A Complete Guide to Determine How Much to Stake on NBA Game

When I first started betting on NBA games, I thought it was all about picking winners. I'd spend hours analyzing matchups, studying player stats, and watching pre-game shows, only to consistently lose money despite being right about who would win more often than not. It took me losing a significant portion of my bankroll to realize what should have been obvious from the start: knowing who will win is only half the battle. The real key to long-term profitability lies in determining the optimal stake for each wager. This is where most casual bettors fail, and where sharp bettors separate themselves from the crowd.

I remember one particular season where I correctly predicted 58% of my NBA picks but still finished down financially because my bet sizing was all over the place. I'd bet $500 on a game I felt strongly about, then $100 on another that I was less certain of, but my losses on the big bets wiped out all my smaller wins. That experience taught me that without a disciplined staking strategy, even good handicapping can't save you. Over the years, I've developed a framework that has consistently helped me maintain profitability, and it revolves around three core principles: bankroll management, edge calculation, and emotional discipline. Let me walk you through how I approach each game now.

Your bankroll is your ammunition, and protecting it should be your number one priority. I always recommend starting with what I call your "risk capital"—money you can afford to lose completely without impacting your lifestyle. For me, that initial bankroll is divided into units, with each unit representing 1-2% of the total. So if you have $1,000 to start with, each unit would be $10-$20. The beauty of this system is that it automatically scales your bets to your current bankroll size. When I'm up, my unit size increases; when I'm down, it decreases. This prevents the dreaded "chasing losses" scenario where bettors dramatically increase their stakes after a losing streak, often digging themselves into a deeper hole. I've found that keeping each bet between 1% and 3% of my total bankroll has allowed me to withstand the inevitable losing streaks that every bettor experiences.

Now, not all bets are created equal, and this is where many bettors miss opportunities. I use a simple but effective method to quantify my confidence in each pick. If I believe a team has a 55% chance of covering the spread rather than the implied 50% (assuming standard -110 odds), that represents a 5% edge. The Kelly Criterion formula suggests betting a percentage of your bankroll equal to your edge divided by the odds. In practice, I use a fractional Kelly approach—usually half-Kelly or quarter-Kelly—to be more conservative. For instance, if I calculate a 10% edge on a particular game, a full Kelly would suggest betting 10% of my bankroll, but I'd typically scale that down to 2.5-5% instead. This approach has saved me from overbetting on games where I was overly confident. Last season, I identified what I thought was a massive edge on a Warriors vs. Rockets game—Golden State was missing two starters, but the line hadn't adjusted enough in my estimation. My calculations suggested a 12% edge, which would have meant a substantial bet using full Kelly. Instead, I stuck with my quarter-Kelly approach and placed 3% of my bankroll on Houston. The Rockets won outright, and while I didn't maximize my profit, I avoided potential disaster when the game came down to a last-second shot.

The psychological aspect of betting is where most people struggle, myself included in my early days. I've learned to recognize when I'm betting for the wrong reasons—boredom, frustration, or the desire to make back losses quickly. Now, I have a strict rule against making impulsive bets. Every wager I place is documented in my betting journal beforehand, including my calculated edge and predetermined stake. This removes emotion from the equation in the moment. Another technique I use is what I call "confidence tiers." I categorize my bets into A, B, and C plays based on the strength of my analysis and edge calculation. A-plays are my strongest convictions with edges over 7%, B-plays are moderate confidence with 3-7% edges, and C-plays are smaller edges under 3%. My stake sizes reflect these tiers, with A-plays getting 2-3% of my bankroll, B-plays 1-2%, and C-plays 0.5-1%. This system has helped me avoid overbetting on games where I liked a team but didn't have a significant mathematical edge.

Let me give you a concrete example from last season that illustrates my complete process. The Lakers were playing the Grizzlies as 4-point favorites, but my research indicated Memphis had value. Anthony Davis was questionable with back spasms, and the Grizzlies were on the second night of a back-to-back, which typically causes an overreaction in the line. After running my numbers, I calculated the Grizzlies had approximately a 54% chance of covering against the implied 48% from the -110 odds—a 6% edge. With a bankroll of $5,000 at the time, my quarter-Kelly calculation suggested a bet of 1.5% or $75. Since this fell into my B-play category, I placed $75 on Memphis +4. The Grizzlies lost by 2 points, covering the spread, and the win added $68 to my bankroll. More importantly, it reinforced the value of sticking to a predetermined staking plan regardless of outcome.

Of course, no system is perfect, and I've had to adapt my approach over time. Early in my betting career, I placed too much emphasis on recent performance, falling victim to what behavioral economists call "recency bias." I'd see a team on a hot streak and overestimate their chances, leading to overbetting. Now, I incorporate more regression to the mean in my calculations. I also pay close attention to situational factors that the market might be slow to react to—things like scheduling spots, rest advantages, and emotional letdown spots. These qualitative factors help me identify edges that pure statistics might miss. Another adjustment I've made is being more selective with my bets. In the past, I felt compelled to bet on games every night, but experience has taught me that sometimes the best bet is no bet at all. Last season, I placed only 47 bets on NBA games despite there being over 1,200 games in the season. This selectivity ensures I'm only betting when I have a genuine edge.

Ultimately, determining how much to stake on an NBA game comes down to balancing mathematical precision with psychological discipline. The system I've developed works for me, but it requires constant maintenance and honest self-assessment. I regularly review my betting journal to identify patterns in both my wins and losses, adjusting my approach based on what the data tells me. The most important lesson I've learned is that successful betting isn't about winning every wager—it's about making mathematically sound decisions over the long run. If you can maintain discipline with your staking strategy while continuously working to improve your handicapping, you'll put yourself in position to profit regardless of short-term variance. Remember, the goal isn't to be right—it's to make money.