Methodology
Last updated: August 2026
The numbers you see on a Tipsterline profile — win rate, yield, ROI, profit, average odds — are not typed in by anyone: they are computed from the picks that were actually published. This page explains exactly what goes into each number, what stays out, and why we built it that way. If a figure in the app ever looks off to you, this page is the reference.
1. One pick, one record
The unit of measurement on Tipsterline is the pick. Every published pick creates one performance record and only one, no matter how many selections it contains. A five-leg parlay counts exactly like a single: one observation, one stake, one outcome.
That record is never multiplied by selections, followers or engagement. A tipster with fifty thousand followers and one with twenty have their statistics computed by the same formula over the same base. Popularity does not improve a track record.
- Follower count does not change any metric.
- Likes, comments and reposts do not change any metric.
- The price of a pick does not change any metric.
- A five-leg parlay is still a single pick, with one stake and one outcome.
2. The four states of a pick
Pending
The event has not settled yet. It neither adds to nor subtracts from any statistic.
Won
The prediction came in. It contributes positive profit based on the effective odds.
Lost
The prediction did not come in. The full stake is recorded as a loss.
Void
The event was cancelled or postponed, or the bet was refunded. It stays out of every calculation.
Only won and lost picks count
We call those two states settled picks. Pending and void picks are excluded entirely from win rate, yield, ROI and averages: they neither help nor hurt. That is why wins plus losses rarely add up to a tipster's total published picks.
3. Effective odds
Every profit calculation uses the effective odds. Those are the adjusted odds when the tipster recorded the price they actually got; when no adjustment was recorded, the originally published total odds are used.
The adjustment exists for two practical reasons: odds move between publishing and betting, and inside a parlay a voided leg changes the final price of the ticket. Recording the real price is more honest than leaving the published one, and the system always takes the more specific of the two.
4. Profit per pick
Won pick
profit = stake × (effective odds − 1)
This is net winnings, not the returned amount: the odds minus the 1 that represents the stake coming back.
Lost pick
profit = − stake
Pending or void pick
profit = 0
Profit is expressed in units (u), not in dollars or euros. A unit is a relative measure of performance: each person decides what a unit is worth for their own bankroll. Tipsterline neither holds nor takes part in that money.
A tipster's total profit is simply the sum of the profit of all their settled picks.
5. Win rate
Win rate
win rate = won picks ÷ settled picks × 100
The denominator is settled picks, not published ones. A tipster with 40 wins, 60 losses and 30 pending sits at a 40% win rate — 40 out of 100 settled — not 30.8%. Pending picks enter the count when they settle, not before.
Win rate on its own says very little. Hitting 70% at odds of 1.20 loses money over time; hitting 35% at 3.50 makes money. Read it next to yield and average odds, never in isolation.
6. Yield and ROI
Yield and ROI
yield = ROI = total profit ÷ sum of stakes of settled picks × 100
On Tipsterline, yield and ROI are the same number
We spell this out because elsewhere in the industry they do not always mean the same thing: some people compute ROI against a theoretical bankroll and yield against the amount staked, which produces two different percentages for the same history. Here there is only one figure — profit divided by the total actually risked on settled picks. If you compare a Tipsterline yield with another platform's, check what the other one is measuring first.
The stake of pending and void picks never enters the denominator. We only divide by what was genuinely risked and already has an outcome, so leaving picks open indefinitely does not flatter the percentage.
7. Average odds and average stake
Both averages are computed over settled picks only, and the average odds use the effective odds. A pending pick at 4.00 does not inflate the average while it remains unsettled.
They provide context about a playing style: average odds of 1.35 with a 4% yield describes a very different profile from average odds of 3.10 with exactly the same yield. The first needs to hit constantly; the second survives long losing runs.
8. Worked examples
A winning pick
Stake of 2u at effective odds of 1.80. Profit = 2 × (1.80 − 1) = +1.60u.
The same pick, lost
Stake of 2u. Profit = −2.00u. The full stake is lost, whatever the odds were.
A short history
10 settled picks, 20u risked in total and +1.00u of accumulated profit. Yield = 1.00 ÷ 20 × 100 = 5%. If 6 of those 10 were wins, the win rate is 60%.
9. Who settles a pick
The tipster is the one who marks whether their own pick won, lost or was voided. Tipsterline does not settle picks automatically against a results provider, and you should know that before reading any statistic on this platform or on any other.
What the system does enforce is that the result cannot be touched afterwards. As soon as a pick is marked won, lost or void it becomes locked: the tipster can no longer change it, nor settle it again with different odds, a different stake or different selections. There is no second chance to rewrite a result that went badly.
- Profit and effective odds are computed on the server from the recorded stake and odds; whatever number the app sends is not trusted.
- Every status change is stored with its author and timestamp.
- If a pick is settled in a way that does not match the real outcome, any user can report it and the correction goes through review, leaving a trace of the dispute.
Settling a pick incorrectly has consequences
Reports are not decorative. A tipster who marks results that do not match reality accumulates strikes, and on reaching three active ones the account is automatically suspended for seven days. Serious or repeat cases end in permanent removal from the platform. The result is declared by the tipster, but once declared it is out of their hands: it stays reviewable, disputable and consequential.
11. A record that cannot be dressed up
A pick cannot be edited or deleted once its event has started. Before kickoff there is a short window — 30 minutes from publication — to fix a typing mistake, and only while the pick is still pending and has no buyers.
Outside that window the record is fixed. Nobody can raise the odds after seeing the result, delete the pick that went wrong, or change the stake after the fact.
That is the whole difference between a screenshot and an auditable track record. Anyone can show a picture of a winning slip; nobody can show a complete history that could not be retouched.
12. Sample size matters
A high yield over few picks says almost nothing. Variance in sports betting is enormous: a run of five or six correct picks in a row is perfectly normal by chance and produces spectacular percentages that do not hold up over time.
That is why every tipster profile shows the yield next to the number of settled picks behind it. Always read both figures together: the more settled picks support a percentage, the more reliable that figure is; with very few picks, treat it as an anecdote rather than a measure of skill. Lists and rankings have to summarise for space, so before judging anyone by a percentage, open their profile and look at the full sample.
There is also no magic number beyond which a track record becomes safe. No history, however long or however good, guarantees future results. Statistics describe what already happened; they do not predict what comes next.