RallyIQ documentation

Everything RallyIQ does, across badminton, tennis, pickleball, padel, table tennis, and squash. Pick a section on the left, or read straight through.

RallyIQ helps tournament directors run competitive racquet-sport events: split a field into fair, rating-tight pools, schedule every round so that everyone plays the same number of games, record results as they happen, and rate every match on how players performed against expectation - with per-player analytics on top.

The things you can run

• Multi-round events - a field split into rating-tight pools, redrawn each round, with a live leaderboard.
• Social and competitive sessions - everyone plays the same number of games, partners rotate, and sitters umpire.
• Rated play - every match feeds the rating engine, so the field sorts itself as the event goes on.

Quick start

The fastest path to your first match:

1. From the dashboard, select New Tournament.
2. Name the event, choose singles or doubles, and add your players.
3. Generate the round - RallyIQ builds the pools, the matches, and the umpire and serve order for you.
4. On the day, open the event, tap a match, and enter the score. Standings recalculate instantly.
5. Share the live link so players and spectators can follow along.

Creating a tournament

Select New Tournament and set the basics:

• Sport - sets default scoring rules (win score, win-by-2, sideout vs rally).
• Match type - singles, doubles, or fixed-partner doubles.
• Players - add them by hand, import a roster, or sync from a rating provider.
• Format - how the draw is structured (see Tournament formats).

Match types & partners

Singles

One player per side - the simplest event to run and the cleanest to seed by rating.

Doubles

Two players per side. Partners can rotate each round (social) or stay fixed all event (competitive).

Fixed-partner pairs

Lock a pair together so they are always drawn and seeded as one unit through pools, playoffs, and medal rounds.

Tournament format

RallyIQ runs pool-based round-robin events. The field is split into pools of 4 to 6 players of similar standard, and inside each pool everyone plays everyone as far as the arithmetic allows.

Each round is drawn fresh from current ratings, so as results come in the pools re-form around how people are actually playing. See how pools are sized and the draw engine for exactly how the field is cut and scheduled.

Seeding, courts & scheduling

Seed by rating to keep the strongest players apart until late, then let RallyIQ assign courts and time slots automatically. The schedule balances court usage and rest between games, and reflows if you add or remove a player.

Recording scores

Tap a match, enter the game scores, and save. Win-by-2 and your chosen win score are enforced, so a tie can never slip through. Standings, court status, and the public link all update the moment you save.

The leaderboard

The leaderboard is ordered by rating, which already reflects how each player performed against expectation rather than just how many matches they won. A player who beat a stronger field ranks above one who beat a weaker field with the same win count.

Alongside the ordering, each player carries their matches played, wins and losses, and points won and conceded, so you can see what produced the rating rather than taking it on trust.

Editing & removing games

Made a typo? Reopen any match, correct the score, and standings recompute. Remove a no-show or add a game late and the schedule and rankings adjust without breaking the rest of the draw.

Analytics & insights

This is what separates RallyIQ from a draw-and-score tool. Every rated match feeds an analytics layer that turns the numbers into plain sentences a player can act on, rather than a table they have to interpret.

After each round, every player gets their own short read of how it went, drawn from their actual results:

• Where you beat expectation - the round in which you most outperformed what your rating predicted.
• Your toughest draw - the round with the strongest average opposition.
• Your biggest rating climb - which round moved your rating most, and by how much.
• The partner effect - how many rounds your pairing lifted your rating rather than held it back.
• Where you stand - your position in the field, and whether your schedule was harder than average.

These are computed directly from results and are deterministic: the same results always produce the same read, so two players comparing notes see consistent explanations rather than differently-worded guesses.

How pools are sized

Before a round can run, RallyIQ has to turn a list of players into groups. Two questions: how many groups, and who goes in each.

A pool is a small group of players who all play each other inside one round, usually on one court. Pools exist so everyone gets plenty of games instead of standing around, and so players meet others of a similar standard.

Every pool holds 4 to 6 players

Fewer than four cannot make a doubles rotation. More than six means too much sitting out and rounds that drag. Some turnouts simply cannot be split: 7 players has no legal arrangement (one pool of 7 is too big, and 4 + 3 breaks the minimum), and RallyIQ tells you rather than inventing something unfair.

The number you enter is a cap, not an order

Treat the pool count as "how many courts I have". It is an upper bound. If your number is impossible, RallyIQ uses the nearest legal count and hands you a short notice explaining exactly why - it never silently ignores you.

Four-player pools are avoided first

This is the part that surprises people. A four-player pool is exactly two teams, so it collapses the moment one person is late or leaves. RallyIQ therefore treats "how many 4-pools" as the first thing to minimise, ahead of keeping every court busy.

The practical consequence: RallyIQ may deliberately leave courts idle. With 36 players and 12 courts entered, it uses 7 pools (six of 5, one of 6) rather than nine pools of 4. If someone reports "the draw isn't using all my courts", this is the policy working as intended, not a fault.

Cut where the skill gaps are, not for equal sizes

RallyIQ ranks everyone strongest to weakest and draws the pool boundaries at the biggest jumps in standard, so each pool holds players of similar ability. Sizes are allowed to come out uneven on purpose.

Ten players rated 9.0, 8.9, 8.8, 8.7 then 5.0, 4.9, 4.8, 4.7, 4.6, 4.5 have an obvious cliff after the fourth. Splitting 5-and-5 would drag one strong player down among the club players and make both pools lopsided. RallyIQ splits 4 and 6, right on the cliff, and every match in both pools is close. Move the cliff to after the sixth player and the boundary moves with it, to 6 and 4.

