Every figure published on this site falls into one of two buckets: numbers a stranger can check against a live public record, and numbers that describe a simulation, or a past, that cannot be re-run. Most sites only tell you about the first bucket and let the second borrow its credibility. This page names the second bucket outright.

Every figure attached to DAITIQ Markets is simulated. The ~$185K of capital was paper money, the fills were sandbox fills, and no real dollar was ever at risk. Real-money trading introduces slippage, partial fills, liquidity constraints, and market impact that paper trading cannot replicate — so if those four systems had traded real money, the results would have been different, and almost certainly worse, because simulation errors flatter the trader.

Those numbers are published as a record of engineering and of judgment, not as investment results. The honest headline is that even as a simulation the project lost: roughly $23,800 behind simply holding an index fund over the same window. That gap is stated on the Markets case study precisely because it is unflattering.

This site once quoted win rates and an ROI figure from the original fall-2025 sports modeling work, before ParlayGenIQ had a public record. In July 2026 they were withdrawn from every page. The problem was methodology: they were graded after the fact against closing lines, with results backfilled once games had ended — a backtest wearing a track record's clothes. Only a handful of those picks existed before kickoff. Nobody could have placed those bets at those numbers, which is the test a published win rate has to pass.

They are not restated here, even as a retraction, because repeating a withdrawn number is republishing it. They do not come back unless the underlying picks are re-audited against bet-time lines — and the removal itself is in the site's public git history.

The ParlayGenIQ record is the opposite case, and the standard the rest of the site is held to: 4,700+ graded public picks since Feb 2026, reading 54.3% win rate and +1.2% ROI as of 2026-08-19.

• →Every pick is timestamped before the game starts, graded against the line it was published with, and never edited or deleted after the fact.
• →The live record at parlaygeniq.com/track-record is the authority — the figures above are synced from its public API daily and may lag it slightly.
• →Every change to the figures quoted on this site is a commit, so the history of what was claimed, and when, is inspectable rather than taken on faith.

If a stranger cannot check a number, it either carries a label that says so — like the paper-trading figures above — or it does not get published. When a figure already published turns out to fail that test, it is withdrawn publicly rather than quietly. That has happened once. This page exists so that if it ever happens again, it will be obvious.