Level Gaps in Doubles: How 0.5 Changes a 2v2
Quick Answer: Half a level between two pairs is worth roughly a 70/30 match — the favorites should win, and losing one night in three is normal. One player being half a level better moves the pair's average only a quarter, closer to 60/40. But averages flatter uneven pairs, because in 2v2 the opponents get to choose who plays the ball.
Someone in your crew is a 3.0 and someone is a 3.5. On paper that is a rounding error. On court it decides who serves first, who gets lobbed all night, and whether anyone enjoys the third set.
So here is what a half-level gap is actually worth, with the arithmetic shown, plus the part the arithmetic gets wrong. If you want the scales themselves first — what the numbers are and where they come from — start with the guide to padel levels and ratings.
What half a level is worth in win probability
Levels are not designed to be turned into percentages. The 0-7 scale that booking platforms use — Playtomic describes its own as running from 0 to 7, with 7 at the top — is a description of a player, not a unit of measurement.
But you can build a working translation, and it is worth building, because "3.0 vs 3.5" tells you nothing until it becomes a percentage.
Every Elo-style rating turns a rating gap into a win probability with the same curve: a gap of 400 points means the favorite wins about 91% of the time, and the curve is smooth all the way down to a coin flip at zero. So the only choice to make is how many rating points one level is worth. Set one level at 300 points and half a level lands on a 70/30 match — which is about how a half-level mismatch feels: the better pair should win, and does, most of the time.
That gives a table. It applies to the gap between the two pairs, using each pair's average level — and the padel win probability calculator will run the same conversion for any two teams, so you can check your own crew's numbers rather than reading off the nearest row:
| Gap between pair averages | Better pair wins about | What that feels like |
|---|---|---|
| 0.0 | 50% | A real match. Nobody knows. |
| 0.25 | 61% | Close. The favorite feels it in the tight games. |
| 0.5 | 70% | Comfortable but not safe. One loss in three. |
| 0.75 | 78% | The result is expected. The margin is the interest. |
| 1.0 | 85% | One night in six or seven goes the other way. |
| 1.5 | 93% | A practice session with a scoreboard attached. |
Two things to take from it.
First, 0.5 is not a wall. A 70/30 pair loses roughly one match in three. When that happens nobody has been robbed and nobody has secretly improved overnight — 30% is just what 30% looks like from the inside.
Second, a level gap on one player is worth half of what you think. If three of you are 3.0 and one is 3.5, that player's pair averages 3.25. The gap between the pairs is 0.25, not 0.5, so you are looking at 61/39 — genuinely competitive. This is the single most useful thing on this page: to work out a 2v2, average each side first, then compare the averages. You can watch the same math run on real numbers in the padel Elo calculator, which averages each pair exactly this way before it prices the result.
And a caveat that matters: this is arithmetic from a model, not measured padel data. There is no official conversion between a 0-7 level and an Elo scale, and the 300-points-per-level choice is mine. Treat the percentages as the shape of the thing.
Why the average flatters an uneven pair
Now the part the table gets wrong.
Averaging assumes both players will be involved about equally. Padel is the sport where that assumption is least true. The other pair can see which of you is weaker inside two games, and after that they get to choose who plays the ball: serve to them, lob over them, hit the diagonal into their backhand, and never once hit to the player they are afraid of.
So an uneven pair does not play at its average. It plays somewhere between its average and its weaker player, and the gap between those two numbers is exactly how much the opposition can exploit. A 3.0 with a 4.0 is not a 3.5 pair. It is a pair with a 4.0 who is standing still and a 3.0 who is playing every ball of the match.
This has three practical consequences.
- The weaker player sets the ceiling, not the average. They are going to hit the most balls, so their bad shot is the pair's bad shot.
- The stronger player's job changes. Covering more court is worth less than making the weaker player's job smaller: take the middle, call loudly, and stop trying to win points from positions where nobody wins points.
- The rating will read the match through the average anyway. Any system that averages a pair — ours included — prices the result against a 3.5 pair that did not exist on court. That is one honest reason ratings sometimes feel unfair, and it is a reason to care about margins rather than one result.
The tactic nobody admits to, and its price
Playing to the weaker opponent is not cheating. It is the most basic tactic in doubles, every coach teaches it, and at any level of competition it would be strange not to.
Among friends it costs something anyway. The stronger opponent stands there for 40 minutes; the weaker one plays a match they cannot win and finishes the night convinced they held everyone back. Do it for three weeks running and you have not built a competitive crew, you have built a crew where one person quietly stops replying to the message.
