NBA Bet Sizing Strategies: Bankroll Management and Unit Allocation
Available in US
The worst month of my betting career was not my worst win rate – it was the month I sized my bets terribly. I had a 56% winner but lost money because I bet 5 units on my losers and 1 unit on my winners. The emotional confidence that drove my sizing was exactly backwards from my actual edge. That expensive lesson taught me that sizing discipline matters as much as selection skill.
Bet sizing is the connector between having edge and realising profit. You can identify value perfectly but destroy returns through poor allocation. You can identify value imperfectly but maximise returns through optimal allocation. The maths of betting is unforgiving – sizing mistakes compound just like selection mistakes.
Unit System Fundamentals
The unit system is the foundation of responsible betting. A unit represents a standard bet size, typically 1-2% of total bankroll. This standardization allows consistent tracking and prevents emotional bet sizing.
Defining your unit starts with defining your bankroll. This must be money you can genuinely afford to lose – not rent money, not savings, not money with any other designated purpose. Betting money should be disposable income allocated specifically for betting entertainment and potential profit.
I recommend 1% of bankroll as a standard unit for most bettors. A $10,000 bankroll means $100 standard bets. This size provides enough cushion to survive normal variance while being meaningful enough to maintain engagement. Smaller bankrolls might use 2% units; larger professional bankrolls often use 0.5% or less.
Fixed unit sizing means betting the same amount on every play regardless of confidence. This is the safest approach for bettors still developing their skills. It removes the emotional sizing decisions that often hurt returns and provides clean data for performance evaluation.
The alternative – variable unit sizing – increases bets on higher-confidence plays. This can maximise returns for skilled bettors but requires accurate confidence calibration. Most bettors overestimate confidence on their losers and underestimate it on their winners, making variable sizing dangerous without proven track records.
Kelly Criterion
The Kelly Criterion is the mathematically optimal bet sizing strategy for maximising long-term bankroll growth. It prescribes betting a percentage of bankroll equal to your edge divided by the odds you receive.
The formula: Bet Size = (Edge / Odds) x Bankroll. If you have 5% edge on a -110 bet, Kelly suggests betting (5% / 110%) x Bankroll = 4.55% of bankroll. This is aggressive sizing that most recreational bettors should not use at full strength.
Full Kelly creates extreme variance. The mathematically optimal path involves large bets that can produce significant drawdowns. A 10-bet losing streak at Kelly sizing would devastate a bankroll even if long-term expectation was positive. The psychological toll exceeds what most humans can handle.
Fractional Kelly reduces variance at the cost of slightly slower expected growth. Half-Kelly (betting half of what full Kelly recommends) produces 75% of the expected growth rate with much smaller drawdowns. Quarter-Kelly produces 56% of growth rate with minimal variance. I use quarter-Kelly as my ceiling for any single bet.
The challenge with Kelly is accurately estimating edge. The formula assumes you know your true probability. Overestimating edge leads to oversizing bets and faster bankroll depletion. Most bettors should be conservative with edge estimates when using Kelly-based approaches.
Bankroll Management
Managing bankroll extends beyond individual bet sizing to overall gambling portfolio construction.
Segregate betting bankroll from life money completely. I maintain separate accounts and never move money between them except during designated review periods. This separation prevents the temptation to chase losses with funds that should not be risked.
Set stop-loss limits for sessions, days, and weeks. If I am down 5 units in a day, I stop betting until tomorrow regardless of remaining opportunities. This rule prevents tilt-driven decisions when emotions are running hot. The games will be there tomorrow; protecting capital is more important than any single day’s action.
Establish win targets as well. If I am up 10 units in a day, I often reduce subsequent bet sizes or stop entirely. This locks in profits and prevents the common mistake of giving back gains through overconfidence. Winning days should end with money secured, not with continued risk-taking.
Review and adjust bankroll allocation periodically. If a winning streak has grown my bankroll significantly, I recalculate unit size upward. If a losing streak has depleted it, I recalculate downward. The percentage-based approach automatically scales, but explicit review ensures discipline.
