Live · Season 1 · straight from the servers ·
crHow creativity20 gets 8
Every scored map, best first. Each row is the map's weight, how close the best is to the record and where it sits in the field, the clean, jumps and mod factors, the score and what it counts after the decay weight. Add the last column up, multiply by Form, and you have the rating: this is the whole global ranking for one player.
Where it comes from
counts, best first
- 1AmberWadi100.0%
Each bar is score × weight. The best map counts in full, the 25th at 29%.
1 map · by the day the best was run
Σ counts = 8.12 × Form 1.001 · shown as 8
straight from the servers ·
Click a column to sort. score × weight = counts · Σ counts = rating, read in rank order.
Top three carry
100.0%
Every scored map is in the top three right now, so each one is load-bearing.
A new map counts at
× 0.95
Slot #2 is open. A fresh clear scoring 8 would add about 7.60; one that beats an existing row moves in above it and pushes the rest down a weight.
Left on the table
6.32
Falls, restarts and quits take that much off across 1 map. A cleaner run on any of them replaces the scored one.
Headroom in the field
62.28
Getting the 1 map outside the top quarter into the top 10% would add about that much at today's weights.
What decay costs
light bar score · solid bar what it counts
- 1AmberWadi8.12 / 8.12
Where each best sits in its field
one dot per map · left is faster
Gap to the record
closest first · from the map leaderboards
- AmberWadi+168.80s · kevj9
Where the rating comes from
counts by difficulty band
- Easy1 map100%
When the bests were set
the scored run behind each row
Around this player
global #30
- #27LaLavaNotRequired3225
- #28YoYowAshyBhaby2518
- #29sksk229125
- #30crcreativity208you
The right column is the rating between you: what it takes to pass the player above, or the cushion over the one below.
Moves that pay
today's weights · biggest gain first
- 1GlacialAbyssSlot #2 is open at weight 0.950. A clear on GlacialAbyss, Blackline, Highwall scoring like your median map would add about that much.+8
- 2AmberWadiA clean run on AmberWadi at the same time: 15 falls and 5 restarts cost the clean factor.+6
- 3AmberWadi−46.67sBeat sk229 on AmberWadi: 46.67s to shave.+2
What if
one row changed, today's weights, the field's personal bests as the yardstick
Map
New best time
now 3:25.249 · record 0:36.446 by kevj9
Rating would read
8±0
global #30 → #30
- Map score
- 8.12 → 8.12
- Field
- top 100% → top 100%
- Performance
- 0.105 → 0.105
- Clean factor
- 0.60 → 0.60
Drag the slider or tick the clean run to see the move.
44 ranked maps are still unscored this season: Balloon Festival (Easy, best 0:29.443), Beach Day (Easy, best 0:24.300), Blossom (Easy, best 0:19.791), Canopy Village (Easy, best 0:54.098), CoralCove (Easy, best 0:18.748), Ember (Hard, best —), Frostpeak (Medium, best 0:37.099), GlacialAbyss (Insane, best —) and 36 more.
Reading the columns
the full model, with formulas →- Best
- Your fastest valid run on the map this season. Only this run is scored; slower ones never hurt.
- Field
- Where that time sits among every player's best on the map. Top 1% is the fastest hundredth; the median is 50%.
- RDS
- The map's weight: its difficulty base, raised by a longer median best, more falls on best runs, more perfect jumps, a wider spread and a lower clear rate, between 20 and 2,000.
- Perf.
- 0.6 × (record ÷ best)^2.5 + 0.4 × (1 − 0.9 × percentile), pulled toward 0.55 while the field is small, ×1.05 while you hold the record.
- Clean
- 1 − 0.04 × (slips beyond a quarter of the map's expected falls ÷ √segments)^1.3 − 0.03 × go-backs^1.3 − 0.01 per restart, never below 0.6; a run with none pays 1.03. Every extra fall or go-back costs more than the last.
- Form
- One multiplier on the whole rating, within 1 ± 0.05: + flawless clears, − a habit of quitting mid-run, + breadth of the catalogue (full at 50%), + jump efficiency above 70%. Shown under the table.
- Jumps
- 1 − 0.1 × (1 − perfect jumps ÷ your jumps); matching the map's perfect count pays 1.03.
- Mods
- Flat mods (No Falls, Exact Jumps) pay 0.4 of their multiplier, capped at ×1.2. Random debuffs slow the rating clock instead, shown as the effective time.
- Weight
- 0.95 to the power of the position minus one. The best map counts in full, the tenth at 63%, the 25th at 29%.
- Counts
- Score × weight. Σ runs the total down the page; the last row is the rating, rounded to a whole number.
Questions
- What is this page?
- The whole global ranking for one player. creativity20's rating is the sum of the last column: every scored map's score times its decay weight, best map first.
- Why does my best map count in full and the others less?
- Scores are sorted best first and the i-th one is multiplied by 0.95 to the power of i. The best counts at 100%, the tenth at 63%, the 25th at 29%. Only the best 25 maps count at all.
- How does one map get its score?
- RDS × performance × clean × quit × mods. RDS is the map's weight from its difficulty and how long and how spread out the field's times are. Performance comes from where your best sits in the field, with a bonus for margin under the median and for holding the record.
- Does a slow run hurt me?
- No. Only your fastest valid run on a map is scored. Slower clears never lower a score; quits on ranked maps do, through the quit factor, until clears dilute them.
- What happens when a new map scores above an existing row?
- It slides in above it and every row below moves down one weight. The map that fell out of the top slots stops counting. The what-if calculator on the Analysis tab shows the exact move.
- Why did my rating change without a run?
- The field moved. Percentiles and RDS are recomputed every 10 minutes as other players set times, so a map you scored weeks ago can drift up or down.
- How accurate are the Analysis estimates?
- They rerun the same formula with today's weights, placing a hypothetical time among the map leaderboard's personal bests (real times) where the plugin uses effective times. Treat the numbers as close, not exact.
- Do mods change the score?
- Flat mods pay 40% of their multiplier, capped at ×1.2. Timed mods earn time instead: half of the bonus they actually delivered comes off the rating clock, and pace and percentile are taken on that effective time, shown under the best.