A unit is the fixed slice of a bankroll that a bettor uses as their basic stake measurement. Instead of saying “I bet $80 on Arsenal to beat Real Madrid,” a bettor who has set their unit at $40 would say “I bet 2 units.” The dollar value behind a unit is personal and usually private; what matters publicly is the number of units risked and won, because that number scales to any bankroll size and lets results be compared across bettors, tipsters, and time periods without anyone needing to disclose how much money is actually involved.

The mechanics are simple. A bettor picks a bankroll — the pool of money set aside purely for betting, separate from rent and groceries — and decides what fraction of it counts as one unit, commonly somewhere between 0.5% and 3%. That figure becomes the baseline stake for an average-confidence bet. Higher-confidence plays might go to 2 or 3 units; lower-confidence plays might drop to 0.5 units. The unit doesn’t move bet-to-bet even though the bettor’s confidence does — it’s the ruler, not the thing being measured.

Units solve a real problem: dollar figures are noisy and easy to misread. A tipster who posts “+$1,200 this month” tells you nothing about whether they’re staking $10 a race or $500 a race, or whether that profit came from one lucky outlier bet. A record of “+38 units across 210 bets” tells you the return relative to what was actually risked, and it stays meaningful whether the reader has a $500 bankroll or a $50,000 one. This is also why serious bettors track performance in units rather than currency — it strips out the effect of increasing or decreasing bet size over time and shows the pure strike rate and pricing edge.

Example

Suppose a bettor sets aside a $5,000 bankroll for the year and fixes their unit at 1%, so 1 unit = $50. Over a week covering the cricket World Cup and the clay-court swing in tennis, they log four bets:

  1. India to beat Australia at 1.85 decimal odds, a strong opinion on team form: staked 3 units = $150. India wins. Profit = 3 × (1.85 − 1) = 2.55 units = $127.50.
  2. A tennis outsider, ranked 74, to beat a seeded player at 3.40, a speculative live-underdog angle: staked 1 unit = $50. The outsider loses. Loss = 1 unit = $50.
  3. Over 2.5 goals in a Champions League group match at 1.91, a moderate-confidence pricing view: staked 2 units = $100. The match finishes 3-1. Profit = 2 × (1.91 − 1) = 1.82 units = $91.
  4. A horse each-way at 6.00 win odds in a competitive handicap, a lower-confidence value play: staked 1 unit = $50 to win. The horse finishes fourth. Loss = 1 unit = $50.

Net for the week: +2.55 − 1 + 1.82 − 1 = +2.37 units, or +$118.50 on a $5,000 bankroll. Reported in units, this is a clean, comparable +2.37u week regardless of whether next month the bettor’s bankroll has grown to $8,000 and their unit has grown to $80 with it. The percentage return and the units-won figure stay legible; only the dollar total changes.

Key Points

  • Size the unit to the bankroll, not the bet: Pick a percentage — 1% is a common, conservative starting point — and let the dollar value of a unit rise or fall as the bankroll does. Chasing losses by suddenly staking 5 units after a bad week is the single most common way units get misused.
  • Match unit size to genuine confidence, not to the odds: A short-priced favorite at 1.30 isn’t automatically a bigger stake than a value price at 4.50; stake size should track how strongly the bettor believes their number beats the market’s, not how “safe” the bet feels.
  • Use units to judge a tipster or your own record honestly: A service or bettor claiming profit should be judged on units won per units staked (ROI), not on raw dollar profit, since raw profit can be inflated simply by betting bigger.
  • Keep the unit stable enough to compare weeks and months: Recalculating the unit after every single bet defeats the purpose; most serious trackers only rebase the unit periodically (monthly or quarterly) as the bankroll materially changes.
  • Never let “units” hide reckless staking: A bettor who calls a bet “5 units” because they’re chasing a loss is still making an oversized bet — relabeling a stake in units doesn’t make bad bankroll management prudent, it just makes it harder for outsiders to spot.
  • State the base unit when sharing records publicly: “+20 units” means very different things at 0.5% versus 5% of bankroll; disclosing the staking plan (even without dollar figures) is what makes a units-based record actually verifiable.