A New Way to Play Rocket League 2v2
What 840 games and 128,845 touches, ranked 2s from Plat to SSL and the pros’ own, reveal about 2v2 positioning. Where the second man stands, what wins games, and the plays almost nobody uses.
Freeze a Plat game and an SSL game at the same moment and you can’t tell which is which. So we measured where the second man stands at every rank, to find out whether the central second man of the Pacifist System helps you win, and whether anyone uses what it gives them.
when their teammate touches the ball,
at every rank from Plat to SSL
840 games, 128,845 live touches
from finding them
Every rank holds the same shape. Central about a third of the time, straight behind in one lane about one in five, both on the same side more than half the game, at every rank and for the pros.
Fifteen claims were written down before they were scored on the games kept back to test them. Fourteen held, and the one that failed is in full below.
Proof that a pass or a play causes the goal. The ball goes to the second man when the pass is on. It’s replay data, not an experiment.
The trailer and the ground pivot, too rare for a fair number. And whether a duo that plays the four plays all game wins more, because almost nobody does.
Where should the second man stand, and does anyone use it?
In 2s, one of you is on the ball. That’s the first man. The other is the second man, waiting for what comes next. In the Pacifist System the second man stays central, in the middle of the pitch.
From the middle every option is still open. Your teammate can drop it back to you, pass it across to you, or pull it back to you from the corner. Follow straight behind them in their lane instead, and the pass into the middle has nobody to go to.
So Study 008 asks two things. Does being central actually help you win? And are players using what it gives them?
840 games, every frame of every game.
Pacifist Labs pulled ranked 2v2 replays from Ballchasing with every frame. About 150 games at every rank, Platinum to Supersonic Legend, plus the pros’ own 2v2 games. That’s 840 games and 128,845 live touches.
| Set | Games | Played | What it answers |
|---|---|---|---|
| Ranked 2s | 746 · about 150 a rank | April to September 2026 | The shape, what wins, the central pass and the four plays |
| The pros’ own 2s | 94 · 13 pros | January to October 2026 | Whether the top plays a different shape, described, never claimed |
| Kept back | 169 of the 840 · 154 ranked | One game in five, by a fixed hash | The test for every claim |
| Our games | 10 · Dave and Preceps | GC1 to GC2, all ten won | What the Pacifist shape looks like, not part of the 840 |
At every touch in live play, the player who touched it is the first man and their teammate is the second man. The second man is central when they’re within 1,000 uu of the middle of the pitch, across it. That’s a band 2,000 uu wide down the middle, a quarter of the pitch’s 8,192 uu width. They’re in the first man’s lane when they’re behind the first man and within 700 uu of them across the pitch. (uu is the game’s own unit of distance. A car is about 118 long.)
A ball to the teammate means the second man gets the next touch, with the ball having travelled 1,000 uu or more and the two of them 1,000 or more apart. For the shape over the whole game we read every frame, not just the touches.
2v2 positioning is the same at every rank. Even the pros.
When your teammate touches the ball, where’s the second man? In the middle quarter of the pitch about a third of the time. At Platinum, at SSL and for the pros.
Second man central when their teammate touches it
Straight behind their teammate in the same lane, about one in five. Again, every rank. Over the whole game, both players are on the same side of the pitch more than half the time, at every rank. And they sit about the same distance apart.
| Rank | In their lane | Same side | One behind the other | Apart |
|---|---|---|---|---|
| Platinum | 22.8% | 54.9% | 21.5% | 2,923 uu |
| Diamond | 21.3% | 53.8% | 21.1% | 2,956 uu |
| Champion | 21.6% | 55.0% | 21.5% | 2,909 uu |
| Grand Champion | 22.3% | 56.0% | 22.8% | 2,810 uu |
| Supersonic Legend | 23.2% | 54.9% | 23.8% | 2,749 uu |
| The pros | 23.8% | 55.4% | 24.0% | 2,696 uu |
SSL teammates sit about one and a half car lengths closer together than Plats, 174 uu, and the pros about two, 227 uu. That’s really it.
Higher ranks aren’t playing a different shape. They make fewer mistakes in it.