The usable compromise: decide who the format is protecting before the first serve. Two versions work.
- Play it straight for one match. Full targeting, honest score, everyone knows what it was. Then change pairs.
- Constrain the target. No lobbing the weaker player, or serve-and-return only to the strong side, or the rule that turns up in every good crew eventually: after a certain point in the game, you have to hit at least one ball to each opponent.
Constraints feel artificial for about ten minutes and then they just feel like the format. And the score still counts, which is the point — a match with a rule in it is a real match; a match where the strong pair goes easy without saying so is not.
Pairing four players when the levels do not match
Four players make exactly three pairings, and you should play all three.
Take a crew of 3.0, 3.0, 3.5, 4.0:
| Pairing | Averages | Model odds |
|---|---|---|
| 4.0 + 3.0 vs 3.5 + 3.0 | 3.5 vs 3.25 | 61 / 39 |
| 4.0 + 3.5 vs 3.0 + 3.0 | 3.75 vs 3.0 | 78 / 22 |
| 4.0 + 3.0 vs 3.0 + 3.5 | identical to the first | 61 / 39 |
The pattern is general: split the extremes. Putting your strongest and weakest player together is the closest thing to a fair match a lopsided four can produce, and stacking the top two produces the one match nobody remembers fondly.
Rotating through all three pairings takes about 90 minutes for three short sets, and it does something the single "fair" pairing cannot: it separates you from your partner. Play only one pairing all season and the numbers can never tell whether you are good or whether your partner is — which is the whole reason a crew's own rating is built on rotation.
What a gap does to the numbers afterwards
Ratings pay for surprise, so a level gap changes the price of every result before anyone serves.
Beat a pair you were 70% likely to beat and there is only 30% of surprise to pay for — the move is small. Beat a pair that was 85% likely to beat you and the same system hands you several times as much. Lose the match you were expected to win and you pay the reverse. This is why a strong player in a mixed crew can play well for a month and watch their number drift down: the wins were priced at almost nothing and the two upsets cost full freight.
It is also why margin matters more than the result in an uneven crew. A 6-4 loss as the underdog is real information — you took 40% of the games against a pair expected to take 70% of them — and a system that only counts wins throws it away. If you want to see how a scoreline changes the price of a night, run one of your own results, then the same result with a different margin, through the padel Elo calculator and watch the two numbers separate. And if you are choosing which number to argue about at all, win rate versus rating is the fight worth having first.
Where this leaves your crew
A half-level gap is not a problem to be solved. It is the reason the fixture is interesting: 70/30 is the sweet spot of competitive sport, and most crews are sitting on one without knowing it.
What a crew genuinely lacks is a record. Nobody remembers that the 3.0 took four games off the strongest pairing in April; everybody remembers the 6-0 in May. That asymmetry is what a rating fixes — one scale, every result on it, priced by who you played and by how much.
That is what Rivals does for a crew of four, automatically and privately: a margin-aware rating that stays inside your crew, plus the head-to-head record for every pair of players, and it will show you exactly why each number moved. Nothing to protect, nothing to duck. But you do not need an app to start using this page. Average each side, split the extremes, and play all three pairings.
FAQ
Is a 0.5 level difference in padel a big deal?
Between two pairs, yes — it is roughly a 70/30 match, so the better pair should win and will lose about one match in three. Between two individual players it is much smaller than it looks, because averaging the pair halves it: three 3.0s and one 3.5 produces a 61/39 match, which is genuinely competitive.
How do you work out the level difference in a 2v2?
Average each pair, then compare the two averages. A 4.0 with a 3.0 averages 3.5, the same as two 3.5s on paper. That is what most rating math does, including the calculator on this site — but remember that averaging assumes both players touch the ball equally, and in padel the opponents choose who does.
Should the strongest player partner the weakest?
For a competitive match with four friends of different levels, yes. Splitting the extremes puts the two pair averages closest together, which is the fairest match your four players can produce. Stacking your two best players produces the one match everyone remembers as boring.
Is it rude to target the weaker player in padel?
It is the standard tactic in doubles and nobody should apologize for it in a competitive match. Among friends, three weeks of relentless targeting is how you lose a fourth player. The workable middle is to agree the rules first: play one match straight, then add a constraint — no lobs over the weaker player, or one ball per opponent per game.
Why does my level drop when I play with a much weaker partner?
Because most systems price your result against your pair's average, not against what you personally did. Partnered with someone well below you, your pair is rated higher than it plays, so expected wins pay almost nothing and losses cost full price. Playing varied partners is what eventually separates your number from theirs.