A 32–18 record is a 64.00% win rate, and 70% takes 10 straight wins
A win rate is the share of the games you played that you won. At 32 wins and 18 losses you won 32 of 50 games, so 32 ÷ 50 × 100 = 64.00%. For that record the calculator above shows that reaching 70% takes 10 wins in a row, and that if exactly 30 games remain you need 24 wins out of them (you can afford 6 losses).
The target needs more wins than most people expect because every win also adds a game to the denominator. The win-rate funnel draws two curves, one for winning every game from here and one for losing every game, so you can see how far the rate can still move and where it meets your target line. If ties are in the record, whether they count as games, count as nothing or count as half a win changes the rate, so the calculator puts all three side by side.
How to use it
- Pick how you will enter the record. If you know wins, losses and ties, use "Wins · losses · ties". If a game profile only shows a total game count and a win rate (%), use "Games · win rate".
- Enter the record. A blank ties box counts as 0. You can type 1,000 with a comma if you like.
- Choose a tie rule if there are ties. Once you enter ties, three cards appear (count as games, leave out, tie = ½ win), each showing the win rate under that rule. The card you press is the rule used for the target.
- Enter a target win rate and, optionally, games left. A target alone gives the streak you need. Add games left and you also get the fewest wins needed within them. You can leave games left empty.
- Move the target line on the funnel. Drag the handle at the right edge up or down, or press it and use the arrow keys to change it by 1%. The needed streak and the row of boxes update with it. Hover over the chart to see the range of win rates after that many more games.
- Take the result with you. "Copy result" copies the result as a sentence, and "Copy link" copies an address that holds your record and tie rule so the same calculation opens again.
How a win rate is calculated
Win rate (%) = wins ÷ games counted × 100. With no ties, the games counted are wins plus losses. This calculator rounds to two decimals. A baseball standings figure such as .640 is the same thing as 64.0%.
| Record | Calculation | Win rate |
|---|---|---|
| 5–1 | 5 ÷ 6 | 83.33% |
| 9–4 | 9 ÷ 13 | 69.23% |
| 6–4 | 6 ÷ 10 | 60.00% |
| 32–18 | 32 ÷ 50 | 64.00% |
| 9–7–1 (tie as ½, NFL style) | 9.5 ÷ 17 | 55.88% |
Win rate and win/loss ratio are different numbers. A 6–4 record is 60% as a win rate, while the win/loss ratio (W/L) is 6 ÷ 4 = 1.5. Some game stat pages show W/L on its own, so the calculator lists it in the lines under the funnel.
Ties: count them, leave them out, or call them half a win?
No single rule fits every sport, so follow the league or event you are comparing against. Here is a 10–5–5 record under all three rules. In the NFL a tie counts as half a win, so the calculator starts on that rule for this page. MLB and the NBA play games to a decision, so a tie never arises there, and KBO and NPB standings leave ties out of the rate.
| Rule | Formula | Win rate | Streak to 70% | Used by |
|---|---|---|---|---|
| Count as games | W ÷ (W + L + T) | 50.00% | 14 | When you want the plain share of games won |
| Leave out | W ÷ (W + L) | 66.67% | 2 | KBO and NPB baseball standings |
| Tie = ½ win | (W + T ÷ 2) ÷ (W + L + T) | 62.50% | 5 | NFL standings |
Tie games count as one-half win and one-half loss for both clubs.NFL Tiebreaking Procedures · checked 2026-10-22
The KBO league rules put it the other way: the win rate is wins ÷ (wins + losses), so a tie is simply not in the formula. Changing the rule moves not only the current rate but the streak you need. The same 10–5–5 record needs 14 straight wins to reach 70% if ties count as games, 2 if they are left out and 5 if each tie is half a win. Do not compare figures that use different rules, and say which rule you used when you share a record.
How many straight wins to reach a target?