The one big difference is the one Study 006 found, nobody back. Both players past the ball, nobody between it and your own goal. Platinum teams do it about a third more often than SSLs.
Nobody back, share of live play
So the shape is the same and the mistakes inside it get rarer as you climb. That’s what separates the ranks in these games, not where the second man stands.
Being central doesn’t win on its own. Getting the ball to each other does.
We put both teams in the same game side by side. The team whose second man was central more often won half the time. A coin flip. The team that got the ball to each other more won nearly six in ten.
Same game, the team that did it more won
And in the Pacifist System, that’s what being central is for. It’s not where you stand. It’s what standing there lets you do.
How well are you playing the system?
Pacifist Score reads your replays and tells you how well you’re playing the Pacifist System. It measures your Pacifist pivots too, and scores them on your account and your player card.
See Pacifist ScoreThe pass into the middle comes one time in ten. When it does, a goal follows a third of the time.
Here’s the moment it matters. Your teammate’s got the ball out wide, up near your opponent’s goal, and you’re in the middle. The ball comes back to you about one time in ten, at every rank.
The ball goes to the central second man
When the ball does come to you, a goal follows inside ten seconds 32% of the time. When it doesn’t, 11%.
A goal inside 10 s from the same moment
Not all of those were passed on purpose. Some just ended up there. But nine times in ten the ball never gets to you, and when it doesn’t, the other team gets the next touch 64.5% of the time.
Four plays being central opens up. Every one of them is rare.
That’s just one option. Being central opens up four plays I teach. The drop and the lateral turn up often enough to put a number on. The pivot played along the ground turns up 28 times, enough for a first look, and the trailer barely turns up at all. Each one was turned into rules on the replay frames from my own definition, and checked by eye on drawn examples.
- The drop pass, or drop fifty. You lose a fifty on purpose, or take a ball that’s coming at you, and deflect it softly back into the middle to your teammate.
- The lateral pass. You shoot the ball across, side to side or into the centre, and your teammate comes in and carries on with the play.
- The trailer. You drag the ball towards the middle at an angle, then peel off up the pitch to bump or demo the defenders, while the ball goes to your teammate trailing in the middle.
- The Pacifist pivot. You’re first man in your opponent’s corner, and you drag or pass the ball back towards the centre, where your teammate is waiting. A solid back pass, usually along the ground so you can recover, not a backboard pass that forces your teammate to dive.
Each play is measured against its chances, the same situation whoever touched the ball next, so the rate is fair. Each one happens fewer than five times in every hundred chances, at every rank, and for the pros.
| Rank | Drop | Lateral | Trailer | Pivot |
|---|---|---|---|---|
| Platinum | 2.8% | 4.1% | 0.1% | 0.2% |
| Diamond | 3.1% | 4.2% | 0.2% | 0.1% |
| Champion | 3.7% | 3.9% | 0.2% | 0.3% |
| Grand Champion | 3.0% | 4.0% | 0.0% | 0.1% |
| Supersonic Legend | 3.2% | 2.7% | 0.6% | 0.1% |
| The pros | 3.8% | 3.6% | 0.0% | 0.0% |
SSLs play the drop about as often as Plats and the lateral less. The trailer and the pivot are too rare to compare. But when the plays happen, a goal follows far more often than from the same chance without them.
A goal inside 10 s, with the play and without
And the drop, the lateral and the pivot don’t let in any more goals at the other end. More goals at one end, no more at the other.
| Play | Made | Goal for | Goal against |
|---|---|---|---|
| Drop pass | 575 · 3.2% of chances | 20.3% against 9.6% | 5.4% against 7.4% |
| Lateral pass | 349 · 3.8% | 39.8% against 9.5% | 4.6% against 6.2% |
| Pivot, along the ground | 28 · 0.2% | 21.4% against 10.6% | 1 of 28 against 6.1% |
| Trailer | 10 · 0.2% | 4 of 10 against 9.4% | 0 of 10, too few |
Air and wall passes don’t let in more goals. They cost commitment and boost.
