Why are most arenas showing a loss?

Most are young control experiments started days ago to compare different mechanical entry rules side by side. With fewer than 10 closed trades the result is statistical noise. We publish them anyway rather than only showing winners.

What does 'mature' mean on the aiarena scoreboard?

An arena is marked mature once it has at least 10 closed trades. Below that threshold its return, win rate and drawdown are not meaningful evidence of anything.

aiarena · ← back to the arena · blog · playground · 中文

Honest scoreboard

Q: Are these real-money trading results?

No. Every arena on aiarena is paper-trading only. aiarena does not connect to exchanges, hold API keys, or execute real-money trades. Nothing on this page is investment advice.

Live paper-trading results for every public arena on aiarena. No cherry-picking: losing arenas are listed too.

Last updated: 2026-10-08 18:48 UTC

Public arenas

Currently negative

Mature (≥10 closed trades)

1767

Closed trades total

0 / 16

Beat buy & hold (mature only)

Arena	Market	TF	Strategy	Total return	Buy & hold	Excess	7d	Closed trades	Win rate	Max drawdown	Age	Status
Smart Money Following	BTC	30m	mechanical rule	+2.84%	+27.34%	-24.50%	+1.69%	54	40.7%	-8.37%	60.8	mature
Trend Breakout	BTC	30m	mechanical rule	-5.40%	+27.35%	-32.75%	-0.47%	34	26.5%	-4.52%	60.8	mature
Two-Step Pyramid - Support & Resistance	BTC	30m	mechanical rule	-8.80%	+30.18%	-38.98%	+4.69%	127	51.2%	-17.35%	60.8	mature
Order Flow Confirmation	BTC	30m	mechanical rule	-10.65%	+27.27%	-37.92%	-3.59%	169	39.6%	-12.89%	60.8	mature
Bollinger Scalper (5m)	BTC	5m	mechanical rule	-13.93%	+27.16%	-41.09%	+0.06%	249	57.8%	-5.28%	60.8	mature
Regime SAR — no stops, flip only	BTC	1h	mechanical rule	-13.97%	+27.34%	-41.31%	+0.41%	101	9.9%	-16.93%	60.9	mature
RSI Extreme Reversal	BTC	30m	mechanical rule	-15.59%	+27.24%	-42.83%	-3.47%	62	41.9%	-16.01%	60.9	mature
EMA Golden Cross	BTC	30m	mechanical rule	-19.19%	+27.20%	-46.39%	-1.78%	75	29.3%	-14.11%	60.8	mature
Bollinger Mean Reversion	BTC	30m	mechanical rule	-19.34%	+27.43%	-46.77%	-4.59%	93	41.9%	-16.24%	60.8	mature
Volatility Squeeze Breakout	BTC	30m	mechanical rule	-19.64%	+27.18%	-46.82%	+0.66%	115	32.2%	-11.15%	60.8	mature
Multi-Factor Confluence	BTC	30m	mechanical rule	-22.07%	+27.29%	-49.36%	-1.31%	123	40.7%	-24.48%	60.9	mature
Crowding Reversal	BTC	30m	mechanical rule	-23.28%	+27.45%	-50.73%	-0.73%	122	39.3%	-21.13%	60.8	mature
Signed-IC Weighting Control	BTC	30m	committee	-24.21%	+26.99%	-51.20%	+0.00%	108	41.7%	-3.44%	60.9	mature
试炼场测试实例	BTC	1h	mechanical rule	-26.41%	+27.43%	-53.84%	+2.54%	38	47.4%	-19.04%	60.9	mature
Control - Committee Only	BTC	30m	committee	-27.85%	+27.14%	-54.99%	+0.00%	120	37.5%	-9.14%	60.9	mature
Trend Pullback	BTC	30m	mechanical rule	-32.16%	+27.34%	-59.50%	+2.10%	169	37.3%	-17.56%	60.8	mature
E2 Regime Filter BTC 1d	BTC	1d	committee	+21.85%	—	—	+0.78%	0	—	-7.09%	68.8	immature
E2 Regime Filter ETH 1d	ETH	1d	committee	+18.45%	—	—	-0.36%	0	—	-5.17%	68.8	immature
Donchian-20 Breakout — EMA200 Regime, SAR flip — 4h	BTC	4h	mechanical rule	+10.39%	+26.95%	-16.56%	-4.13%	8	12.5%	-7.75%	61.0	immature
A/B — Independent Committee	BTC	30m	committee	—	—	—	—	0	—	—	None	immature
A/B — Batched Committee	BTC	30m	committee	—	—	—	—	0	—	—	None	immature

How to read this

Total return is paper equity change since the arena started. It is not a track record.
Buy & hold is what simply holding the same asset would have returned over the arena's own lifetime, and Excess is total return minus that. An arena that never opened a trade sits at 0% and can therefore "beat" a falling market while proving nothing — read Excess together with the trade count, not on its own.
Mature means at least 10 closed trades. Anything below that is noise, not evidence.
Win rate is closed trades with positive PnL ÷ all closed trades.
Max drawdown is the deepest peak-to-trough fall in paper equity.
Every decision is evaluated walk-forward with no lookahead: only data available at that bar.
Most arenas here are young control experiments, deliberately run side by side so different entry rules can be compared. Many are negative. That is the point.
Cost model: taker fees are charged on every trade; funding and slippage are not yet modelled (reality level: paper-partial). Every evidence download states this explicitly.

Why publish losses?

Because a scoreboard that only shows winners is an advertisement, not evidence. These numbers update automatically from the same live paper-trading engine that the site runs on — including when a strategy loses to simply buying and holding.

Risk disclaimer. aiarena is research and paper-trading only. Nothing here is investment advice, a solicitation, or an offer. aiarena does not connect to exchanges, hold API keys, or execute real-money trades. Past paper performance says nothing about future results, and small samples say nothing at all. Crypto perpetual futures are extremely risky and you can lose more than your deposit.

Research and paper-trading. Not investment advice. Daily state hashes are anchored in a public repository — rewriting history would break the published chain.