Say you have A wins in B games and want a target rate t (0.7 for 70%). After x more wins in a row you need (A + x) ÷ (B + x) ≥ t. Rearranged, x ≥ (t × B − A) ÷ (1 − t), and that value rounded up to a whole number is the fewest straight wins. From 32–18 with a 70% target, (0.7 × 50 − 32) ÷ 0.3 = 10, and ten wins give 42 ÷ 60 = 70.00%.
The higher the target, the faster the streak grows, because the divisor 1 − t shrinks toward 0 as the target nears 100%.
| Target | Fewest straight wins | Record then | Rate then |
|---|---|---|---|
| 55% | 12 | 62–50 | 55.36% |
| 60% | 25 | 75–50 | 60.00% |
| 65% | 43 | 93–50 | 65.03% |
| 70% | 67 | 117–50 | 70.06% |
| 75% | 100 | 150–50 | 75.00% |
| 80% | 150 | 200–50 | 80.00% |
| 90% | 400 | 450–50 | 90.00% |
Even at the same 50%, the more games you have already played, the harder it is to move. Going from 50% to 60% takes a streak that grows in step with the games played.
| Record now | Games played | Fewest straight wins |
|---|---|---|
| 5–5 | 10 | 3 |
| 10–10 | 20 | 5 |
| 25–25 | 50 | 13 |
| 50–50 | 100 | 25 |
| 100–100 | 200 | 50 |
| 500–500 | 1,000 | 250 |
Once you have even one game that was not a win, a streak can never reach a 100% target. When ties count as games, a tie plays the same role as a loss here. In that case the calculator says 100% cannot be reached instead of giving a streak, and with no games recorded yet it shows that a first win makes it 100%.
Fewest wins needed in the games left
With a fixed number of games left, such as the rest of a season, the question changes. After playing all N remaining games you need (A + k) ÷ (B + N) ≥ t, where k is the number of wins among them. So k is t × (B + N) − A rounded up, and if that is below 0 you could lose every remaining game and still hold the target. The order of wins and losses does not change the final rate, so the k wins do not have to come in a row.
- 32–18, target 70%, 30 games left: 0.7 × 80 − 32 = 24, so finishing 24–6 gives 56 ÷ 80 = 70.00%. In the games left alone that is a win rate of 80.0%.
- 50–50, target 60%, 50 games left: 0.6 × 150 − 50 = 40, so you need 40 wins and 10 losses. With 39 wins it is 89 ÷ 150 = 59.33%, which falls short.
- 50–50, target 60%, 10 games left: even winning all ten gives 60 ÷ 110 = 54.55%, so the target is out of reach within those games. With no cap on games, the fewest straight wins is 25.
- KBO's KT, through September 27, 2026: 81 wins, 4 ties, 49 losses: in this Korean baseball example 10 of 144 games remain and the rate with ties left out is 81 ÷ 130 = 62.31%. To hold .600 for the season they need 3 wins in the last 10, which gives 84 ÷ 140 = 60.00%.
The dotted vertical line on the funnel marks how many games are left. Its top end is the final rate if you win them all, its bottom end the final rate if you lose them all, and the dot on the line is the rate with only the fewest wins. If the target line sits above that vertical line, the games left cannot get you there.
How many straight losses can you take?
If you are already at or above the target, the question flips: how many losses in a row before you drop below it? After y losses the rate A ÷ (B + y) has to stay at or above t, so y ≤ (A − t × B) ÷ t, and that value rounded down is the number of straight losses the target survives. From 32–18 with a 60% target, (32 − 30) ÷ 0.6 = 3.33. Three straight losses leave 32 ÷ 53 = 60.38%, still above the target, and a fourth drops it to 32 ÷ 54 = 59.26%. For a record already past the target, the calculator shows this loss count in place of a streak.
How much does one game move the rate?