Never chase losses. The desire to “get even” after a bad day leads to the worst bet sizing decisions. Each bet should be evaluated independently based on its own merits, not based on what happened in previous bets. Emotional accounting is the enemy of bankroll preservation.
Confidence Tiers
If you choose variable bet sizing, structuring confidence tiers provides discipline that pure intuition lacks.
I use a 3-tier system: standard (1 unit), elevated (2 units), and maximum (3 units). Each tier has explicit criteria. Standard bets meet my baseline threshold for positive expectation. Elevated bets meet baseline plus additional confirming factors. Maximum bets require multiple strong confirming factors with no significant concerns.
The criteria must be defined in advance, not rationalized after the fact. My elevated tier requires: positive expected value above 5%, line value confirmed by reverse line movement, and no significant injury uncertainty. Missing any criterion keeps the bet at standard tier regardless of gut feeling.
Track results by tier separately. If elevated bets are not actually more profitable than standard bets, your confidence calibration is wrong. The data will reveal whether your tiering adds value or just adds variance.
Resist the temptation to create too many tiers. A 10-tier system from 0.5 units to 5 units provides too many decision points and invites emotional manipulation. Simple structures with clear criteria are more consistently applied than complex systems.
Maximum bets should be rare. If you are playing max confidence daily, either your criteria are too loose or you are fooling yourself about confidence levels. I use maximum tier perhaps 2-3 times per month – genuinely exceptional situations where multiple independent indicators align.
Common Sizing Mistakes
Recognising common errors helps avoid them. I have made all of these mistakes at various points in my betting development.
Oversizing early in a bankroll’s life. The temptation to grow quickly leads to aggressive sizing that cannot withstand normal variance. A bankroll needs time to prove that the underlying edge is real before aggressive sizing is appropriate.
Sizing based on previous results. Winning streaks do not increase true probability of future wins. Losing streaks do not decrease it. Each bet has its own probability distribution independent of recent history. Sizing should reflect current bet quality, not past outcomes.
Increasing size to recover losses. The brutal reality of recovery through increased betting is this: down 10 units, doubling bet size does not halve recovery time – it dramatically increases probability of catastrophic loss. Recovery happens through continued positive expectation betting, not through betting larger.
Sizing based on payout rather than edge. A +500 underdog does not automatically deserve larger bets than a -110 spread. Bet size should reflect expected value, which requires estimating true probability against the odds offered. High payouts with low probability can have smaller expected value than modest payouts with high probability.
Ignoring opportunity cost. Money bet on a mediocre opportunity cannot be bet on a great opportunity that appears later. Over-allocation to early-day games might leave insufficient bankroll for late-day games that offer better value. Consider daily and weekly betting budgets, not just single-bet sizing.
Proper bet sizing will not turn a losing strategy into a winning one. But improper bet sizing can absolutely turn a winning strategy into a losing one. The discipline to size correctly is as important as the skill to select correctly.
What percentage of bankroll should I bet on each NBA game?
Standard recommendation is 1-2% of bankroll per bet. Conservative bettors use 1%; aggressive bettors might use 2-3% on high-confidence plays. Never bet more than 5% on any single wager. Proper sizing ensures you can survive normal variance while maintaining meaningful engagement.
What is the Kelly Criterion for sports betting?
The Kelly Criterion calculates optimal bet size as edge divided by odds, multiplied by bankroll. Full Kelly maximises expected growth but creates extreme variance. Most bettors should use fractional Kelly (25-50% of calculated amount) to reduce drawdown risk while maintaining growth potential.
Should I bet more on games I am confident about?
Variable sizing can increase returns if your confidence is well-calibrated. However, most bettors overestimate confidence on their losers. Start with flat betting to establish a track record, then consider tiered sizing only after proving your confidence correlates with actual outcomes.
Created by the ”Betting Stats nba” editorial team.