Only one completed pass in five in ranked 2s goes along the ground. Half come off a wall and the rest go through the air. My worry with an air pass is that it uses boost and commits both of you high, so if anything misses you’re all in. So we wrote down how to check it, then ran it once on every completed pass.
What follows a completed pass, ranked
| What follows | Ground | Wall | Air |
|---|---|---|---|
| Goal against inside 10 s | 4.0% | 3.2% | 3.2% |
| Opponent touch in your third inside 10 s | 27.0% | 26.3% | 21.8% |
| Goal for inside 10 s | 17.3% | 16.5% | 15.5% |
| A fifty as it’s taken | 7.9% | 7.8% | 7.1% |
| Both players off the ground | 33.1% | 81.0% | 74.0% |
| Boost the receiver uses getting to it | 18.5 | 26.1 | 25.6 |
So I was half right. Air and wall passes don’t let in more goals. Compared within the same third of the pitch and the same rank, the odds of a goal against are 0.80 (0.61 to 1.05) for a wall pass and 0.81 (0.59 to 1.10) for an air pass, against a ground pass. What they do cost is commitment. Both of you leave the ground after three in four of them, against one in three along the ground, and the receiver uses about 7 more boost getting to it.
Close enough to be passed to. There’s no perfect distance.
Central doesn’t mean sitting back. When your teammate’s on the ball in the middle or up near your opponent’s goal, and you’re within about a fifth of the pitch behind them, your team gets the next touch nearly half the time. Nearly a third of the pitch back or more, and it drops to under four in ten.
Your next touch, by how far back they are
And there’s no perfect distance. In the games we explored, sitting 2,000 to 3,000 behind let in fewer goals than having the second man ahead of the ball, 4.2% against 6.3%. That was written down as a claim, and it failed on the games kept back, 5.5% against 5.9% with the second man ahead of the ball. Just don’t be too far away.
Following in their lane or sitting off to the side makes little difference on its own, with the second man 500 uu or more behind. Your team’s next touch was 41.2% against 41.6%, and a goal for 17.5% against 17.8%, on the games kept back. What the lane costs you is the pass into the middle.
Ten games played the Pacifist way. A different shape from every rank.
Here’s what it looks like. Ten games of mine with my duo partner Preceps, played with the Pacifist System, GC1 to GC2. We won all ten, 48 goals for and 10 against.
Second man central when their teammate touches it
| Measure | Ours | Our opponents | GC | SSL |
|---|---|---|---|---|
| Second man central, at a teammate’s touch | 70.2% | 30.6% | 31.4% | 31.0% |
| Central when the ball’s out wide in attack, at touches | 62.7% | 18.3% | 16.3% | 18.0% |
| Both on the same side of the pitch | 25.8% | 56.7% | 56.0% | 54.9% |
| Nobody back | 3.3% | 11.8% | 8.3% | 7.7% |
| Both near the ball at once, within 1,200 uu | 4.4% | 9.7% | 9.4% | 9.7% |
| Apart | 3,457 uu | 2,846 uu | 2,810 uu | 2,749 uu |
| Ball to the teammate, at touches | 6.7% | 10.3% | 9.3% | 8.3% |
| Drop passes, share of chances | 9.2% of 65 | 4.0% of 75 | 3.0% | 3.2% |
| Goals let in a game | 1.0 | 4.8 | 3.2 | 3.1 |
Every rank in this study has the second man central about a third of the time. Ours was central seven times in ten. Every rank has both players on the same side more than half the time. We were on the same side a quarter of the time. And nobody back, about one moment in thirty, where a GC team is one in twelve.
It’s a role-based game. One of us on the ball, the other is the option. If the pass is there, we open up. If not, we take it back. And we were both on the ball at the same time half as often as a GC team.
Here’s what surprised me. We didn’t pass to each other more. If anything, less. We hadn’t mastered the passing yet, and the trailer didn’t exist at the time. What changed was the shape, and we let in one goal a game, where a GC team lets in three.
A replay shows what happened, not why.
Every result here comes from replays, not an experiment. We compared teams inside the same game, measured each play against the same chance without it, and checked the claims on games kept back from finding them. That rules a lot out. It doesn’t prove that a position or a play causes the goal.
