Kelly Criterion Calculator: Optimal Bet Sizing

Finding value bets is only half the battle. Once you have identified an edge, you need to decide how much to stake. Bet too little and you leave money on the table. Bet too much and a losing streak can wipe out your bankroll. The Kelly Criterion calculator solves this problem by telling you the mathematically optimal stake size based on your edge and bankroll.

This page gives you a free Kelly Criterion calculator designed specifically for sports betting. You will also learn why most experienced bettors use fractional Kelly (Half or Quarter Kelly) rather than the full formula, how to estimate your win probability, and how to avoid the common mistakes that cause even sharp bettors to over-stake. Explore our full suite of betting calculators to find other tools that complement your Kelly staking strategy.

Use the Kelly calculator below to get started. Enter your bankroll, the odds you are getting, and your estimated win probability. The calculator defaults to Quarter Kelly because it offers a safer path to long-term growth while protecting you from the variance that comes with imperfect probability estimates.

Quick Start: Calculate Your Kelly Stake

The Kelly Criterion calculator helps you determine the optimal percentage of your bankroll to wager on a single bet. Unlike flat betting where you stake the same amount regardless of edge size, Kelly adjusts your stake based on how much of an advantage you believe you have.

Before using the calculator, you need three pieces of information:

• Bankroll: The total amount of money you have set aside for sports betting. This should be money you can afford to lose without affecting your daily life.
• Odds: The price you are getting on your bet. The calculator accepts American odds (like -110 or +150), decimal odds, or fractional odds.
• Win Probability: Your honest estimate of how likely your bet is to win. This is the hardest input to get right, and getting it wrong is the main reason Kelly can lead you astray.

The calculator returns three outputs:

• Full Kelly Stake: The mathematically optimal bet size assuming your probability estimate is exactly correct. Most bettors should not use this number.
• Half Kelly Stake: A more conservative approach that still captures most of the growth while reducing volatility significantly.
• Quarter Kelly Stake: The safest option that experienced bettors often prefer, especially when probability estimates are uncertain.

Why Quarter Kelly is the default: The full Kelly formula assumes your probability estimate is perfect. In reality, even the best handicappers have margin of error in their predictions. Quarter Kelly protects you from overconfidence and reduces the emotional pain of drawdowns. If your edge is real, you will still grow your bankroll over time. If your edge is smaller than you thought, Quarter Kelly prevents catastrophic losses.

Remember that sports betting involves risk and no calculator can guarantee profits. The Kelly Criterion is an educational tool that helps you think about bet sizing in a disciplined way. Always bet only what you can afford to lose.

How the Kelly Criterion Works (and the Snippet-Ready Formula)

The Kelly Criterion is a formula developed by John Kelly at Bell Labs in 1956. Originally designed for optimizing signal noise in telephone lines, gamblers and investors quickly realized it could be applied to any situation where you are making repeated bets with an edge.

In plain English, the Kelly Criterion tells you what percentage of your bankroll to bet so that you maximize the long-term growth rate of your money. It balances two competing goals: betting enough to take advantage of your edge, while not betting so much that a string of losses devastates your bankroll.

The formula in its simplest form is:

Kelly Percentage = (bp - q) / b

Where:

• b = the decimal odds minus 1 (or the net profit per dollar wagered)
• p = your probability of winning
• q = your probability of losing (which equals 1 - p)

Variable	Definition	Example
b	Net profit per unit wagered (decimal odds - 1)	At -110 odds: b = 0.909
p	Your estimated probability of winning	You estimate 55% chance: p = 0.55
q	Your probability of losing (1 - p)	q = 1 - 0.55 = 0.45
Kelly %	Optimal percentage of bankroll to wager	(0.909 x 0.55 - 0.45) / 0.909 = 5.5%

Understanding the output: The Kelly Criterion returns a percentage of your bankroll. If the formula gives you 5%, and your bankroll is 1000 dollars, your optimal stake is 50 dollars. If the formula returns a negative number, it means you have no edge and should not place the bet.

Negative Kelly values: When the calculator shows a negative Kelly percentage, the odds are not in your favor. This happens when your estimated win probability is lower than the break-even probability implied by the odds. In these situations, the correct stake is zero. A disciplined bettor passes on bets with negative expected value.

The Kelly Criterion is mathematically elegant but rests on a critical assumption: that your probability estimate is correct. In the real world of sports betting, your estimates contain error. The next sections explain why this matters and how to adjust your approach.

