Staking Plan

Staking Plan

A bettor fires on NFL games by gut feel. First bet: $50, wins $45. Second bet: $50, loses $50. Third bet — he's "sure" about this one — $200. Loses $200. A week later he's sitting on half a bankroll despite carrying a genuine +2% ROI edge. Not bad luck. No staking plan. Bet sizes swung between 5% and 20% of roll with zero structure, and one ordinary downswing did fatal damage. A staking plan is the rule that keeps emotion out of bet sizing. Without it, even a sharp edge gets washed out by variance inside a couple of months.

What It Actually Is

A staking plan is a systematic method for sizing every bet. It's the foundational bankroll management layer, working alongside accurate edge estimation and variance awareness.

Why it matters:

• Ruin protection. Bet size must be small enough that a bad run can't wipe the bankroll. Without a plan, bettors instinctively size up when confident or after losses, both increase ruin risk.
• Discipline over emotion. A pre-written rule blocks the psychological traps: pressing after a win, chasing after a loss.
• Long-run convergence. Variance demands a large sample before results reflect true edge. A staking plan keeps the bankroll alive long enough to get there.
• Honest performance tracking. With consistent sizing, ROI by bet type is measurable. With random sizing, true edge is impossible to calculate.

Connections to related concepts: a staking plan is the practical application of the Kelly Criterion and Risk of Ruin. Kelly gives the mathematically optimal bet size but requires accurate edge estimates. In practice, most bettors use simplified plans, flat stakes or percentage staking, because precise edge estimation is hard.

Without a staking plan, any betting or poker strategy is doomed. The edge might be real, but wrong bet sizing will destroy the bankroll before that edge ever shows up in the results.

Categories of Staking Plans

All staking plans fall into three categories based on how bet size is determined:

Flat staking. Bet size is fixed regardless of current bankroll or recent results. Example: always $50 per event. The simplest plan, and the most resistant to emotional decisions.

Percentage staking. Bet size is a percentage of the current bankroll. Example: 2% per bet. Bankroll grows, bets grow proportionally. Bankroll shrinks, bets shrink.

Progressive systems. Bet size adjusts based on recent results. Martingale doubles after a loss. Fibonacci steps up by sequence. These systems are mathematically dangerous.

The right category depends on experience, emotional discipline, and how confidently you can estimate your edge. Most disciplined bettors settle on percentage staking at 1–3% of bankroll per bet.

Flat Staking

Flat staking is the simplest plan: one fixed size per bet, no exceptions. Common setup: 1% of starting bankroll as one unit, everything wagered at one unit.

Advantages:

• Simplicity. No math before each bet.
• Tilt protection. Size doesn't change after a bad run, blocking the impulse to chase losses.
• Clean tracking. All bets equal, ROI calculates directly.
• Beginner-friendly. No precise edge estimate required.

Disadvantages:

• Slow growth. Bankroll doubles but bet size stays the same. Effective edge as a percentage shrinks.
• No downswing adjustment. Roll drops by half but stakes are unchanged. Ruin risk climbs.
• Ignores edge strength. Every bet is equal regardless of how strong or thin the edge is.

When flat staking fits: first 6–12 months of betting, bankroll under $5,000, uncertain edge, or emotional instability. Most bettors shift to percentage staking or a 1-3-5 unit system once they've built some experience.

Percentage Staking

Percentage staking sizes every bet as a fraction of current bankroll. Standard ranges:

• Very conservative: 0.5–1% per bet
• Standard: 1–2%
• Aggressive: 3–5%
• Very aggressive: 5–10% (ruin risk climbs fast)

Concretely: $5,000 bankroll, 2% staking, every bet is $100. Bankroll grows to $6,000, next bet is automatically $120. Drops to $4,000, bet is $80.

Advantages:

• Automatic scaling. Bankroll grows, bets grow. Bankroll drops, bets drop.
• Ruin protection. Reducing size during drawdowns prevents rapid wipeout.
• Flexible. Percentage adjusts to personal risk tolerance.

Disadvantages:

• Needs frequent recalculation. Bankroll shifts daily, so does bet size.
• Still ignores edge confidence. All bets sized equally, same as flat staking.
• Correlation-sensitive. Bets on the same market cluster raise actual portfolio variance well above the theoretical sum.