Some of it runs both ways. A pass is played when it’s on, so the moment was already a good one. And the team playing better on the day passes more, scores more and wins more, which is why passing more and winning go together without one proving the other.
The sample has limits too. Ballchasing uploads lean towards keen players. Rank is the replay’s band on Ballchasing. The pros’ games are 94, enough to describe and not enough to claim. And the plays are rules on replay frames, checked by eye on a handful of examples each, so some real plays will be missed and some near misses counted.
Stay central. Look for them. Learn the four plays.
The rest of this section is my coaching, built on what held in the data. It isn’t data itself.
- Stay central, and close enough to be passed to. When your teammate’s on the ball, be in the middle of the pitch, within about a fifth of the pitch behind them. Not in their lane, and not sat deep. If you’re not sure how far, use string theory.
- Out wide near their goal? Look for them. When you’ve got the ball out wide in attack and your teammate’s in the middle, the pass is the option almost nobody takes.
- Learn the four plays. The drop, the lateral, the trailer and the Pacifist pivot. Not to force them, but so you’ve always got another option.
- Keep it on the ground when you can. A ground pass lets the passer recover and keeps both of you on your wheels. An air or wall pass isn’t more likely to cost a goal, but it takes you both off the ground and costs the receiver boost.
The Pacifist System is a circular system. Instead of rotating out while your teammate dives in, you rotate with the ball and feed it back to yourself, or back to your teammate if they’re there. Nothing’s forced. If you’ve got a perfect solo set-up, take it. You never have to force things, because you’ve always got another option.
| Question | What to look for |
|---|---|
| Where am I when they touch it? | In the middle of the pitch, not straight behind them in their lane. |
| Am I too far back? | Within about a fifth of the pitch behind them. Nearly a third back and your team keeps the ball less. |
| Can they pass it to me? | If the ball can’t reach you, you’re not an option, you’re a spectator. |
| I’m out wide in attack. Who’s central? | Look for them before you force a shot or a pinch. |
| Are we both past the ball? | Nobody back is the mistake that separates the ranks. Take it back. |
Fifteen frozen claims, fourteen held. Still observational.
| Measure | Result | Read it with this |
|---|---|---|
| Second man central | Plat 32.7% to SSL 31.0%, the pros 30.3% · kept back, every rank 30.1% to 33.5% | Held |
| In their lane | 21.3% to 23.8% at every rank and for the pros · kept back, Plat 22.0%, SSL 23.0% | Held |
| Same side, distance apart | 53.8% to 56.0% · 2,696 to 2,956 uu · kept back, SSL 9.3% closer than Plat | Held, the distance only just |
| Nobody back | Plat 10.5% to SSL 7.7% · kept back 10.5% against 7.8% | Held, matches Study 006 |
| More central, same game | Won 49.7% (46.2 to 53.2) · kept back 51.3% | Held, a coin flip |
| More balls to the teammate, same game | Won 57.6% (54.3 to 61.1) · kept back 61.7% | Held, the better team has the ball more anyway |
| The central pass when wide in attack | Made 7.5% to 13.4% · a goal 32.1% against 10.9% · kept back 31.8% against 9.8%, 44 moments | Held, not cause |
| How far back | Next touch 44.7% within 2,000 behind, 38.4% past 3,000 · kept back 44.2% against 37.6% | Held |
| In their lane against off to the side | Kept back, next touch 41.2% against 41.6% · a goal for 17.5% against 17.8% | Held, little difference |
| The drop and the lateral | Made 3.2% and 3.8% · a goal for 20.3% and 39.8% against about 10% · a goal against 5.4% and 4.6% against 6% to 7% | Held, not cause |
| The pivot along the ground | 28 in 746 games · a goal for 21.4% against 10.6% · against 1 of 28 | A finding, redefined after the test |
| The trailer | 10 possibles in 746 games | Too rare for a number |
| Playing more of the four plays | Won 53.6% · with passing held level, odds 1.08 (0.91 to 1.25) | Not written down first, a finding |
| Air and wall passes | Goal against 3.2% against 4.0% along the ground · both off the ground 74% to 81% against 33% | Written down first, exploratory |
| Our ten games | Central 70.2% · same side 25.8% · nobody back 3.3% · 1 goal let in a game | One duo, picked from a winning run |
And here’s the one that failed. It was written down before the games kept back to test it were opened, and it’s reported as it came out.