Worked Example: -110 Spread With a Small Edge

Let us walk through a concrete example using a standard NFL point spread. You are betting the Kansas City Chiefs -3.5 at -110 odds. After analyzing the matchup, you believe the Chiefs have a 55% chance of covering the spread.

Step 1: Identify the inputs

• Bankroll: 2000 dollars
• Odds: -110 (which equals 1.909 in decimal, so b = 0.909)
• Your estimated win probability: 55%
• Implied probability at -110: approximately 52.4%
• Your edge: 55% - 52.4% = 2.6%

Step 2: Calculate Full Kelly

Kelly % = (0.909 x 0.55 - 0.45) / 0.909 = 5.5%

Full Kelly stake = 2000 x 0.055 = 110 dollars

Step 3: Apply fractional Kelly

Kelly Fraction	Calculation	Stake
Full Kelly (100%)	2000 x 5.5%	110 dollars
Half Kelly (50%)	2000 x 2.75%	55 dollars
Quarter Kelly (25%)	2000 x 1.375%	27.50 dollars

Why the difference matters: If your 55% estimate is actually closer to 53% (a common margin of error), Full Kelly would have you significantly over-betting. Quarter Kelly at 27.50 dollars gives you room for error while still capturing the positive expected value of the bet.

This example also shows why you need to understand expected value before using Kelly. If you cannot estimate your true win probability with some confidence, the Kelly output is meaningless. Learn more about identifying edges in our Expected Value betting guide.

Fractional Kelly: Half, Quarter, and Why Pros Rarely Use Full Kelly

Full Kelly is theoretically optimal for maximizing long-term bankroll growth, but it comes with a major practical problem: volatility. Even with a genuine edge, Full Kelly produces stomach-churning swings that most bettors cannot handle.

Consider this: Full Kelly is designed to maximize the logarithm of your wealth over time. This means it accepts a roughly 50% chance of losing half your bankroll at some point, and a meaningful chance of losing 90% or more during extended drawdowns. Most people would abandon their betting strategy long before realizing the long-term gains.

Why fractional Kelly is better for sports betting:

• Probability estimates are never perfect: The Kelly formula assumes you know your exact edge. In sports betting, even the sharpest handicappers operate with significant uncertainty.
• Variance is brutal: Sports have small sample sizes compared to casino games. A 55% bettor can easily lose 10 bets in a row, which at Full Kelly would devastate your bankroll.
• Psychological capital matters: The best strategy is one you can actually follow. Smaller stakes reduce emotional stress and prevent panic decisions.
• Correlated bets: Sports bettors often have multiple positions that share risk factors. Full Kelly does not account for this correlation.

Kelly Fraction	Growth Rate vs Full	Volatility vs Full	Best For
Full Kelly (100%)	100%	100%	Theoretical optimum (rarely used)
Half Kelly (50%)	75%	50%	High confidence in probability estimates
Quarter Kelly (25%)	50%	25%	General sports betting (recommended default)
Eighth Kelly (12.5%)	25%	12.5%	High-variance props and uncertain edges

The key insight is that Half Kelly gives you 75% of the growth rate while cutting volatility in half. Quarter Kelly still captures meaningful growth while being much more forgiving of estimation errors.

The simple rule: The less confident you are in your probability estimate, the smaller your Kelly fraction should be. If you have a proven model with years of data, Half Kelly might be appropriate. If you are making subjective judgments about matchups, Quarter Kelly or less is wiser.

Professional bettors and hedge funds that use Kelly-type systems almost universally use fractional approaches. The mathematical elegance of Full Kelly matters less than the practical reality of surviving variance and staying in the game.

When to Use Half vs Quarter Kelly

Choosing between Half and Quarter Kelly depends on several factors related to your confidence level, the market you are betting, and your personal risk tolerance.

Use Half Kelly when:

• You have a quantitative model with a long track record of accuracy
• You are betting liquid markets (major spreads and totals) where the lines are efficient
• You have verified your edge across a large sample size (hundreds of bets minimum)
• Your bankroll can absorb a 30-40% drawdown without affecting your emotional stability
• You are betting uncorrelated events (not multiple bets on the same game)

Use Quarter Kelly when:

• You are making subjective assessments rather than using a formal model
• You are betting props, player markets, or other high-variance bets
• You have a smaller sample size or are newer to advantage betting
• You want to minimize the impact of estimation errors
• You are placing multiple bets that may share common risk factors

Use Eighth Kelly or less when:

• You are betting highly volatile prop markets
• Your probability estimates are educated guesses rather than data-driven
• You are experimenting with a new approach or market
• You want maximum protection against ruin even at the cost of slower growth

The transition between fractions should be gradual. Start with Quarter Kelly and move to Half Kelly only after demonstrating consistent accuracy over a meaningful sample. Most serious bettors find that Quarter Kelly provides the best balance between growth and emotional sustainability.

