Maple is built around three commitments that shape every part of the product, from how a backtest is presented to how AI is used.
Evidence over excitement
A strategy that "feels" right is not the same as a strategy that has been shown to work. Maple is designed to slow that instinct down — to insist on a full backtest, a look at the trade-by-trade detail, and a check for overfitting before any conclusion is treated as reliable.
Robustness over perfect backtests
A backtest with a spectacular headline return and no resilience to small changes in its parameters, its time window, or the sequence of its trades is not a good result — it's often a warning sign. Maple treats a strategy that performs reasonably across a range of conditions as more valuable than one that performs brilliantly in exactly one narrow, precisely-tuned configuration.
Better questions over easy answers
Maple will not tell a user a strategy is "good" or "bad." It will show what the data says, where the strategy's weaknesses are, and what assumptions its apparent edge depends on — and leave the judgment to the person doing the research. The goal is a trader who asks sharper questions of their own ideas, not one who has outsourced that judgment to a black box.
What this rules out
This philosophy is incompatible with promising returns, selling signals, or presenting AI-generated commentary as a substitute for a trader's own judgment. Maple does not do any of those things, by design.