| Claim | What came out | Verdict |
|---|---|---|
| Sitting 2,000 to 3,000 uu behind the first man lets in at least 1.5 points fewer goals than the second man ahead of them | 5.5% (4.3 to 6.8) against 5.9% (5.1 to 6.9), explored 4.2% against 6.3% | Failed |
The strongest evidence is how consistent it is. The same shape at every rank, on the games we explored and the games kept back, and the plays followed by more goals at every step we checked. What it isn’t is proof of cause. And it isn’t peer reviewed.
The detail behind the numbers.
The sample
Ranked 2v2 replays from Ballchasing, read frame by frame by the Pacifist Labs scorer. 847 games came down with every frame, about 150 per rank band, Platinum to Supersonic Legend, from April to September 2026, and the pros’ own 2v2 games from 13 pros, January to October 2026. Five were a second upload of a match already in, found by score, length, touch count and goal times, and are left out. In two pro games the pro couldn’t be picked out of the lobby, so they’re out too. That leaves 840 games, 746 ranked and 94 of the pros’, and 128,845 live touches. In the pros’ games only the pro’s own team counts as the pros. Rank is the replay’s band on Ballchasing.
Touches, live play and the shape
Live play runs from the end of each kick-off to the next goal or kick-off. Every team is turned to attack the same way. At each touch the toucher is the first man and their teammate the second man. Central is within 1,000 uu of the middle across the pitch, in the lane is behind the first man and within 700 uu of them across it, goal side is between the ball and their own goal. The shape over the whole game uses every live frame, one value per team per game. Same side is both players on one side of the middle. One behind the other is both within 700 uu across. Nobody back is both players past the ball.
Balls to the teammate and the central pass
A ball to the teammate is the second man getting the next touch, the ball having travelled 1,000 uu or more and the two of them 1,000 or more apart. Out wide in attack is the first man’s touch in the attacking third, more than 2,600 uu from the middle, with the second man central. The middle and attacking thirds are everything outside the first man’s own third, which ends 1,700 uu back from halfway. A goal for or against counts if it’s scored inside 10 s of the moment.
The four plays
A chain is one player’s run of touches with nobody else touching in between. The drop is a fifty, an opponent touching within 0.35 s, or a ball coming at the first man, deflected 300 uu or more back to end nearer the middle, taken by the teammate within 2,000 of the middle and not deep in their own third. The lateral goes 1,500 uu or more across, more across than down, ends nearer the middle and is taken by the teammate coming forward. The trailer is a dribble that starts 1,500 or more from the middle and ends 1,000 or more nearer it, straight to the teammate, with the dribbler gone on up the pitch. The pivot is a spell on the ball with a touch in their corner or wing area, played back 500 uu or more to the teammate within 1,500 of the middle, with the ball no higher than 400, the passer no higher than 300 and the teammate no higher than 150. Each play is counted against its chances, the same situation whoever touched it next.
Ground, wall and air passes
A pass is a completed pass in the replay’s own events, the teammate touching it next. Ground uses the pivot’s height rule. Wall is otherwise any ball that comes within 300 uu of a wall, a corner or the ceiling on its way. Air is the rest. Pressure is an opponent touching the ball in the passing team’s own third inside 10 s. The odds within thirds and ranks are Mantel-Haenszel. These definitions were written down before the first run and not changed after it, but the study was already open, so every figure is exploratory.
Intervals
Intervals are 95% bootstrap intervals that resample whole games, so touches from the same game move together. The shape figures resample teams. For ground, wall and air passes the rates use Wilson intervals and the odds within thirds and ranks Mantel-Haenszel intervals. The win odds for the four plays come from one logistic regression, per standard gap in plays per hundred chances and in passing, each with the other held level.
