction accuracy: for each dancer, compare the predicted probability to what actually happened (1 = made finals, 0 = didn't)
The Brier score is the average of (prediction - actual)² — lower is better
Pick the weight that minimizes Brier score for each (level, role) combination

The result: these weights produce predictions that are 2.2% more accurate than using a uniform weight of 0.6 for everyone.

Edge cases

No calibration data available: If there aren't enough historical entrants in your strength tier at a particular event, the calculation falls back to the event's base rate.

New events: For events with only 1–2 years of history, the base rate is noisier. The calculation still works but the predictions are less reliable.

Event renames: Events that changed names (like "Novice Invitational" → "H-Town Throw Down") are treated as the same event — all historical data from both names is combined.

Future events: Your odds page includes upcoming events from the WSDC calendar. For events you've never competed at, the calculation uses whatever historical data exists for that event.