Estimate Win Probability (The Edge Bridge)

The Kelly Criterion is only as good as your probability input. Enter an accurate probability and you get a sensible stake size. Enter a probability that is off by a few percentage points and Kelly can have you severely over-betting or under-betting.

This is the critical gap that many Kelly explanations skip: before you can size your bets, you need to actually have an edge. And to know if you have an edge, you need to estimate win probability more accurately than the betting market.

Why probability estimation is hard:

The odds set by sportsbooks already reflect sophisticated probability assessments. The -110 line on a spread implies roughly 52.4% probability (after accounting for vig). To use Kelly profitably, you need to believe your estimate is better than the market's estimate, and you need to be right about that belief.

Three practical approaches to estimating win probability:

Method	Description	Pros	Cons
Market-Based Estimate	Use sharp books or no-vig lines as your baseline, then adjust based on your analysis	Starts from an informed position; reduces overconfidence	Hard to beat efficient markets; requires access to sharp lines
Model-Based Estimate	Build a statistical model using historical data and relevant factors	Quantifiable and testable; can identify systematic edges	Time-intensive; requires data and technical skills
Line Shopping	Find odds significantly better than the market consensus	The edge is built into the better odds; less estimation required	Edges are small; requires multiple accounts

The expected value connection: Win probability and expected value are two sides of the same coin. If you believe a bet has a 55% chance of winning at -110 odds, you are also saying the bet has positive expected value of approximately 5%. Our EV Calculator guide helps you visualize this relationship and find plus-EV bets before sizing them with Kelly.

Common probability estimation mistakes:

• Ignoring the vig: The break-even probability at -110 is 52.4%, not 50%. Many bettors forget this and overestimate their edge.
• Recency bias: Weighting recent games too heavily relative to season-long performance.
• Confirmation bias: Finding reasons to bet teams you want to bet rather than objectively assessing probability.
• Overconfidence: Thinking you can consistently estimate probabilities within 1-2% accuracy when even experts struggle to do this.
• Ignoring market efficiency: Sportsbooks employ teams of analysts and use sophisticated models. Beating them consistently requires either specialized knowledge, faster information, or identifying systematic blind spots in their pricing.
• Confusing luck with skill: A few winning bets do not prove you can estimate probabilities accurately. You need hundreds of bets to separate skill from variance.

How small errors compound: Consider what happens when your probability estimate is off by just 3 percentage points. If you believe a bet has a 55% chance of winning but the true probability is 52%, your perceived 5% Kelly stake becomes severely over-sized. Over many bets, this error compounds. You are betting as if you have a meaningful edge when you actually have almost none after accounting for the vig. This is why fractional Kelly provides such important protection.

Building estimation skill over time: The best approach for most bettors is to start conservatively and track results meticulously. Record your estimated probability for every bet, then compare your predictions to actual outcomes over time. After several hundred bets, you can calculate whether your estimates are calibrated. If you estimated 55% winners actually win at 55%, your estimates are accurate. If they win at 50%, you are systematically overconfident and should adjust.

The honest truth is that most recreational bettors do not have reliable probability estimates. If you are not sure whether you can beat the market, you probably cannot. In that case, Kelly-based staking may give you false confidence in bets that do not actually have positive expected value.

Read our comprehensive Expected Value betting guide to develop a framework for identifying genuine edges before applying Kelly bet sizing.

Kelly for Sports Betting: Real Market Examples

Theory is useful, but seeing Kelly applied to real betting scenarios makes the concepts concrete. Here are worked examples across different market types, demonstrating how Kelly adapts to different odds and edges.

Example 1: NFL Spread at -110

You are betting Buffalo Bills +3 at -110 against the Kansas City Chiefs. After analyzing the matchup, you estimate Buffalo has a 54% chance of covering.

