3 Staking Plans

Staking appropriately is an essential aspect of any Advantage Player. Stake too high and you risk the chance of ruin. Stake too low and you fail to optimise the return from your edge.

There are many staking different types of staking strategies. At bookiebashing we categories them as follows:

• Unit Loss
• Unit Win
• Kelly Staking

Each of the above staking plans can be fixed (relative to the size of the bankroll) or proportional. 

There are other staking plans (such as progressive staking and Unit-impact staking) that we deem them to be inappropriate to an advantage player at bookiebashing.

The three variables that affect when a punter may want to adjust a stake are bankroll, EV and odds size.

Unit-Loss Staking

Under Unit loss staking is a staking plan where we stake exactly the same unit for every bet, regardless of EV or odds. So for example, we may stake £1 at Evens and £1 at 10/1 and £1 at 100/1.

This strategy initially sounds sub-optimal, but it has its benefits. Primarily it is a very easy plan to follow, especially when placing multiples and System Bets such as Lucky 15s.

Unit-Win Staking

Under Unit-Win Staking we adjust the stake so that we win the same amount of liability on every bet. For example if we want to win £200 in every bet then we will stake £200 at evens, £100 at 2/1 and £20 at 10/1.

To Calculate Unit-Win Staking:

Kelly-Staking

Kelly Staking was developed by John Kelly in 1956 to optimise the returns given that longer odds hold more expected value. is derived from a paper from John Kelly called “A New Interpretation of Information Rate” which he wrote for AT&T to assist with signal noise issues over long distance telephone lines. John Kelly hypothesised that a logarithmic utility function would provide a staking methodology that would have a better return than Unit-Loss or Unit-Win staking if we can consistently bet on high value bets over a large enough sample size.

A comparison using real data from bookiebashing trackers

The Real-life betting histories to Jan-22) have been analysed from the BB Horse Racing Tracker, Golf Tracker, Bet Tracker and Darts bets from the x180s tool. The stakes have been normalised in each of the graphs below so that 1000 unit stakes were placed in total under each betting history.

Horse Racing Tracker

Bet Tracker

Golf Tracker

 Darts bets (from Bet Tracker)

 Commentary on real life Bet Histories

Kelly staking outperforms Unit-Win staking under every set of data. It is worth noting that the sample sizes of the Golf and Darts bet histories are several orders of magnitude lower than the Horse Racing and Bet Tracker bet histories.

Proportional Staking

The above graphs consider Fixed Staking plans – the staking is fixed relative to the size of the starting bankroll and does not change. It may be preferable to adjust staking as a bankroll changes this has the benefit of maximising returns as a bankroll grows, and also protects against the risk of ruin if a losing run is encountered.

A comparison of staking plans (differences)

The three variables that affect when a punter may want to adjust a stake are bankroll, EV and odds size.

Replaceable (or “loseable”) bankrolls

At bookiebashing we recommend the concept of a “replaceable” bankroll. When deciding how much money to assign to a particular strategy

DO NOT assign more money to that strategy than you can afford to comfortably lose in entirety. Sports betting can be highly volatile and there will be frequent, long losing runs.

DO assign an amount of money to a bankroll that is “replaceable”. You may want to replace the bankroll under different conditions, or you may decide not to continue with a a particular strategy.

Bookiebashing Golf Staking

The bookiebashing golf staking methodology is a derivative of Kelly Staking. We decided to slightly change the logarithmic nature of Kelly to optimise our staking when betting 10-15% of the field Each Way to win a single event. More on the bookiebashing golf staking methodology can be found here.

Maximum Drawdown

The maximum drawdown is the maximum observed loss from a peak to a trough of a portfolio. Here is the historical graph for the Golf Options we put up each week on bookiebashing, with advised stakes:

The largest drawdown was between bets 284 and 653 from £3,386 to £2,361 (a period which last a few days over 1 year). Whilst we advise a bankroll of £1000 under this staking plan – were you to have started at exactly bet 284, this bankroll would have been busted at the lowest trough one year later before we started catching up with Expected Value again.

Compounding

In the analyses above we have assumed that Single bets have been placed.

One of the most effective tools in an advantage players armoury is the ability to compound EV. If we have two single independent bets that are 105% EV then we can create 110.25% EV (1.05 * 1.05) bet by combining the two bets into a double. The EV of the double is more than the sum of its individual parts.

The effects of compounding can be dramatic. This is mostly seen in Horse Racing and Coupons tracker results histories. Whilst the average ROI of a single horse on the horse racing tracker is 105% (as of Jan 2022), many bet histories will have ROIs of 20% or more due to the placement of Lucky 15s, Lucky 31s etc.

We recommend using doubles, trebles, Lucky 15s and Lucky 31s across all trackers wherever possible. The bookiebashing systems bet calculator can be used to analyse the effect on EV when using system bets. It can also be used to calculate returns and p/l:

The Bookiebashing Variance Model

The bookiebashing variance model is a Monte Carlo Simulation that will analyse the expected returns of a strategy over a user defined number of simulations given ROI, odds and bankroll.