After n games at a win rate p, one more win lifts the rate by (1 − p) ÷ (n + 1) and one more loss lowers it by p ÷ (n + 1). That is why a single game weighs less the more you have played, and the "Next game" line in the calculator shows this.
| Record now | If you win | If you lose |
|---|---|---|
| 5–5 (10 games) | 54.55% | 45.45% |
| 10–10 (20 games) | 52.38% | 47.62% |
| 25–25 (50 games) | 50.98% | 49.02% |
| 50–50 (100 games) | 50.50% | 49.50% |
| 250–250 (500 games) | 50.10% | 49.90% |
| 500–500 (1,000 games) | 50.05% | 49.95% |
When you talk about a change, keep percentage points and percent apart. Going from 50% to 60% is a rise of 10 percentage points, but relative to the starting value it is a 20% rise.
How far can you trust a win rate from few games?
A win rate summarizes the past and is not the probability of winning the next game. If you assume every game is won with the same chance, though, the record lets you estimate the range that chance probably lies in. The "95% range of true win rate" in the calculator uses the Wilson interval (Wilson, 1927). The NIST statistics handbook presents the Wilson method for a proportion's confidence interval, and Brown, Cai and DasGupta (2001) recommend it over the simple approximation that is often used when the number of games is small.
| Record | Games played | 95% range |
|---|---|---|
| 6–4 | 10 | 31.3–83.2% |
| 12–8 | 20 | 38.7–78.1% |
| 30–20 | 50 | 46.2–72.4% |
| 60–40 | 100 | 50.2–69.1% |
| 180–120 | 300 | 54.4–65.4% |
| 600–400 | 1,000 | 56.9–63.0% |
Six wins in ten games cannot tell you the true level is above 50%, and the low end of the range only climbs above 50% at around 100 games. Real games differ in opponent and conditions, so the assumption does not hold exactly. Read this range as a feel for how much a rate swings on a small number of games. It is the thick translucent bar over the current dot on the funnel.
When a game screen shows only games and win rate
Many game profiles show a total game count and a win rate (%) instead of wins and losses. Enter them in "Games · win rate" and the calculator multiplies games by the rate, rounds to get the wins, and then works the same way. At 250 games and 52.4% that is 131 wins and 119 losses (52.40%), and 60% takes 48 straight wins.
The rate on screen is already rounded, so it may not match the game count exactly. At 7 games and 50% you would need 3.5 wins, which cannot be a whole number. The calculator then uses the rounded count (4 wins, 57.14%) and says so in the status line. If you can find the exact wins and losses, entering them in "Wins · losses · ties" is more accurate.
Common mistakes
- Averaging win rates to combine them: the average of 1–0 (100%) and 1–9 (10%) is 55%, which is wrong. The combined record is 2–9, a win rate of 18.18%, so add wins and losses first, then divide.
- Thinking 10 percentage points means winning 10% of your games played: from 50–50, ten more wins only reach 54.55%. Getting to 60% takes 25 in a row.
- Mixing set wins with game wins: in a best-of-three, where one match has several sets, the set win rate and the match win rate differ. Count in one unit only.
- Comparing win rates that use different tie rules: 66.67% with ties left out and 50.00% with ties counted both come from the same 10–5–5 record.
- Reading a displayed 100.00% as a true 100%: 999,999–1 is 99.9999%, which only shows as 100.00% after rounding to two decimals. The calculator decides whether the target is met from the unrounded value.
Limits
- It assumes no ties happen in the games still to come. If one does, fix the record and calculate again.
- The streaks, loss counts and wins needed are arithmetic conditions. They do not predict your chances of winning or your standing. For clinching a title, the baseball magic number calculator handles that separately.
- The 95% range assumes the same chance of winning every game and no effect of one game on the next. In games where matchmaking or opponent strength changes, or sports with home and away games, the real uncertainty can be larger.
- Each box accepts up to 999,999 games. With a very high target the streak you need can be larger, and when it goes past the chart's range it is shown with an arrow and a number only.
- Your record is calculated in your browser only and is never sent to a server or stored. "Copy link" puts the record in the address, so keep that in mind when you share it.