• Bankroll: 5000 dollars
• Decimal odds: 1.909 (b = 0.909)
• Win probability: 54%
• Implied probability at -110: 52.4%
• Edge: 1.6%

Kelly % = (0.909 x 0.54 - 0.46) / 0.909 = 3.4%

Stake Type	Percentage	Dollar Amount
Full Kelly	3.4%	170 dollars
Half Kelly	1.7%	85 dollars
Quarter Kelly	0.85%	42.50 dollars

Example 2: NBA Total at -110

You are betting Under 224.5 in an NBA game at -110. Your model projects a total of 220 points, giving you a 58% chance of the under hitting.

• Bankroll: 5000 dollars
• Win probability: 58%
• Edge: 5.6%

Kelly % = (0.909 x 0.58 - 0.42) / 0.909 = 11.8%

This is a large stake because the edge is significant. But notice the risk: if your model is wrong and the true probability is 53%, you would be massively over-betting. Quarter Kelly at 2.95% (147.50 dollars) provides protection.

Example 3: Underdog Moneyline at +140

You are betting the Detroit Lions moneyline at +140 as home underdogs. You estimate they have a 45% chance of winning outright.

• Bankroll: 5000 dollars
• Decimal odds: 2.40 (b = 1.40)
• Win probability: 45%
• Implied probability at +140: 41.7%
• Edge: 3.3%

Kelly % = (1.40 x 0.45 - 0.55) / 1.40 = 5.7%

Stake Type	Percentage	Dollar Amount
Full Kelly	5.7%	285 dollars
Half Kelly	2.85%	142.50 dollars
Quarter Kelly	1.43%	71.50 dollars

Example 4: Player Prop (High Variance)

You are betting Patrick Mahomes Over 285.5 passing yards at -115. You estimate a 57% chance based on matchup analysis.

• Bankroll: 5000 dollars
• Edge: approximately 4%

Props deserve extra caution. They are higher variance than sides and totals, lines can be less efficient, and limits are often lower. For props, consider using Eighth Kelly rather than Quarter Kelly.

Kelly % = approximately 7.8% (Full) Eighth Kelly = approximately 1% = 50 dollars

This smaller stake accounts for the higher uncertainty in prop markets and protects your bankroll from the variance inherent in these bets.

Key Patterns Across These Examples

Looking at these four examples together reveals important patterns for applying Kelly in practice:

Edge size determines stake size: The NFL spread example with a 1.6% edge produces a much smaller Kelly stake than the NBA total with a 5.6% edge. This is exactly how Kelly should work. You bet more when you have more confidence in a larger edge, and less when edges are thin. However, thin edges are also where estimation errors hurt most, so fractional Kelly becomes even more important.

Odds format affects the calculation: Notice how the underdog moneyline at +140 produces a larger Kelly percentage than the spread at -110 for a similar edge. This is because the potential payout is higher on plus-money bets. Kelly accounts for this automatically, but you should understand why your stake sizes vary across different odds.

Market type influences your Kelly fraction: The progression from sides to totals to props generally means moving from Half Kelly toward Eighth Kelly. This reflects increasing uncertainty in your probability estimates and higher variance in outcomes. Spreads and totals in major sports are the most efficient markets with the most reliable probability estimates. Player props involve more variables and less liquid markets, warranting extra caution.

Always verify your edge makes sense: Before placing any Kelly-sized bet, ask yourself whether your edge is realistic. A 5% edge on a major NFL spread is rare. A 10% edge is almost certainly an estimation error on your part. If Kelly suggests a stake that feels too large, your probability estimate is probably too confident. Trust the feeling and reduce your fraction.

Seasonal and situational adjustments: Your Kelly fraction might change throughout a season. Early in the year when team compositions are uncertain, use smaller fractions. Later in the season when you have more data, you might move toward larger fractions if your model has proven accurate. Similarly, playoff games often have more public money and sharper lines, potentially reducing your edge even if your analysis is sound.

Convert Kelly Percent to Betting Units You Can Follow

Many bettors track their wagers in units rather than dollars. A unit is simply a standardized bet size, typically representing 1% of your bankroll. Using units makes it easier to compare performance across different bankroll sizes and time periods.

Kelly outputs a percentage of bankroll, which translates directly into a units framework. However, most bettors use flat units (1 unit per bet) while Kelly suggests variable sizing based on edge. You can get the benefits of both by using Kelly to determine when to bet more or fewer units.