A player sitting right on a boundary who is closer to the top of the pool below than the bottom of the pool above gets dropped down into the tighter fit. This is why two friends of different standards usually land in different pools - by design, so the games stay competitive.

The draw engine

Once pool membership is fixed, RallyIQ builds the actual matches: who plays, who sits, who umpires, and who serves first.

It optimises four things, in this order of priority:

1. Equal play - inside a pool, every player plays the same number of games. This is a hard rule, not a preference.
2. Partner variety - no two players ever partner each other twice in the same round.
3. Opponent spread - you face the other pool members as evenly as the arithmetic allows.
4. Rest spacing - your sit-outs are spread through the round, so you never rest twice back to back.

What each pool size actually plays

Why a six-player pool cannot be perfect

This is arithmetic, not a shortcoming. For everyone to partner everyone exactly once and play the same number of games, the sums only work out when the pool size leaves a remainder of 0 or 1 when divided by four. Four and five qualify. Six does not - it would need seven and a half matches.

So RallyIQ chooses equal play: a six-pool plays 6 matches and everyone plays exactly 4, rather than 7 matches where four players get five games and two get four. The price is that each player misses one possible partner, and faces three opponents twice and two opponents once. Both are mathematically forced. A six-pool showing a missed partner is optimal, not broken.

Umpires and first servers

Every player sitting out a match is assigned to umpire it, so the sitters are never idle. A four-player pool has no sitters and therefore no umpire. The first serve goes to whichever playing member has served least, decided in the original schedule order so that the rest-spacing reshuffle cannot skew it.

Draws are reproducible: the same roster always produces the same draw, because every tie is broken by a fixed rule rather than randomly.

Scoring formats

A scoring format is the recipe for how one game is won. RallyIQ treats it as a swappable setting rather than something baked into the engine.

Built in today: first to 11, first to 15, first to 17, and a 21-point rally format. New formats can be registered without touching the rating maths. Each format defines its target score, whether there is a hard ceiling, and how tight games typically are at that length.

One rating scale across every format

The obvious worry with mixed formats is that a longer game hands out bigger rating swings simply because there are more points in it. RallyIQ removes that: performance is measured as a fraction of the game's target score, not as a raw point count. Beating expectation by 10% of the target moves your rating the same amount whether the target was 11 or 17.

The one deliberate exception is the 21-point rally format, which is calibrated slightly differently because rally scoring produces tighter games.

The rating engine

RallyIQ has its own rating engine. The short version: your rating moves on how you played against expectation, not on whether you won.

How a rating moves

Before a match, RallyIQ averages each team's ratings and converts the gap between them into a predicted scoreline. Afterwards it compares what actually happened to that prediction. Beat the prediction and your rating rises; fall short and it falls.

Two consequences that surprise people, and both are intended:

• You can lose a match and gain rating - if you were expected to be beaten heavily and you kept it close.
• You can win a match and lose rating - if you were the heavy favourite and only scraped home.

Partners do not always move together

Most rating systems give both partners an identical change. RallyIQ adjusts each partner by how far their own rating sits from the team average. A strong player carried by a weaker partner gains less from a win and is marked down harder for a loss; the weaker partner gains more and is punished less. Credit follows what each player was expected to contribute.

New players settle quickly, established ones stay steady

Ratings move fastest when RallyIQ knows least about you, then settle smoothly as you log matches. There is no cliff edge where one extra match suddenly halves how much your rating can move. Players with few matches are flagged provisional, so a volatile early rating does not distort later draws.

The guard rails

• Scale - ratings run from 2.0 to 8.0. A player with no history starts at 4.0.
• Severe mismatches are skipped - if the two teams' average ratings differ by more than 1.5, the match is excluded from rating entirely. A hopeless mismatch says little about anyone's skill and would unfairly punish the weaker pair.
• No single game dominates - one match can move a rating by at most +0.40 or −0.25. The upward limit is deliberately larger than the downward one, so one bad night costs less than one good night earns.
• Corrections are clean - fix a score from an earlier round and RallyIQ replays the whole tournament from the starting ratings rather than patching the old numbers. The corrected history is exactly what it would have been had the score been right first time.

Awards & badges

At the end of an event, RallyIQ generates a podium plus achievement badges to hand out and share. Your leaderboard row shows up to three badges, with a +N chip if you hold more; hover it to see the rest.

There are two kinds, and they behave completely differently.

Milestone badges - earned by your own totals

The three rating badges stack: a player rated 6.8 with 12 matches holds all three at once.

Superlative badges - earned fresh every round

These are a snapshot of who is leading right now, not a permanent honour. Every time a round is processed, all superlative badges are stripped from everyone and re-awarded for that round alone. Losing one is expected behaviour, not a fault - it simply moved to whoever earned it this time.

Spreadsheet export

Export the whole event as a multi-sheet Excel workbook - registrations, rounds, matches, results and the leaderboard - for your own analysis or your club's records. Individual sheets can be exported on their own if you only need one.

Sharing a live link

Every event has one public live link. Share it once and players and spectators follow brackets, court status, and standings from their phones in real time - no account needed to watch.

Private player links & self-scoring

Give each player or team a private link so they can enter their own court's scores. RallyIQ keeps everything in sync, so you do not have to run between courts collecting results.

After the event

When the event finishes, the leaderboard, every round's results and each player's own analytics stay available at the share link, so players can look back at how their rating moved and which round was their toughest draw. The full workbook export is the archive copy.