Should you bet the same amount every time?
Before you develop the kind of long-run sample and calibrated confidence to vary your bet sizes, flat staking — betting the same amount on every play — is the right default. Pick a unit size (1–2% of your bankroll is standard) and stick to it regardless of how confident you feel.
The temptation to bet more on your "best bet of the week" is dangerous early on because confidence in sports betting is largely uncorrelated with accuracy until you have thousands of graded bets and a calibrated model behind it. Flat staking removes the most common way bettors amplify their mistakes.
Should beginners focus on one sport?
A bettor with deep knowledge of one conference, one division, or one sport will consistently outperform a bettor who spreads attention across every game on the board. The sportsbook has analysts covering every game full-time; your edge comes from the specific spots where your information or angle is better than average.
Pick one sport to focus on for a full season, ideally one you already watch and understand well. Betting what you actually know — team tendencies, coaching tendencies, injury contexts — is far more sustainable than scanning for statistical anomalies across 15 sports simultaneously.
Track everything from day one
Most recreational bettors have no idea whether they are winning or losing. Memory is selective and loss-blind — people remember their big wins vividly and blur their losses. Tracking every single bet from the start builds the data you need to evaluate your process.
Track: date, game, market, line you took, closing line, result, units won/lost. The minimum you need to know whether your process is working: win rate per market, units per bet type, and CLV average. Without this data, you are flying blind.
Shop lines across multiple books
Open accounts at three to five sportsbooks and check the number before placing every bet. A difference of half a point on a spread or 10 cents of juice adds up to several units over a season — often more than the edge in the bet itself.
Start with major US books (DraftKings, FanDuel, BetMGM, Caesars) and add a sharp book (Pinnacle if available in your region) for market comparison. The Pinnacle closing line is the most efficient reference price available publicly.
Judge process, not short-run results
The hardest thing about sports betting is that good decisions and bad decisions look the same for weeks or months. A sharp bettor with a real edge can lose 20 consecutive bets by variance alone; a lucky guesser can win 15 in a row. Short-run results tell you almost nothing useful.
Evaluate whether your lines were good when you made them — not whether the bet won. Did you get a better number than the close? Was your reasoning correct even if the outcome went the other way? A process focus keeps you from making reactive, result-driven adjustments to a strategy that might be working fine.
How big should a unit be on a $1,000 bankroll?
Between $10 and $20, using the standard 1–2% of bankroll. Your bankroll is money set aside only for betting, separate from rent and bills. A unit is the fixed amount you risk on a normal bet.
The small unit is there so you survive bad stretches. A 20-unit drop costs $400 at $20 a unit, which is 40% of the bankroll. At $10 a unit the same drop costs $200, or 20%. Drops of that size can happen to a bettor with a real edge over a long season.
Recalculate the unit on a schedule, such as monthly, not after every result. If the bankroll falls to $800, 2% becomes $16. Raising the unit to win losses back faster is chasing under another name.
How many bets before your record means anything?
Several hundred at least, and a thousand is better. A win rate measured over 100 bets has a standard error of about 5 percentage points. Over 1,000 bets that shrinks to about 1.6 points.
Compare that with the break-even rate of 52.38% at -110. A 55% record over 100 bets is only about half a standard error above break-even, well inside normal noise. The same 55% over 1,000 bets is about 1.7 standard errors above it, which starts to mean something.
Underdog bettors need even larger samples, because each result swings the record further. Closing line value helps here. It settles down in far fewer bets than profit does, which is why it is worth tracking from the start.
What is expected value, and how do you calculate it?
Expected value is the average amount a bet would win or lose if you could place it many times. Multiply each outcome by its probability, then add the results. A positive number means the price is in your favor; a negative one means it is not.
Take a $100 bet at +150 that you think wins 42% of the time. A win pays $150 and a loss costs $100. The expected value is 0.42 × $150 minus 0.58 × $100, which is $63 minus $58, or +$5 per bet. The break-even rate at +150 is 40%, so your estimate sits 2 points above it.
The answer is only as good as your probability estimate. If the true chance is 38%, the same bet is worth -$5. That is why the next step is checking your estimates against the closing line and recording how they hold up.