Should pushes count as losses?

No. The standard convention excludes them from the win-rate denominator, because a push is a bet that never got decided. Counting them as non-wins would penalise records on markets where pushes are common. The push rate is shown separately.

Why is my win rate above 50% but my profit negative?

Because 50% is not break-even. At −110 you need 52.38% just to stand still. And if your losing bets carried larger stakes than your winning ones, the record and the profit can end up on opposite sides of zero regardless of the price.

Is ROI on handle or on bankroll?

Handle — total amount wagered. Bankroll-based return measures growth, which depends on how often you turn the bankroll over, so two bettors with identical skill can post very different numbers.

How big does a sample need to be?

Larger than a season. Distinguishing a real 2–3% edge from break-even usually takes well over a thousand bets. The sample size calculator gives the figure for your specific edge, and closing line value gives you a faster read on whether the process works.

Why do some figures show an em dash?

Because they are undefined rather than zero. With no decided bets there is no win rate; with no handle there is no ROI. Printing 0% would read as a result rather than as an absence.

Betting Record Calculator

Wins, losses, pushes and money in. This turns them into the figures that actually describe a record.

Wins

Losses

Pushes

Total wagered

Every dollar put through — your handle.

Net profit or loss

Negative for a losing period.

Unit value

Average odds

Optional — lets the record be compared against the win rate it needed.

Win rate

53.85%

112 from 208 decided bets. At 52.38% break-even, this record clears the bar by 1.47% — which is what the win rate actually has to be judged against.

Total bets: 215
Decided bets: 208
Push rate: 3.3%
ROI: +4.30%
Net profit: +$1,850.00
Average stake: $200.00
Profit per bet: +$8.60
Net units: +9.25u
Break-even win rate: 52.38%

Win rate excludes pushes from the denominator — the standard convention, since a push is not a bet you won or lost. Push rate is reported separately so nothing is hidden. Note that a win rate above break-even does not guarantee a profit: unequal stakes across bets can put the two figures on opposite sides of zero.

Runs entirely in your browser · nothing is sent or saved

How it works

Most bettors know their record and very few know what it means. This takes the raw counts and produces the derived figures — win rate, ROI, profit per bet, net units — along with the one comparison that makes any of them meaningful: the win rate the prices actually required.

Pushes are excluded from the win rate denominator. That is the standard convention and the right one: a push is a bet that never got decided, so counting it as a non-win would understate a record with lots of them. The push rate is reported separately so nothing is concealed by that choice.

ROI divides by handle — the total wagered — not by bankroll. That is what makes it comparable across bettors staking very different amounts, and it is why a 3% ROI means the same thing whether the bankroll behind it is $1,000 or $100,000.

The number to look at is the gap between your win rate and the break-even rate at your average price. A 54% record sounds strong until you notice the average price was −140, where break-even is 58.3%. Win rate in isolation says almost nothing; win rate against the price it needed says nearly everything.

Two figures are deliberately withheld rather than shown as zero. With no decided bets there is no win rate — a record of pure pushes has not established one. And with no handle there is no ROI. Printing 0% in either case would read as a result rather than an absence, which is worse than an em dash.

The formula

decided bets   = wins + losses
win rate       = wins / decided bets      (pushes excluded)
push rate      = pushes / total bets

ROI            = net P/L / total wagered
profit per bet = net P/L / total bets
average stake  = total wagered / total bets
net units      = net P/L / unit value

break-even at average price = 1 / average decimal

  win rate > break-even  →  the prices were beaten
  (though unequal stakes can still leave the
   record and the profit on opposite sides of zero)

Pushes leave the win-rate denominator but stay in the bet count. ROI's denominator is handle, never bankroll.

A worked example

112–96–7 over a season, $43,000 wagered, $1,850 profit, $200 units, average price −110.

Win rate is 112 / 208 = 53.85% — pushes excluded, so the denominator is 208 rather than 215. Break-even at −110 is 52.38%, so this record beat its prices by 1.47 points.

ROI is 1,850 / 43,000 = +4.30%, or $8.60 of profit per bet across 215 wagers. In units, that is +9.25u.

A good season, and worth being precise about how good: 4.3% ROI over 215 bets is well within the range that variance produces from a break-even bettor. The win rate confidence calculator puts a range around it, and the sample size calculator says how many more bets it would take to be sure.

Common questions

Should pushes count as losses?: No. The standard convention excludes them from the win-rate denominator, because a push is a bet that never got decided. Counting them as non-wins would penalise records on markets where pushes are common. The push rate is shown separately.
Why is my win rate above 50% but my profit negative?: Because 50% is not break-even. At −110 you need 52.38% just to stand still. And if your losing bets carried larger stakes than your winning ones, the record and the profit can end up on opposite sides of zero regardless of the price.
Is ROI on handle or on bankroll?: Handle — total amount wagered. Bankroll-based return measures growth, which depends on how often you turn the bankroll over, so two bettors with identical skill can post very different numbers.
How big does a sample need to be?: Larger than a season. Distinguishing a real 2–3% edge from break-even usually takes well over a thousand bets. The sample size calculator gives the figure for your specific edge, and closing line value gives you a faster read on whether the process works.
Why do some figures show an em dash?: Because they are undefined rather than zero. With no decided bets there is no win rate; with no handle there is no ROI. Printing 0% would read as a result rather than as an absence.

The guide behind this calculator

Related calculators

For informational and analytical purposes only. These tools do not predict outcomes and do not recommend wagers.