Skip to content
Duelwise
Sign in

The AI, measured

One opponent per game. Strength numbers come from tests or counted matches; an unmeasured level is in training.

Crosscurrent against people

Human win rate vs Marlow, by week

0%15%30%Sep 10Sep 17Sep 24Oct 1Oct 8Oct 15Oct 22Oct 29Week of Oct 29: 0% (0 of 9 matches), too few matches to trust0%
Each point is a week of finished human matches against Marlow.solid: 30 or more matches. Dashed and hollow: too few matches to trust.
Crosscurrent is a scripted solver with nothing to train.

Crosscurrent is a scripted solver, not a learned model: each level is a different algorithm, so there is nothing to train. Your match is recorded for your report and the board.

  1. v1, champion Nov 3Each level beat the level below, and Pip beat random legal play, from both seats over 400 games a seat. No human matches used.
What is site Elo?

Site Elo is a strength number we measure by playing the levels against each other. Easy is pinned at 1000 and every other level is placed from its results against the one below. A gap of 200 points means the stronger side is expected to win about three games in four (76 in 100, counting a draw as half). It is not a FIDE or chess.com rating, so it says nothing about how you would score against people.

Crosscurrent agent

Scripted solver: does not learn from site matches, because there is nothing to train.

Humans win vs Hard, last 7 days

no matches

No counted Hard matches this week.

Current version

v1

promoted Nov 3

Matches vs humans

11

11 passed the filters

Champion strength

2085

site Elo, measured

Human win rate by level

weekly, last eight weeks, counted matches only

Not enough weekly results to chart yet.

Levels, measured strength and limits
Easy79.6% (76.6% to 82.2%, 800 games)score vs random1000
Medium85.6% (83.0% to 87.8%, 800 games)score vs easy1309
Hard94.6% (92.8% to 96.0%, 800 games)score vs medium1807
Champion83.2% (80.4% to 85.6%, 800 games)score vs hard2085

Head to head

stronger level's score against the level below, with its confidence interval

  • Medium vs Easy85.6%stronger wins, 800 games83.0% to 87.8%
  • Hard vs Medium94.6%stronger wins, 800 games92.8% to 96.0%
  • Champion vs Hard83.2%stronger wins, 800 games80.4% to 85.6%

Version history

  1. v1All levels measured. First version. No human matches used. Nov 3

Human matches, Crosscurrent

where they go

  • Played17
  • Finished11
  • Passed every filter11
  • Allowed for training0
  • Used for training0

Nothing is used for training. 0 opted-out matches never leave the database. 11 not used: this agent is scripted and has nothing to train.

  • 5 resigned
  • 1 abandoned

What changed

measured, not described

No measured change is recorded for this version.