How to Track Price Sensitivity in Betting Results

# How to Track Price Sensitivity in Betting Results

The direct answer

Price sensitivity asks whether a recorded betting result still works when the accepted odds are slightly worse. Recalculate the complete sequence at realistic price bands, such as the advised price, 2 per cent lower, 5 per cent lower and the actual price taken. If a small reduction turns apparent profit into loss, the method is highly price-sensitive and execution is part of the edge.

This is not a reason to invent a generous price. It is a reason to preserve timestamps, accepted odds and every qualifying bet so the test can be reproduced.

Why the advised price is not enough

A strategy might be published when a selection is 3.00, but many readers may only obtain 2.80. Those prices imply different break-even probabilities:

The formula is `1 / decimal odds`. A move from 3.00 to 2.80 looks small on a screen, yet it raises the win rate needed to break even by 2.38 percentage points.

If the original edge was narrow, that difference can consume it.

The minimum data to record

For every qualifying selection, keep:

1. date and event; 2. market and selection; 3. advised price and timestamp; 4. actual accepted price and timestamp; 5. bookmaker or exchange, with commission where relevant; 6. stake and staking rule; 7. result, including voids, dead heats and deductions; 8. a closing or reference price where it has a clear definition.

Screenshots can support a record, but a structured ledger is needed for calculation. The BetOwl calculator directory provides supporting tools for checking turnover, returns and return on investment.

Recalculate the whole sequence

Do not test price sensitivity only on the winners. Apply each price rule to every recorded bet in its original order.

For a one-point win bet at decimal odds `d`, profit is `d - 1`. A loser returns minus one point. A void returns zero. For a 5 per cent worse-price test, multiply the decimal profit portion by 0.95 and then add the stake back:

`tested decimal odds = 1 + ((original odds - 1) x 0.95)`

That distinction matters. Reducing the whole decimal price by 5 per cent can produce a different result because the returned stake is not profit.

A worked illustration

Take five one-point bets with results win, loss, win, loss, loss and accepted prices of 2.50, 3.20, 1.90, 4.00 and 2.10.

Only the winning prices affect the realised return in this small sequence, but every bet remains in the denominator for turnover and return on investment. Over a full record, price changes also affect which bets qualify if the method has a minimum-price rule. That second effect should be tested separately.

Two different sensitivity tests

Execution sensitivity

Keep the selections fixed and reduce every accepted price by a declared amount. This asks how much slippage the recorded selections could withstand.

Qualification sensitivity

Re-run the original selection rule at nearby minimum prices or value thresholds. This asks whether the strategy depends on one fragile cut-off.

Do not combine the two and call the result one test. A method can survive worse execution on the same bets but change substantially when the qualifying set is rebuilt.

Use bands that could really happen

Choose bands from observed price availability, not from the result you want. Useful comparisons might include:

Closing price is supporting evidence, not proof of profit. Define the market, timestamp, source and commission. Comparing an early bookmaker price with a late exchange price without adjustment is not a clean like-for-like test.

Watch the biggest winners

A long record can appear robust because one or two large-priced winners dominate the total. Publish the result with the largest winner removed, then with the largest three removed. This does not erase legitimate wins. It shows how concentrated the conclusion is.

Also group returns by price band. If every positive point came from selections above 20.00, the method has a different risk profile from one producing steady returns around 2.00.

The sample-size guide explains why a large bet count can still provide weak evidence when outcomes are dependent or returns are dominated by rare events.

Price sensitivity and staking

Test prices before adding an elaborate staking method. Variable stakes can hide a fragile selection record because larger bets may happen to land on the best results.

Start with one-point level stakes, then compare actual staking separately. The guide to fixed and percentage staking explains how cash exposure changes even when the selections remain identical.

A publishable audit table

A transparent strategy review should show:

Keep the code or spreadsheet formula with the report. A sensitivity claim that cannot be reproduced is only a description.

Responsible limits still come first

Better price tracking does not make gambling safe or guarantee a return. Decide an affordable budget independently of the analysis and use account limits where helpful. Never raise stakes to compensate for taking a shorter price.

BetOwl's responsible gambling centre lists practical controls and support routes.

Frequently asked questions

How much worse should the test price be?

Use bands supported by the actual difference between advised and obtainable prices. Two and five per cent are useful illustrations, not universal standards.

Should I reduce decimal odds by 5 per cent?

Reduce the profit portion, not the returned stake, unless the test explicitly says otherwise. Use `1 + ((odds - 1) x 0.95)`.

Does beating the closing price prove a strategy works?

No. It can be useful supporting evidence if the reference is consistently defined, but it does not guarantee realised profit or future performance.

Should void bets remain in the record?

Yes. Record them and define how they affect turnover. A high void rate can reveal a timing or market-definition problem.