The frozen claims
One game in five was held back by a fixed hash of its Ballchasing id before anything was opened. The claims were written from the other games, committed to the study’s records, then scored once on the games kept back, with whatever they said reported. Claims 9 to 15 were written after the games kept back had been opened for claims 1 to 8, but before any of their own measures were run on them. The times below are the commit times, UK time.
| Claims | Committed | Result committed | Held |
|---|---|---|---|
| 1 to 8, where the second man stands, what wins and how far back | 20:48 | 20:49 | 7 of 8 |
| 9 to 11, the team shape over every frame | 21:45 | 21:46 | 3 of 3 |
| 12 to 15, the four plays | 22:19 | 22:20 | 4 of 4 |
Five matches turned out to be in twice, and one had a copy on each
side of the split. Every check was run again without them, on 169
games kept back. Every verdict stayed the same and no estimate moved
by more than half a point, and the figures on this page are the rerun
ones. Each claims file was committed before its result, a minute
apart each time. Later work, the plays redefined in my words, the four plays and
winning, and ground against air passes, was done with both halves
open, so it’s reported as findings, never claims. The sha256
fingerprints of the claims file as each set was committed are
8de14af4a06b7b2af2075ffd990f8740512ff9e54a9e00d3e1238f11a0852585 for claims 1 to 8,
c4c1dafcd437ce778205194716aff3fc3ba392ff5021ee4410877ae6c4ea27b3 for 9 to 11, and
4a0258518047b7eb77426ef473cbb2e72829f95ebc007a1e4edacaa1321eff1e for 12 to 15.
Sources and remaining limits
Figures were checked against the Study 008 result files and the frozen claims. These identify the internal evidence. They are not public download links.
The study is observational. A position or a play that goes with scoring might also go with being the better team on the day, and the data can’t fully separate the two. No result here establishes a causal effect on the next touch, the next goal or the game.
What the findings show and what’s still open.
What’s the middle quarter of the pitch?
A band down the middle, 2,000 uu wide, a quarter of the pitch’s width. Not one of four boxes. If the second man is in that band when their teammate touches the ball, we count them as central.
Is 2v2 positioning really the same at SSL as at Plat?
Near enough. Central about a third of the time, in the lane about one in five, on the same side more than half the game, and only about one and a half car lengths closer together. The big difference is nobody back, 10.5% of the time at Plat against 7.7% at SSL.
Is following in my teammate’s lane bad?
Not on its own. With the second man well behind, the lane and off to the side keep the ball and score about as often. What it costs is the pass into the middle, because nobody’s there.
Are air passes too risky?
Not for goals against. Air and wall passes let in no more goals than ground passes, and an air pass is followed by less pressure in your own third. What they cost is commitment. Both of you leave the ground after three in four of them, and the receiver uses about 7 more boost getting to the ball.
Does this prove the Pacifist System wins?
No. Our ten games show the shape can be played and won with, but they’re one duo and picked by me. The study shows every rank plays the same shape, that passing goes with winning, and that the plays being central opens up are followed by far more goals and no more against.
Is the coaching proven by this study?
No. It’s my coaching, aimed at what the data found. It hasn’t been tested as an intervention.
Can I see how I play it?
Yes. Pacifist Score reads your replays and tells you how well you’re playing the Pacifist System. It measures your Pacifist pivots, and scores them on your account and your player card. Score comes with free access to the Dojo, a training programme built around your weaknesses and the time you’ve got to train.
Can they pass it to you?
Every rank plays the same shape. Being central doesn’t win on its own, getting the ball to each other does, and the plays being central opens up are followed by far more goals and no more against. Almost nobody plays them.
Stay central. Look for them. Learn the four plays.
See how well you play the system.
Pacifist Score reads your replays, tells you how well you’re playing the Pacifist System and scores your Pacifist pivots. The whole system is in the Digital Edition, built over more than five years. Or start free with the Pro Match mini report from one 2v2 replay. Study 008 informs that coaching. It is not a validation of every part of the system.
PACIFIST LABS Shaping the Future of Rocket League Decision-Making.