Seasonal

Buy league-early, sell league-late

LeagueAllflame

Allflame — day 38ScreenedScreened 343 strategies over 89 currencies.

Buy early in a league, sell later. Every figure below pays the worst end of the range at both ends.

Strategies
74
passed the screen
Screened
343
hypotheses tested
Best worst case
+553%
the weakest league of the top strategy

How much of this is noise

7%false discovery rate

About 7% of the table is expected to be noise, and there is no way to say which — so spread across it rather than backing one name.

LeaguesFound against chanceFoundChance
4231.26
5100.24
630.03
What this is not
  • Not a prediction of the next league. It is what happened in the past ones, and a league that starts with an unusual currency breaks the pattern.
  • Not a per-name claim. No single strategy clears a Bonferroni bar; four leagues cannot produce a p-value that small however real the effect.
  • Not free capacity. The trade needs divine on day 3 of a league, which is exactly the population that cannot be large.

Hinekora's Lock

Per league

1.5×12×20×Mirage 9.4×Mercenaries 9.6×Allflame 9.3×Keepers 6.5×d0d30d45

Buy window still open

Nothing left to enter on day 38.

Every strategy here buys in the first week of a league. This is a table for the next league, not for today — which is the honest answer rather than an empty page.

Buy window has passed

Kept visible so the next league can be planned from them.

24 currencies, best window each
BuyWindowWorst leagueTwo bounds, one measurementThe bold number is the screen: every league profitable while paying the top of the band going in and taking the bottom coming out. That is a taker, and it is deliberately the harsh end — a strategy that survives it has margin for error. The green number is the same hold for someone who posts orders and waits, which is the other end of the same measured band. On a hold of weeks, waiting hours to fill costs nothing, so it is not an optimistic fudge here the way it would be on an instantaneous cycle. Across the strategies that sweep at both bounds the worst case rises a median of 52.8 points. The screen still runs on the bold number: 84 further strategies sweep only at the favourable end and are not shown, because selecting on the bound that flatters them is how this project has been fooled twice.MedianLeaguesCapacity
Hinekora's Lock +7 more windowsday 0 → 30+553%+746%+835%4/4~256
Reflecting Mist +4 more windowsday 7 → 45+360%+525%+721%4/4~879
History +3 more windowsday 3 → 30+217%+300%+249%4/4~572
Veiled Exalted Orb +5 more windowsday 3 → 30+180%+250%+280%6/6~706
Mirror of Kalandra +4 more windowsday 3 → 30+178%+206%+207%4/4~13,489
Unrequited Love +4 more windowsday 7 → 45+123%+141%+204%4/4~5,245
Mirror Shard +5 more windowsday 3 → 45+111%+129%+188%4/4~2,743
The Price of Devotion +3 more windowsday 3 → 30+107%+200%+161%4/4~475
The Immortal +4 more windowsday 7 → 45+86%+138%+157%4/4~2,919
House of Mirrors +4 more windowsday 7 → 45+73%+99%+141%4/4~26,315
Deafening Essence of Scorn +2 more windowsday 7 → 30+73%+156%+114%5/5~356
Orb of Dominance +1 more windowsday 7 → 30+60%+156%+125%4/4~768
Exceptional Eldritch Ichor +1 more windowsday 7 → 30+50%+175%+62%4/4~243
Awakener's Orbday 7 → 45+50%+129%+116%4/4~293
Exceptional Eldritch Ember +1 more windowsday 7 → 21+41%+210%+74%4/4~217
Seven Years Bad Luckday 7 → 30+33%+150%+67%4/4~500
Horned Scarab of Awakeningday 7 → 45+32%+200%+69%4/4~304
Grand Eldritch Ember +1 more windowsday 7 → 21+25%+85%+30%4/4~693
Deafening Essence of Spiteday 7 → 45+24%+86%+60%4/4~518
Dense Fossilday 7 → 45+24%+69%+25%4/4~331
Syndicate Medallionday 7 → 21+20%+75%+50%4/4~2,044
Essence of Horrorday 7 → 45+11%+86%+29%4/4~814
Fine Delirium Orbday 7 → 30+10%+63%+13%4/4~431
Vivid Crystallised Lifeforce +1 more windowsday 3 → 45+6%+37%+158%5/5~1,715