Start With What the Record Includes
A useful record identifies the sport, market, product tier, date range, odds, stake size, wins, losses, pushes, units, and ROI. If those details are missing, the headline number may not tell the full story.
Different sports and markets should be separated when the underlying products or decision processes are different.
Look for Consistent Rules
The grading process should be stable enough that the same rules can reproduce the published record. That includes how voids, pushes, line changes, odds, and timing are handled.
Consistency matters because a picks service can look stronger or weaker depending on which results are included and how they are grouped.
Connect Results to Product Expectations
A customer should be able to see how the record connects to what the product actually delivers. Free, paid, sport-specific, and market-specific offers can have different access levels and reporting boundaries.
Tracked results are one input in a decision. Responsible bankroll management, legal eligibility, and personal risk tolerance still matter.
Combine Returns by Stakes, Not by Averaging Percentages
Suppose a hypothetical MLB record earns 5 units on 100 units risked (5% ROI), while an NFL record loses 2 units on 10 units risked (-20% ROI). Averaging those percentages gives -7.5%, which is not the combined ROI. Combined profit is 3 units on 110 risked, or approximately +2.73%.
Even a correctly weighted total can hide differences between sports, tiers, and market types. Read the separated records first and keep research or backtest results out of a customer-facing aggregate. Use comparable date ranges and clearly stated stake and grading conventions.
Key Takeaway
ESA uses tracked results to support accountability, not certainty. The goal is to make the product boundary and performance context clear enough for informed decisions.