Fifty-five wins and forty-five losses look like a winning record on its own, but that number means nothing until it is priced against the odds those bets were placed at. A 55% win rate at -110 barely clears the break-even line, while the same record at plus money can be a real edge.
This calculator turns a bankroll, a unit size, a record, and an average price into the numbers that actually describe a betting result: dollar value of one unit, units won, profit at one unit a bet, ROI on money risked, and the win rate needed just to break even at those odds.
Start with bankroll, the total amount set aside for betting. Next comes unit size as a percentage of that bankroll, the fixed slice risked on a single bet. After that, enter wins and losses for the record being measured, followed by the average odds those bets were placed at.
The result shows what one unit is worth in dollars, the record stated as a win percentage, units won across the sample, what that record is worth in profit if every bet had been sized at one unit, ROI on money risked, and the break-even win rate required at the average odds entered. Read units won and ROI together, since either one alone hides part of the picture.
Unit size is just bankroll times the unit percentage. A $1,000 bankroll at 2% makes one unit $20.00, the amount risked per bet regardless of how the record swings.
Units won equals wins times (decimal odds minus 1) minus losses. At -110, a win pays back a little under one unit of profit, so 55 wins produce just under 50 units before the 45 losses are taken off, leaving units won at +5.00.
From there, profit at one unit a bet is units won times the dollar value of a unit: 5.00 times $20.00 is $100.00. ROI is units won divided by bets made, 5.00 divided by 100, or 5.00%. Break-even win rate is simply the implied probability of the average odds, 52.4% at -110, the number a 55.0% record has to clear to be worth anything at all.
Unit sizing exists so a five-game losing streak and a five-game winning streak get measured the same way, in units, instead of in whatever dollar amounts felt right at the time. A bettor who sizes bets by feel, larger after a win, smaller after a loss, ends up with a record that can’t be compared to anyone else’s or even to their own past results.
The common mistake is judging a record by win percentage alone. A 55% win rate sounds strong until it’s checked against a 52.4% break-even line at -110, where it barely survives, or against a lower break-even line at plus odds, where the same 55% is a clear edge.
A sharp bettor holds unit size constant, tracks ROI over a large enough sample to mean something, and treats any record as incomplete until it’s been run against the odds it was actually priced at.
What counts as a good ROI? There’s no single number that applies everywhere, since it depends on the odds being bet and the size of the sample. A 5.00% ROI on 100 bets at -110, the result of the worked example, reflects a record that clears break-even by a small but real margin.
Why can a winning record still lose money? Because win percentage says nothing about price. A bettor going 55-45 at heavy favorite odds well below -110 can show units won that are negative even with more wins than losses, since each win pays less than each loss costs.
How many bets does it take before ROI means anything? Small samples swing hard in both directions, so a handful of bets at any ROI figure doesn’t say much. The break-even win rate stays fixed at a given price, but the record needed to prove real skill against it takes volume to establish.
Does unit size have to stay fixed forever? It should stay fixed for the length of any sample being measured, since changing it mid-stream breaks the comparison. Bankroll can be reassessed and the percentage adjusted going forward, but a single record should be sized consistently from bet one to the last bet counted.