Basic conversion:

If your standard unit is 1% of bankroll (100 dollars on a 10000 dollar bankroll), then:

• A 1% Kelly stake = 1 unit
• A 2.5% Kelly stake = 2.5 units
• A 0.5% Kelly stake = 0.5 units

Quarter Kelly Output	Units (if 1 unit = 1%)	Recommended Stake
0.5% or less	0.5 units	Half unit (minimum threshold bet)
0.5% to 1.5%	0.5 to 1.5 units	1 unit (standard bet)
1.5% to 2.5%	1.5 to 2.5 units	2 units (confident bet)
2.5% to 4%	2.5 to 4 units	3 units (high confidence)
Above 4%	Above 4 units	Cap at 3-4 units (verify edge)

Why cap your maximum stake: Kelly can suggest very large bets when you have a large perceived edge. A 10% Kelly stake might seem reasonable if you truly have a 10% edge, but such edges are rare in sports betting. Capping your maximum at 3-4 units protects you from overconfidence and ensures that no single bet can devastate your bankroll. Even professional bettors with proven track records rarely stake more than 3% of their bankroll on a single wager. If your Kelly calculation suggests more than 4 units, treat it as a red flag that your probability estimate may be too aggressive.

Recalculating your unit size: As your bankroll grows or shrinks, your unit size should adjust. Many bettors recalculate their unit size weekly or monthly based on their current bankroll. This ensures you are always betting an appropriate percentage rather than a fixed dollar amount. Some bettors only recalculate after significant changes, such as a 20% increase or decrease in bankroll. Others update their unit size after every session. Find a rhythm that works for you, but avoid adjusting so frequently that you are chasing short-term results.

The units approach combined with Kelly thinking gives you the best of both worlds: consistent tracking and performance measurement, plus intelligent adjustment based on edge size. For more on unit-based betting systems, see our guide on betting units explained. For a complete framework on protecting and growing your betting bankroll over time, read our bankroll management guide.

Kelly vs Flat Betting: Growth vs Drawdowns

Should you use Kelly or stick with flat betting where every wager is the same size? Both approaches have merit, and the right choice depends on your confidence in your edge estimates and your tolerance for variance.

Flat betting advantages:

• Simplicity: Every bet is the same size. No calculations required.
• Emotional stability: You never have to second-guess whether a large bet was appropriate.
• Harder to blow up: A losing streak hurts the same amount regardless of which games you lost.
• Works without probability estimates: You can flat bet successfully without knowing your exact edge, as long as you have one.

Kelly betting advantages:

• Faster growth: When edges are real and estimates are accurate, Kelly compounds your bankroll more efficiently.
• Automatic scaling: You bet more when you have bigger edges and less when edges are thin.
• Bankroll protection: Kelly naturally reduces stakes as your bankroll shrinks, preventing total ruin.
• Forces discipline: You cannot bet on negative EV spots because Kelly returns zero or negative stakes.

Factor	Flat Betting	Kelly Betting
Ease of use	Very easy	Requires probability estimates
Growth potential	Moderate	Higher (with accurate estimates)
Drawdown risk	Predictable	Variable (lower with fractional Kelly)
Psychological difficulty	Low	Higher (variable stake sizes)
Best for	Most recreational bettors	Serious bettors with proven models

When NOT to use Kelly:

• You cannot reliably estimate win probabilities better than the market
• You are betting recreationally without tracking detailed records
• Your edges are so thin that even small estimation errors flip them negative
• You struggle emotionally with variable bet sizing
• You are betting highly correlated positions (same game, same player)

The hybrid approach: Many successful bettors use a middle ground. They use flat units for most bets and increase to 2-3 units only when they have high confidence in their edge. This captures some of the Kelly upside while maintaining the psychological benefits of mostly-consistent sizing.

Progression matters: If you are new to sports betting, starting with flat betting makes sense. You can focus on developing your handicapping skills without adding the complexity of variable stake sizing. Once you have a track record of several hundred bets and can demonstrate that your probability estimates are reasonably accurate, you can introduce Kelly concepts gradually. Start with a very conservative fraction like Eighth Kelly to see how it feels before moving to Quarter Kelly.

The psychological factor: Do not underestimate the mental component. Losing a large Kelly-sized bet can shake your confidence far more than losing several smaller flat bets that add up to the same amount. If you find yourself second-guessing your stakes or avoiding bets because the Kelly stake feels too large, you are better off with flat betting until your confidence and track record improve.

