LongShort
PvP Bitcoin trading arena where skill earns crypto
10 followers
PvP Bitcoin trading arena where skill earns crypto
10 followers
LongShort is a PvP simulated Bitcoin day trading arena. Trade leveraged LONG/SHORT through monthly seasons, climb a skill-based leaderboard, and prove you're an apex trader. No deposit, no wallet needed to play. Up to 5% of WICK token supply is reserved for beta testers.
How does the leaderboard actually settle at the end of a season, and is it just PnL or does risk-adjusted return get factored in somehow?
@ayferrlhu Great question — and yes, it's risk-adjusted, not raw PnL (the raw PnL is just to show).
Settlement: the leaderboard is live, it re-ranks on every closed trade, from any user, in real time. At the end of each month a snapshot freezes the standings and the top 20 are published as that season's official result.
PnL is shown for transparency, it's the visible merit of a trade, but ranking is driven by Apex Score, an algorithm I built to measure who's actually good rather than who got lucky. It weighs a few things:
Skill: the main component: the overall quality and consistency of your trade record.
Progress: grows from WICK earned through real trading results.
Risk: reads your average leverage, but risk on its own is never enough to rank you.
Synergy: rewards traders who pair strong skill with genuine risk control, not reckless size.
The whole thing is deliberately tuned to reward strong, repeatable records over a single lucky 100x hit, can't be faked by lucky. That's the part I cared about most. Happy to go deeper on the formula if you're curious, you can read it in the DOCS in the landing page.
How does the leaderboard actually rank people—is it purely PnL, or does risk-adjusted return like Sharpe ratio factor in so someone grinding 1x leverage doesn't get steamrolled by someone going 100x?
@dne4ga7 it's not literally a Sharpe ratio, but it's built to solve exactly the problem you're describing — raw PnL alone never ranks you (or a lucky guy going 100x would rank first in a season, but not by skill). The whole thing runs on a score I built called Apex Score. It's public in DOCS page in the landing page, so here's the actual formula:
ApexScore = (S × 0.70) + (P × 0.15) + (R × 0.10) + (Y × 0.05)
S = Skill (70%) — the dominant term, and it's multiplicative: pnl_factor^0.40 × profit_ratio_factor^0.25 × avg_trade_ratio^0.20 × win_rate^0.15. Because it multiplies, one weak dimension drags the whole thing down,you can't paper over a bad win rate with a couple of lucky moonshots.
P = Progress (15%) — log10(1 + WICK earned), log-scaled so grinding volume has diminishing returns.
R = Risk (10%) — log10(1 + avg leverage). Note the log: the gap between 25x and 100x is tiny (~1.4 vs ~2.0), so cranking leverage barely moves the needle.
Y = Synergy (5%) — sqrt(S × R), which only pays out when skill and risk show up together.
It took a while to develop the Apex Score. It is the result of a lot of reading on what makes a trader actually good, and training the AI on those criteria until the formula felt fair. The formula is editable I can add more to synergy, reduce progress. I will adapt each season to have the best % on each factor.
So to your "1x-vs-100x" worry: leverage is a real but minor, heavily log-dampened input, it can't carry a bad trader. The 70% Skill term is pure quality and consistency (profit ratio, avg win vs avg loss, win rate), and a reckless 100x gambler tanks all of those the moment they blow up. A disciplined trader ranks well whether they run 2x or 25x, because the score rewards the record, not the size of one bet.
Real example from our docs, a solid trader (Skill 0.65, 25x avg) scores ~0.79 (PRO tier), while a low-skill farmer (Skill 0.20, 10x avg) scores ~0.42. Same game, wildly different scores, and leverage isn't what separated them.
Fair callout though: it's not a true risk-adjusted-return metric like Sharpe, I optimized for "reward consistent skill, punish luck and recklessness" over textbook risk-adjustment. If you've got thoughts on baking in something more Sharpe-like, I'm genuinely interested, this is exactly the kind of feedback the beta is for.
How does the leaderboard ranking actually get calculated across monthly seasons? Is it purely PnL or do they factor in consistency and risk-adjusted returns somehow?
@egekoumwgdr I developed a formula, I called it Apex Score. After a lot of machine learning how to put in an algorithm how to define a pro trader from a lucky guy. Read the DOCS in the landing page for more details, or the reply to others in here. Basically: Apex Score = (S × 0.70) + (P × 0.15) + (R × 0.10) + (Y × 0.05), where S = Skill (70%), P = Progress (15%), R = Risk (10%), Y = Synergy (5%).