A workable rule of thumb: a sharp with a +3% ROI edge should stake 2–3% (near 1/4 Kelly). A recreational bettor with an uncertain edge should stay at 1–1.5%.

Kelly Criterion as a Staking Plan

The Kelly Criterion is the mathematically optimal staking plan under a logarithmic utility function. The formula:

Kelly% = (edge × odds - 1) / (odds - 1)

Concretely: 5% edge at odds of 2.10. Kelly = (0.05 × 2.10 - 1) / (2.10 - 1) = 0.0455 / 1.10 = 4.1%. Bet 4.1% of bankroll.

In practice, Full Kelly is rarely used because ruin risk remains meaningful, around 13.5% even for a perfect player. Pros use Fractional Kelly:

• Full Kelly: maximum growth rate, 13.5% ruin risk
• 1/2 Kelly: 75% of Full Kelly growth, 2% ruin risk
• 1/4 Kelly: 50% of growth, ruin risk under 0.1%

Most sharps run 1/4 or 1/2 Kelly. You capture most of the growth while cutting variance significantly.

Use the Kelly calculator to size bets by your edge and current odds.

The key limitation: Kelly demands an accurate edge estimate. Overestimate your edge by 2%, thinking +5% when it's really +3%, and Kelly pushes you to bet 1.5× the optimal size, materially increasing ruin risk. Fractional Kelly is the compromise: slightly suboptimal growth, protection against edge-estimation errors.

Progressive Systems: Martingale and Others

Progressive systems adjust bet size based on recent results. They're intuitively appealing to newcomers but mathematically treacherous.

Martingale. Double after every loss. When you finally win, you recover all prior losses plus one unit. The problem: a run of 7–10 consecutive losses will exhaust any realistic bankroll. At odds of 2.0, the probability of a 7-loss streak is (0.5)^7 = 0.78%, which will show up regularly across 100 bets. Long-run ruin risk approaches 100%.

Fibonacci. Bet size follows the Fibonacci sequence: 1, 1, 2, 3, 5, 8, 13... Step forward after a loss, step back two after a win. Less extreme than Martingale, same core problem: a prolonged downswing drains the roll.

Labouchere. Write a sequence of numbers (e.g., 1-2-3-4-5). Bet the sum of the first and last numbers. Win, cross both off. Lose, add the amount wagered to the end of the list. Goal: clear the whole list. Same flaw: a downswing expands the list endlessly, demanding ever-larger bets.

Oscar's Grind. Hold bet size flat after a loss. Increase by one unit after a win, until a target profit is reached. Less extreme, but still psychologically painful on a downswing.

Every one of these systems shares the same fundamental flaw: they ignore actual result distributions. Extended losing streaks are statistically inevitable, and any progressive system will eventually hit a run the bankroll can't survive. Professional bettors don't use them.

Try the Labouchere calculator to see exactly how quickly a typical downswing burns through a roll.

Confidence-Based Staking: The 1-3-5 Unit System

The unit rating system sits between flat staking and Kelly. The idea: classify bets by edge strength and assign proportional sizes.

Typical structure:

• 1 unit: thin edge (e.g., +1–2% ROI), standard wager
• 3 units: solid edge (+3–4% ROI), elevated confidence
• 5 units: strong edge (+5%+ ROI), maximum conviction

Where one unit = 1% of bankroll. At a $5,000 roll, 1 unit is $50 and 5 units is $250.

Advantages:

• Accounts for edge strength. Not all bets are equal, the plan reflects that.
• Moderate complexity. Sits between pure flat staking and full Kelly math.
• Forces honest classification. Pre-defined categories make you assess each bet objectively.

Disadvantages:

• Depends on honest grading. If everything feels like "5 units," the plan fails.
• Doesn't account for bet-level variance. A parlay and a straight bet can end up rated identically.
• Requires calibration. Over time, your "5-unit" bets should actually correlate with higher realized ROI.

Experienced regulars often prefer a 1-3-5 or 1-5 system over Full Kelly, because Kelly is too sensitive to edge-estimation errors.