The bottom line: Kelly is a tool for bettors who have genuine, measurable edges and the discipline to estimate probabilities honestly. For everyone else, flat betting with good game selection is likely to produce better long-term results.

Correlation, Parlays, and Portfolio Exposure Caps

The Kelly Criterion assumes each bet is independent. In sports betting, this assumption often breaks down. Multiple bets on the same game, same player, or same narrative can create correlated risk that Kelly does not account for.

Why correlation matters:

If you bet the Chiefs spread and the Chiefs team total over in the same game, both bets share common risk factors. If the Chiefs offense struggles, both bets lose together. Kelly treats each bet independently, so it would size both as if they had no relationship. This can lead to significant overexposure.

Common correlation traps:

• Same-game parlays (SGPs): By definition, SGP legs are correlated. Stacking player props in an SGP compounds this.
• Multiple bets on the same team: Spread, moneyline, and totals for the same game share injury risk, weather, and game flow.
• Player props from the same game: If a game becomes a blowout, multiple player unders might hit together.
• Narrative-driven bets: Multiple bets based on the same thesis (like expecting a team to bounce back) are correlated.

Exposure caps for safety:

To manage correlation, experienced bettors use exposure caps that limit total risk regardless of what Kelly suggests.

• Per-game cap: Maximum 3-5% of bankroll exposed to any single game, across all bets
• Per-player cap: Maximum 2% of bankroll on any individual player
• Per-day cap: Maximum 10-15% of bankroll at risk on any given day
• Correlated position rule: When betting correlated positions, treat them as a single bet for Kelly purposes

Practical rules for correlated bets:

• If you bet a spread and a total in the same game, size each at half of what Kelly would suggest for a single bet
• Avoid stacking props from the same game unless you reduce each stake to Eighth Kelly or less
• For SGPs, calculate Kelly on the parlay itself, not the individual legs
• When multiple bets share a common risk factor (injury, weather, referee), aggregate your exposure

Parlays and Kelly: Traditional Kelly assumes single bets, not parlays. For parlays, you need to estimate the probability of the entire parlay hitting, which requires accurate assessment of correlation between legs. Most bettors should avoid applying Kelly to parlays and instead use minimal stakes for these higher-risk wagers.

The goal is to survive long enough to realize your edge. Correlation creates hidden risk that can blow up your bankroll even when individual bets seem reasonable. When in doubt, reduce stakes and cap exposure.

Kelly in Excel: DIY Calculator + Template Setup

Building a Kelly calculator in Excel or Google Sheets gives you full control over inputs and allows for customization. Here is how to set up a basic Kelly template.

Step 1: Set up your inputs

Create cells for:

• Bankroll (Cell B1)
• Odds in American format (Cell B2)
• Your estimated win probability (Cell B3)
• Kelly fraction you want to use (Cell B4, default to 0.25 for Quarter Kelly)

Step 2: Convert American odds to decimal

In Cell B5, use this formula to convert American odds to decimal:

For positive odds: =(B2/100)+1 For negative odds: =(-100/B2)+1

Or use a combined formula: =IF(B2>0,(B2/100)+1,(-100/B2)+1)

Step 3: Calculate the b value

In Cell B6: =B5-1

This gives you the net profit per unit wagered.

Step 4: Calculate Full Kelly percentage

In Cell B7: =((B6*B3)-(1-B3))/B6

This returns the Kelly percentage. If negative, you have no edge.

Step 5: Calculate your stake

In Cell B8: =MAX(0,B7B4B1)

This multiplies Full Kelly by your fraction and bankroll, and ensures the result is not negative.

Common Excel errors to avoid:

• Odds format confusion: Make sure you are entering American odds correctly. A -110 bet should be entered as -110, not 110.
• Probability as decimal: Enter 55% as 0.55, not 55.
• Negative Kelly display: Use MAX(0,formula) to prevent showing negative stakes.
• Rounding errors: Format the output to show appropriate decimal places for your currency.

Extending the template:

Once you have the basic formula working, you can add:

• A row for each bet to track multiple wagers
• Automatic implied probability calculation from odds
• Expected value calculation alongside Kelly stake
• Running bankroll tracker that updates after each bet
• Conditional formatting to highlight negative Kelly bets in red
• A summary section showing your total exposure and daily betting limits
• Charts to visualize your bankroll growth over time

Advantages of a DIY spreadsheet: Building your own Kelly calculator forces you to understand the math behind the formula. When you construct the formulas yourself, you develop intuition for how changes in probability and odds affect your stake sizes. This understanding is valuable even if you later switch to using an online calculator.

