Strategies

How much to stake is a separate question from which tip to take. A good tip staked badly still loses money, and no staking method turns a bad tip into a good one.

Reading the Chart

The green line is your bankroll, bet by bet, on our own settled bets at the odds they really had. The dotted line is where you started; the red areas are drawdowns, every stretch spent below your previous best. The faint lines are the other strategies, for scale — tap one to bring it to the front.

Under the figures, one square per bet: green won, red lost, grey was cancelled. The outlined block is the longest run of losses in the record.

Every Bet, Listed

Under the chart is every ticket the strategy met, in kickoff order: the stake, how it ended, and the bankroll afterwards. Add the rows up and you get the headline figure. Rows with no stake say why the strategy passed.

Challenge has to wait. Doubling after a win means knowing the win first, and our tickets overlap — most of them start before the one before them has finished. The list marks those rows, and Challenge therefore plays far fewer bets than the rest. Nothing here can stake money that other bets are still holding. A ticket counts as settled three hours after its last kickoff, which is an estimate.

What Works

• Flat staking — the same share of your bankroll every time, 1% to 5%. It cannot blow up, and it is the honest default.
• Challenge — a progression that doubles after a win instead of after a loss. It has its own help entry; open it from the Bets tab.

Why the fractions of Kelly are gone

Kelly sizes a bet by the edge, and the edge needs a probability you can trust. Ours was measurably wrong exactly where it mattered — our own numbers ran high precisely where the model disagreed most with the bookmaker — so the confidence figure was removed from the app, and with it the only honest input Kelly had. Rather than keep a formula fed by a number we no longer stand behind, the strategies here are flat staking and Challenge.

Why Challenge sits beside them

Challenge is a progression too, but it runs the other way: it never chases a loss. What it asks for is five wins in a row, and that is rarer than it feels. What a failed run costs depends on the odds — at 2.00 it is the stake the run began with, below 2.00 it is more — and four failed runs end a challenge without ending the plan, so a long enough losing streak still empties a bankroll. The Challenge help entry has the whole rule.

The calculator does not argue with you about this. Set your starting stake and it tells you what the last bet of the ladder costs and how often five wins in a row actually happen — worked out from our own settled tickets rather than a textbook. Check the record for how our tips have really done before you pick a stake.

What This Screen Never Claims

No staking method turns a negative expectation into a positive one. Only a more accurate model does that, and that work lives elsewhere in the app. Everything here is about not losing your bankroll to variance while the model does its job.