Professional Approaches

Seasoned bettors typically combine multiple staking approaches depending on the bet type:

Syndicate bankrolls. In a team of sharps, the group often runs a shared bankroll with a unified staking strategy. Bet sizes are coordinated to avoid members overlapping and inadvertently moving lines against each other.

Multi-account distribution. Spreading action across multiple books or accounts. Standard sizing on each account: 1–2% of total combined bankroll. This reduces exposure to account restrictions and distributes drawdown risk.

Segregated bankrolls by format. Experienced bettors keep separate rolls for each activity: sports betting, cash poker, tournaments, casino advantage play. This produces honest ROI tracking per format and prevents results from one leaking into another.

Closing line awareness. Bet size scales with price quality. If one book offers 2.10 and another offers 2.05 on the same bet, the sharp sizes up 1.5× at 2.10 and passes on 2.05. Integrated staking plus line shopping.

Position management. Not all bets fire simultaneously. Sharps sometimes bet one side early, then the other side later to profit from line movement. This requires a more advanced staking framework.

More on line analysis: Line Shopping.

Choosing the Right Plan

Picking a staking plan comes down to three things: bankroll size, edge confidence, and emotional discipline.

Small bankroll (under $500). Flat staking at 1–2% of bankroll. Simplicity outweighs optimization at this level.

Mid-range bankroll ($500–$5,000). Percentage staking at 1–3%, or a 1-3-5 unit system. Enough experience to estimate edge, but emotional discipline is still developing.

Large bankroll ($5,000+). Fractional Kelly (1/4 or 1/2) for experienced bettors with solid edge data.

Low edge confidence. Flat staking or very conservative percentage staking (0.5–1%). Without edge certainty, an aggressive plan accelerates ruin.

High emotional discipline. 1/2 Kelly or an aggressive 1-3-5 unit setup are viable.

Low emotional discipline. Flat staking or 1/4 Kelly. Smaller swings mean less psychological pain.

Most sharps follow a natural progression: flat staking first, then percentage staking after 6–12 months, a 1-3-5 system after a year, and Fractional Kelly after two or three years. That path builds variance tolerance gradually instead of forcing it early.

Common Mistakes

1. Bet size out of proportion to bankroll. A bettor puts 10% of roll on a game because he's "sure." On a $1,000 bankroll that's $100 per bet. Ten straight losses, entirely possible on +EV bets, ends the run. The math is unforgiving.

2. Running progressive systems. Martingale, Fibonacci, Labouchere. The logic sounds reasonable: recover losses on the next bet. Reality: a downswing wipes the roll long before recovery arrives. Professionals avoid these entirely.

3. Pressing after an upswing. Five wins in a row and the bettor thinks he's running hot, so he sizes up. That's the gambler's fallacy. Variance has no memory. The plan sets the size, not recent results.

4. Ignoring correlations. Betting 2% each on five different markets within one game (spread, total, team fouls, etc.). Actual portfolio variance is far higher than the sum of individual bet variances.

5. Mixing bet types without separate bankrolls. Applying the same 2% to single-game bets and parlays. Parlays carry dramatically higher variance. 2% on a parlay is often wildly oversized.

6. Changing the plan mid-stream. Starts with 1% flat, bumps to 2% after a good stretch, then 5% after another. A year later the bettor is running 5% flat, far riskier than intended. Stick with the plan for at least 3–6 months before adjusting.

7. No records. Betting without tracking every bet size. When a downswing hits, there's nothing to analyze. Keep a log: spreadsheet or a dedicated bet tracker.

Where It Falls Short

A staking plan manages risk. It does not create edge. No plan converts a negative-EV operation into a winning one. Perfect Kelly applied to zero edge produces breakeven results with high variance, not income.

The plan also can't account for skill drift. If your edge erodes over time, tougher competition, book limits, structural changes, and you keep sizing the same, ruin risk rises quietly. Revisit actual edge estimates every 1,000+ bets.

Life events aren't in the formula either. A medical bill, job loss, or other emergency might force a bankroll withdrawal. Build a buffer above the theoretical minimum so those situations don't force bad decisions at the table.

Finally, a staking plan is a statistical tool. Over 50–100 bets, variance can produce almost any outcome even with a solid plan. Don't scrap the plan because of a rough stretch. If the edge is real and the math is right, discipline over a long sample does the work.