RoflLab | From replay to team feedback

How it works

Three steps between yesterday's scrim and the talk with the player.

Upload the replays

Drop the .rofl files on any screen. The import runs in the background, tags scrims and official matches, and recognizes the opponent from their players.
Confirm who played

Each player has their own accounts, smurfs included, looked up with Riot by Riot ID. Anything ambiguous goes to review; nothing counts until it's right.
Deliver the feedback

Role cards, matchups, objectives, and gold efficiency become priorities for each player, with the evidence and what to look for in the VOD. Export it as a PDF for the meeting.

Features

Every number compared with players in the same position.

Every example below comes from the fictional organization that ships with RoflLab: two splits of Vortex Esports against eight opponents.

Role card

Each player against every player seen in the same position, on your team and on your opponents. Only what holds up in both wins and losses shows up.

KazeTop

Strength:CS/min vs lane opponent
Strength:CS per minute
Weakness:Team damage share
Weakness:Unspent gold at the end

MiraiJungle

Strength:Kill participation
Strength:Vision per minute
Strength:Wards placed per minute
Weakness:Damage/min vs lane opponent

SolusMid

Strength:Damage/min vs lane opponent
Strength:Team damage share
Strength:Damage per gold
Weakness:Vision per minute

Lane matchup

What changes when the lane opponent picks a given champion.

Kaze vs Rumble, 23 matches

Win rate

26%

58%

Gold/min

380

481

CS/min

6.4

8.0

Deaths

7.0

3.9

vs Rumble other matches

Record against expectations

Winning 66% against a team you should beat 65% of the time is different from winning 42% against the best team in the league.

Opponent	W / L	Win rate	Expected
Iron Wolves	16 / 5	76%	71% +5
Black Lotus	27 / 12	69%	67% +2
Kraken Club	40 / 21	66%	65% +1
Titan Forge	25 / 14	64%	63% +1
Aurora Gaming	20 / 20	50%	51% -1
Red Comet	14 / 16	47%	48% -1
Solar Flare	19 / 26	42%	44% -2
Nova Tempest	9 / 22	29%	34% -5

Expected: the chance of winning given each opponent's estimated strength. The difference shows who the team beats more, or less, than it should.

Objectives in wins and losses

The team's share of the total for both sides. The telling cases are staying ahead while losing, or falling behind while winning.

Barons

90%

10%

Dragons

77%

26%

Turrets

75%

24%

Heralds

69%

27%

in wins in losses

Gold efficiency

How much of the gold turns into damage in each player's typical match, compared with the position.

Solus Mid 65%
Lumi Support 56%
Mirai Jungle 48%
Raven ADC 47%
Kaze Top 31%

Feedback ready for the meeting

Priorities for the team, for each lane, and for each player, with the evidence alongside and what to look for in the VOD. Player goals track progress. The PDF includes your notes.

Strength

Support vision

Lumi is above the support baseline in vision per minute and wards, in both wins and losses.

Priority

Unspent gold at the end

Kaze ends matches with more unspent gold than other top laners. Review the recall timings.

Goal

Raven: CS/min ≥ 10

9.6 on average PDF

Only what's more than luck

Numbers that own up to the sample size.

Ten matches don't say what a hundred do. Each rate comes with the range of what's plausible, every opponent's strength goes into the math, and the feedback leaves out what doesn't hold up. Fewer false alarms in the talk with the player.

Win chance, with a 95% interval

Vortex Esports, both seasons. The middle line marks 50%.

Against a typical opponent55% 41% to 69%
Matches up to 30 min56% 47% to 65%
With LeBlanc on the other side beyond chance22% over 23 matches

The first two cross 50%: too early to conclude. The third sits entirely below it, so it's a real pattern.

Built for the team's routine

Each organization in its own space, each person with their own role.

Isolated data

Each organization only sees its own data, and the database itself enforces that.
Roles

Owner, admin, analyst, and viewer. The coaching staff invites whoever they need by email.
Accounts and smurfs

Links keep their history: an account that changed hands keeps counting for the right player on each date.
Change history

Who changed what and when, from the roster to the replays. Files stay private and are never published.

Invite-only access

Got a link? That's how your account starts.

Sign-up is closed. Each team joins through an invite from the RoflLab admin and then invites its own staff. If you already have an account, just log in.

Log in