For a more streamlined experience, use our Kelly Criterion calculator which handles all conversions automatically and provides Half and Quarter Kelly outputs.

Frequently Asked Questions

What is the formula for the Kelly Criterion?

The Kelly Criterion formula is: Kelly Percentage = (bp - q) / b, where b equals the decimal odds minus 1, p equals your win probability, and q equals your loss probability (1 - p). For example, at -110 odds with a 55% estimated win probability, b = 0.909, p = 0.55, and q = 0.45. The calculation gives (0.909 x 0.55 - 0.45) / 0.909 = 5.5% of bankroll. Most bettors use a fraction of this amount for safer staking.

Does the Kelly Criterion really work?

The Kelly Criterion works mathematically when two conditions are met: you have a genuine edge, and your probability estimates are accurate. It has been proven to maximize long-term bankroll growth. However, it requires you to actually have an edge over the betting market and to estimate that edge correctly. Many bettors fail at one or both of these requirements. Additionally, Full Kelly produces extreme volatility, which is why most practitioners use fractional Kelly (typically Quarter Kelly) to reduce risk.

Why is Kelly Criterion risky?

Kelly Criterion is risky because it assumes perfect probability estimates, which are impossible in practice. Small errors in your win probability estimate can lead to significant over-betting. Full Kelly also has high variance, with a roughly 50% chance of experiencing a 50% drawdown at some point. The formula does not account for correlated bets, and using Kelly on multiple related positions can compound risk. These issues are why experienced bettors use fractional Kelly, typically staking only 25% of what the full formula suggests.

Can you go broke using the Kelly Criterion?

In theory, pure Kelly betting cannot reduce your bankroll to exactly zero because you always bet a percentage rather than a fixed amount. However, you can effectively go broke by losing so much that you can no longer place meaningful bets. If your probability estimates are wrong and you are betting without an edge, Kelly will still lose money over time. Over-aggressive Kelly fractions, betting on correlated positions, and using Kelly on negative expected value bets can all lead to catastrophic losses. Fractional Kelly and exposure caps provide important protection.

What is fractional Kelly and which fraction should I use?

Fractional Kelly means betting a fraction of the full Kelly stake, such as half (50%), quarter (25%), or eighth (12.5%). This reduces both potential growth and volatility. For most sports bettors, Quarter Kelly is recommended as it provides meaningful growth while protecting against estimation errors. Use Half Kelly only if you have a proven model with a long track record. Use Eighth Kelly or less for high-variance props or when probability estimates are particularly uncertain. The key principle is: the less confident you are in your estimate, the smaller your Kelly fraction should be.

How do you estimate win probability for Kelly?

Estimating win probability is the hardest part of using Kelly. Three main approaches exist. Market-based estimation uses sharp sportsbook lines as a baseline and adjusts based on your analysis. Model-based estimation uses statistical models built on historical data. Line shopping finds odds significantly better than consensus, where the edge is built into the price difference. Whichever method you use, your estimates must be more accurate than the betting market to have a genuine edge. Most bettors overestimate their ability to beat the market, which leads to over-betting with Kelly.

Kelly Criterion vs flat betting: which is better?

Neither is universally better. Kelly betting produces faster bankroll growth when you have accurate probability estimates and genuine edges. Flat betting is simpler, more emotionally stable, and does not require precise edge calculations. For most recreational bettors, flat betting is the better choice because it is hard to consistently estimate probabilities more accurately than the betting market. Kelly is best for serious bettors with proven models and the discipline to accept variable sizing. A hybrid approach, using flat units with occasional larger bets on high-confidence plays, offers a middle ground.

Responsible Gambling Notice

The Kelly Criterion calculator is an educational tool designed to help you think about bet sizing in a structured way. It does not guarantee profits and cannot eliminate the risk inherent in sports betting.

Key points to remember:

• You must be 21 or older to bet in most US states. Availability varies by state.
• Only bet money you can afford to lose completely.
• The calculator outputs are based on your inputs. If your probability estimates are wrong, the suggested stakes will be inappropriate.
• No betting system can overcome a lack of genuine edge. If you do not have an advantage, disciplined staking will not create one.
• Sports betting should be entertainment, not a source of income or a way to recover financial losses.

If gambling is causing problems in your life or the life of someone you know, help is available.