Kelly Criterion Real Betting Examples

The Kelly Formula:

Kelly % = (b × p - q) / b

Where b = decimal odds minus 1 (your net profit per $1 wagered), p = your estimated probability of winning, and q = 1 - p (probability of losing).

Or equivalently: Kelly % = (Decimal Odds × p - 1) / (Decimal Odds - 1)

A positive result tells you what percentage of your bankroll to wager. A negative result means you have no edge and should skip the bet entirely. For a deeper explanation of the formula itself, check our Simple Kelly Criterion Guide.

The Scenario: You are looking at Chiefs -3.5 vs. Raiders, priced at -110 on the spread. After studying the matchup, reviewing recent performance, and analyzing key injuries, you estimate the Chiefs cover 56% of the time. Your current bankroll is $5,000.

Step 1: Convert American Odds to Decimal

For negative American odds, the conversion is: Decimal = 1 + (100 / |Odds|). So for -110: Decimal = 1 + (100 / 110) = 1.909.

Step 2: Plug Into the Kelly Formula

Kelly % = (Decimal Odds × p - 1) / (Decimal Odds - 1)

Kelly % = (1.909 × 0.56 - 1) / (1.909 - 1)

Kelly % = (1.069 - 1) / 0.909

Kelly % = 0.069 / 0.909 = 7.6%

Step 3: Apply Full Kelly vs. Fractional Kelly

Full Kelly says to bet 7.6% of $5,000 = $380. That is a significant chunk of your bankroll on a single NFL spread bet, and it assumes your 56% win rate estimate is perfectly accurate.

Half Kelly (50%) = 3.8% of $5,000 = $190. This is a much safer approach that protects you against overestimating your edge.

Quarter Kelly (25%) = 1.9% of $5,000 = $95. Conservative, but ideal if you are still building your track record.

Most serious bettors, including us at BetLegend, recommend fractional Kelly because even a small overestimate in your win probability can lead to overbetting. The standard -110 spread bet is the most common scenario you will encounter, so this example is worth memorizing.

The Scenario: The Grizzlies are +180 underdogs at the Celtics. You have done your homework and believe Memphis wins this game 40% of the time given the matchup dynamics. Your bankroll is $5,000.

Step 1: Convert +180 to Decimal

For positive American odds: Decimal = 1 + (Odds / 100). So for +180: Decimal = 1 + (180 / 100) = 2.80.

Step 2: Plug Into the Kelly Formula

Kelly % = (2.80 × 0.40 - 1) / (2.80 - 1)

Kelly % = (1.12 - 1) / 1.80

Kelly % = 0.12 / 1.80 = 6.7%

Step 3: Full Kelly vs. Fractional Kelly

Full Kelly = 6.7% of $5,000 = $335.

Half Kelly = 3.35% of $5,000 = $167.50.

Quarter Kelly = 1.675% of $5,000 = $83.75.

Here is something interesting about underdog bets and the Kelly Criterion: even though you expect to lose this bet 60% of the time, the juicy +180 payout creates enough value that Kelly still recommends a healthy 6.7% position at full size. That is the beauty of value betting on underdogs. You do not need to win often. You just need to win more often than the odds imply. The implied probability of +180 is only 35.7%, but you believe the true probability is 40%, giving you a solid edge.

The Scenario: Orioles +160 vs. Yankees. You give the Orioles a 45% chance to win. Bankroll: $5,000.

Step 1: Convert +160 to decimal: 1 + (160/100) = 2.60

Step 2: Kelly % = (2.60 × 0.45 - 1) / (2.60 - 1)

Kelly % = (1.17 - 1) / 1.60 = 0.17 / 1.60 = 10.6%

Step 3: Full vs. Fractional

Full Kelly = 10.6% of $5,000 = $530. That is extremely aggressive for a single baseball game.

Half Kelly = 5.3% of $5,000 = $265, but most bettors cap individual bets at 5% of bankroll = $250.

Quarter Kelly = 2.65% of $5,000 = $132.50.

When Kelly spits out a number over 10%, it usually means one of two things: either you have found a genuinely massive edge, or you are overestimating your win probability. In MLB especially, where even the best teams lose 60+ games a year, it pays to be skeptical of any estimate above 50%. That is why using a disciplined bankroll management strategy with a hard cap (like 5% max per bet) is so important alongside Kelly.

The Scenario: You like the Panthers -1.5 on the puck line, priced at -130. You have studied the matchup and estimate Florida covers the 1.5-goal spread 52% of the time. Your bankroll is $4,000.

Step 1: Convert -130 to Decimal

Decimal = 1 + (100 / 130) = 1 + 0.769 = 1.769

Step 2: Plug Into the Kelly Formula

Kelly % = (1.769 × 0.52 - 1) / (1.769 - 1)

Kelly % = (0.9199 - 1) / 0.769

Kelly % = -0.0801 / 0.769 = -10.4%

Step 3: Interpret the Result

Kelly says: DO NOT BET. The result is negative, which means your estimated 52% win rate is not high enough to overcome the -130 juice on this puck line. The breakeven win rate at -130 is 56.5%, so your 52% estimate falls well short.

This is actually a great example of how Kelly protects you. Hockey puck lines are notoriously tricky because the -1.5 spread requires a multi-goal victory, and the vig on puck lines is often steeper than standard -110 pricing. Just because you think a team will cover more often than not does not mean the bet has value. The price has to be right too.