Reference date: October 22, 2026 · The KBO league rules are the 2026 edition, and the NFL tiebreaking rules and NIST handbook were checked in their original text the same day. The KBO KT example is the standings through games on September 27, 2026 from the KBO record room.
Sources
- 01KBO — 2026 League Rules (PDF, in Korean)Article 2, win rate calculation: win rate is wins / (wins + losses)
- 02NFL — Tiebreaking ProceduresTie games count as one-half win and one-half loss for both clubs
- 03NIST/SEMATECH e-Handbook of Statistical Methods — 7.2.4.1 Confidence intervalsConfidence interval for a proportion: the method first published by Wilson (1927) and the recommendation of Brown, Cai and DasGupta (2001)
- 04Wilson, E. B. (1927). Probable Inference, the Law of Succession, and Statistical InferenceJournal of the American Statistical Association 22(158), 209–212
- 05Brown, L. D., Cai, T. T., DasGupta, A. (2001). Interval Estimation for a Binomial ProportionStatistical Science 16(2), 101–133
- 06KBO Record Room — Daily Team Standings (in Korean)Through games on September 27, 2026: KT 81 wins, 4 ties, 49 losses (win rate with ties left out .623)
- 07How NPB standings are decided (Japanese) — マジック計算室NPB regular-season standings use win % = W ÷ (W + L); ties are left out of the rate
Frequently asked questions
What are the win rates for 5–1 and 9–4?
Win rate is wins ÷ (wins + losses) × 100. 5–1 is 5 ÷ 6 × 100 = 83.33%, and 9–4 is 9 ÷ 13 × 100 = 69.23%. With no ties, the tie rule you pick makes no difference to the value.
Why does 60% from 50–50 take 25 straight wins and not 10?
Each win also adds a game to the total. Ten more wins give 60–50, and 60 ÷ 110 = 54.55%. You need 75–50 for 75 ÷ 125 = 60.00%, while 74–50 is 74 ÷ 124 = 59.68%, which is short.
Is a 10–5–5 record 50% or 66.67%?
It depends on how ties are counted. Counting ties as games gives 10 ÷ 20 = 50.00%, leaving them out gives 10 ÷ 15 = 66.67%, and a tie as half a win gives 12.5 ÷ 20 = 62.50%. The KBO league rules use wins ÷ (wins + losses), so ties are left out, while the NFL counts a tie as half a win and half a loss.
If I need 24 wins in 30 games left, must they come in a row?
No. The final rate depends only on how many games you win, not on their order. From 32–18 with a 70% target and 30 games left, finishing 24–6 gives 56 ÷ 80 = 70.00%. If a tie comes up in those games, fix the record and calculate again.
How do I calculate when a game screen only shows games and win %?
Switch the entry mode to "Games · win rate" and enter both values, and the calculator works back to the wins. At 250 games and 52.4%, 250 × 0.524 gives 131 wins and 119 losses, and 60% takes 48 straight wins. If the on-screen rate is rounded and the wins do not come out whole, it rounds to the nearest whole win and tells you.
A 6–4 and a 60–40 record are both 60%. What differs?
How tightly the record pins down true ability. Treating every game as the same chance and using the Wilson interval, 6–4 gives a 95% range of 31.3–83.2%, and 60–40 gives 50.2–69.1%. One game also moves the rate less the more games you have played.
Can I average win rates when combining seasons or characters?
No. Add the wins and losses separately, then divide again. The average of 1–0 (100%) and 1–9 (10%) is 55%, but the combined record 2–9 is 2 ÷ 11 = 18.18%. Averaging rates over different game counts gives the smaller sample too much weight.
With a loss on record, can a win streak reach 100%?
No. From 1–1, however many you win, the game count stays one above the wins, so you approach 100% without reaching it. 999,999–1 is 99.9999% and shows as 100.00% at two decimals, but the calculator judges by the unrounded value and does not count a 100% target as met.