Now, what if the line was better? Let's say you shop around and find Panthers -1.5 at +110 instead. Let's recalculate:

Decimal = 1 + (110/100) = 2.10

Kelly % = (2.10 × 0.52 - 1) / (2.10 - 1) = (1.092 - 1) / 1.10 = 0.092 / 1.10 = 8.4%

Same game, same probability estimate, but the plus-money price flips the Kelly result from "do not bet" to an 8.4% position. That is the power of line shopping, and it is why every serious bettor should have accounts at multiple sportsbooks.

The Scenario: Warriors/Suns Over 228.5 at -115. You estimate the over hits 59% of the time. Bankroll: $6,000.

Step 1: Convert -115 to decimal: 1 + (100/115) = 1.870

Step 2: Kelly % = (1.870 × 0.59 - 1) / (1.870 - 1)

Kelly % = (1.1033 - 1) / 0.870 = 0.1033 / 0.870 = 11.9%

Step 3: Full vs. Fractional

Full Kelly = 11.9% of $6,000 = $714. That is a massive bet, and you would need supreme confidence in your 59% estimate.

Half Kelly = 5.95% of $6,000 = $357, capped at 5% = $300.

Quarter Kelly = 2.975% of $6,000 = $178.50.

Totals bets can be some of the most reliable edges in sports betting because the market sometimes overreacts to team narratives while the pace and efficiency numbers tell a different story. But 11.9% is still a flashing warning sign that you should double-check your assumptions. Try our Kelly Criterion Simulator to see how bets of this size play out over hundreds of wagers.

The Scenario: You want to parlay two legs: Packers -3 at -110 (you estimate 57% to cover) and Nuggets ML at -150 (you estimate 64% to win). The combined parlay odds offered by your sportsbook are +250. Your bankroll is $5,000.

Step 1: Calculate the True Combined Probability

For a parlay, multiply the probabilities of each independent leg:

Combined probability = 0.57 × 0.64 = 0.3648 (36.48%)

Step 2: Convert +250 to Decimal

Decimal = 1 + (250/100) = 3.50

Step 3: Apply Kelly to the Parlay

Kelly % = (3.50 × 0.3648 - 1) / (3.50 - 1)

Kelly % = (1.2768 - 1) / 2.50

Kelly % = 0.2768 / 2.50 = 11.1%

Step 4: Fractional Kelly

Full Kelly = 11.1% of $5,000 = $555.

Half Kelly = 5.55% of $5,000 = $277.50, capped at 5% = $250.

Quarter Kelly = 2.775% of $5,000 = $138.75.

Parlays are where the Kelly Criterion gets especially interesting. The formula works the same way, but you need to be honest about the combined probability. Notice that even though each individual leg has a decent edge, the combined probability drops below 37%. You are going to lose this parlay roughly two out of every three times. That is why fractional Kelly matters so much with parlays. Full Kelly on a bet you lose 63.5% of the time will cause wild bankroll swings. Quarter Kelly keeps you in the game. For more on using Kelly with parlays, see our Kelly for Parlays guide.

The Scenario: Lakers -110 to cover the spread. You have studied the game and honestly estimate them at 51% to cover. Bankroll: $5,000.

Step 1: Convert -110 to decimal: 1 + (100/110) = 1.909

Step 2: Kelly % = (1.909 × 0.51 - 1) / (1.909 - 1)

Kelly % = (0.9736 - 1) / 0.909

Kelly % = -0.0264 / 0.909 = -2.9%

Result: Negative Kelly = No Edge. Skip this bet.

This is the most important Kelly Criterion example on this entire page, because the hardest part of sports betting is not finding winners. It is having the discipline to pass on bets where you do not have a real edge. A 51% win rate on a -110 bet means you are essentially flipping a weighted coin with a house vig working against you. The breakeven point at -110 odds is 52.4%, so your 51% estimate puts you on the wrong side of the math.

Here is a useful rule of thumb: for standard -110 bets, you need to win at least 52.4% of the time just to break even. That means Kelly will return a negative number for any win probability below that threshold. And honestly, if your estimate is anywhere close to 52%, the edge is so razor-thin that even a slight miscalculation wipes it out. That is why many sharp bettors set a minimum Kelly threshold of 2% before pulling the trigger, as we discuss in our full Kelly Criterion guide.

Every example above shows calculations for full, half, and quarter Kelly. Here is a quick reference for when to use each:

Full Kelly (100%) maximizes long-term bankroll growth in theory, but it assumes your win probability estimates are perfectly accurate. In reality, they almost never are. Full Kelly leads to massive bankroll swings that most bettors cannot stomach.

Half Kelly (50%) is the most popular choice among professional bettors. It sacrifices roughly 25% of the theoretical growth rate but cuts the variance nearly in half. If you have a solid track record of at least 500+ graded bets, half Kelly is a strong approach.

Quarter Kelly (25%) is ideal for beginners, bettors still building their track record, or anyone who values bankroll preservation over maximum growth. You give up about 44% of the theoretical growth rate, but you dramatically reduce your risk of ruin.

Not sure which fraction is right for you? Our Fractional vs. Full Kelly breakdown goes deep on the tradeoffs, and the Kelly Simulator lets you see the difference in action over thousands of simulated bets.