Debrief · Supporter Beta

22 Ranked Games Under the Microscope

Or, an introduction to GamingDiver Debrief as a performance improvement tool

I played my entire Ranked Season 101 (2v2 Tier VII) in one sitting. This meant 22 games in Plymouth, and was an opportunity to analyze and improve my season play using Debrief, GamingDiver's in-development replay analyzer. I queued solo all night: every teammate was a random ally, every game a new pairing.

Every number below was pulled from the replay files themselves: position tracks, salvo trades, HP timelines, smoke puffs, spotting intervals, kill feeds. No memory, no vibes. If you're a supporter, the games in this article are live links. Open one in Debrief and scrub the timeline yourself.

Plymouth, Tier VII Royal Navy light cruiser · open the build I ran

The season at a glance

The season did not start out smoothly

I started the evening on Radar Plymouth played like a proper radar cruiser: open water, light up targets. Five games in, the record was 2 wins, 3 losses, I was averaging 30,573 damage, and even the wins were quiet, my two lowest-impact wins of the whole night. The problem was simple: no one was taking out destroyers in a 2v2 format, so the Radar either wasn't getting used, or when it was used, was marginally useful.

After watching each replay after my match, it was clear what was wrong. It didn't make sense to pass up on Smoke Generator when partnered with another player for spotting in a game mode with scarce DD presence. So I switched up to a Smoke build. The first games on the new build were still rough while the old open-water habits wore off (the replays catch a lone desperate puff at 3:32 of the worst game of the night, sixteen seconds before I sank). Then it clicked. From 6:06 PM onward the smoke goes down proactively, around the three-minute mark, at the first cap fight, game after game; Debrief shows the puff timings in 13 of the final 17 games.

The rest of the season: 15W – 2L, averaging 68,615 damage. More than double the radar hour. Same ship, same format, same evening. The build didn't change what Plymouth is; it changed when she gets to start playing.

Session arc, in order: W W L L L → W W W W W W W W W L W W W W L W W

Insight 1: In 2v2, your ship is the win condition

The single strongest correlation in the whole dataset:

• Every game my ship survived, we won. All sixteen. 16W – 0L.
• When my ship sank: 1W – 5L.

In a two-ship format there is no team to carry your corpse. Every plan that ends with "and I probably die doing it" is a coin flip at best.

Insight 2: The closer games

My random teammate sank before the end in 9 of my 17 wins. That's not a typo: more than half my wins were finished alone. Three were true 1v2s; five more were duels after trading one-for-one. Debrief timestamps every beat, so here are the three best, exactly as the replays tell them:

Mountain Range, 116,922 damage is the season highlight, and the game the video below comes from. It shows how Plymouth closes out a game after the teammate goes down: torpedo hit on the Congress at 4:16, sank it at 5:02, lost my teammate's Jean Bart at 5:43, then the endgame vs a BA Tirpitz: smoke the reset @7:21, sonar lights the push @7:38, torpedo the chase @8:52. The spotting track in the viewer shows the Tirpitz vanish when my smoke goes down, then flick back on as he pushes into sonar range, seventeen seconds later. He never got back out.

8:42. Plymouth sits inside her own smoke next to the burning Tirpitz at B cap; the Congress and Jean Bart wrecks mark the earlier trades. Ten seconds later the torpedo lands.

Islands of Ice, 95,181 damage is the pure 1v2. My teammate's Scharnhorst went down at 4:43 with both enemies alive. Torpedoes into the Nottingham at 5:18 and 5:32, sank it at 5:35, fifty-two seconds after going 1v2. Then a reset smoke @7:44 and a flooding campaign against the Atlântico: eight torpedo hits, twenty flood ticks, sunk at 9:36.

New Dawn, 101,047 damage is the other true 1v2. My teammate's Bismarck went down at 5:58; the Rodney that killed him lived exactly thirty-seven more seconds (sunk 6:35). Four torpedo hits kept the flooding on the second cruiser while the clock and the caps did the rest. Won on points.

Plymouth's whole kit (smoke, sonar, a fast-cycling battery, torpedoes) is built for exactly this script: erase their vision advantage, punish the push, flood the runner.

Insight 3: My losses were not my teammates' fault

This one stung. Averages across wins vs losses:

Read that last row again. My random allies did more damage in my losses than in my wins. The losses weren't "matchmaking diff". They were games where my guns went quiet: zero kills in all five losses, hit count collapsed to a third. When Plymouth lands 60+ shells a game we win; when she lands 29 we don't. The replay data is brutally unsentimental about this.

Insight 4: Where I actually die

Debrief tracks every ship's position continuously, so you can ask: where was everyone at the moment I sank? All six of my deaths share one shape:

• I'm in or on the edge of the cap ring,
• both enemies are inside brawling distance,
• and my teammate is roughly twice as far from me as the nearest enemy.

That's the pattern: over-holding a contested cap in a smoke cruiser while the crossfire closes. Smoke makes the ring feel safe, but smoke doesn't stop two ships pushing the puff with sonar up.

The worst game of the season (Mountain Range, 17:11) is the teardown case: an open-water crossing into a Constellation + Congress sniper comp, no island line, and the replay shows my smoke going down at 3:32, sixteen seconds before I sank. 10,472 damage, my only game under 22k all night. Two starts later, the proactive-smoke pattern shows up and the season turned.

Insight 5: The comps that fight back

• With a battleship teammate: 14W – 3L. With a cruiser teammate: 3W – 2L. Plymouth wants a big target to shoot around.
• Against BB+CA enemy comps: 14W – 3L. Against double-cruiser: 3W – 1L, and 0W – 1L against the one CA+DD comp. Fast comps that can hunt the smoke are the uncomfortable games.

The full season log

Every game, in order. Open any linked game in Debrief.

† short recording; result confirmed, but the replay captures only the opening minutes, so it isn't featured above.

More of the Debrief toolkit

The tactical replay is the headline, but the panels under it did just as much work this season:

Damage forensics infer each player's commander build (fire-tick sizes encode Fight Fire with Fire rank, heal totals count Repair Party uses, salvo cadence falls out of the hit log). Next to it: your top score contributors and a kill feed where clicking a sinking jumps the playback to that moment.

The commander-build panel is how I knew the Mountain Range Tirpitz was running a full survival spec (Fight Fire with Fire R2, three Repair Parties, 36.5k healed) before I committed to the endgame: he was going to out-sustain a gun duel, so the finish had to be a torpedo.

Team rosters show every ship's consumable loadout as chips (this is where the radar-vs-smoke question stares back at you) plus a Base XP estimate per player. Below, damage-dealt-to bars are scorecard-verified where screenshots were attached; the striped portions are flagged as estimates rather than passed off as fact.

That honesty matters: when Debrief can't determine something, it says so on the card instead of guessing quietly. The Base XP number itself comes from our fitted formula (R² ≈ 0.87 on the scorecard corpus), broken down term by term in the viewer so you can see exactly where your 3,237 came from.

This is what Debrief is for

The Debrief viewer for the Mountain Range game: tactical replay on the left, match briefing and the Base XP formula breakdown on the right.

Everything above came out of the replay files: continuous position tracks for every ship, salvo-by-salvo damage trades, HP timelines, smoke activations, spotting intervals (that's how we timestamped "sonar lights the push @7:38"), cap-ring geometry, ribbon-level hit breakdowns, kill attribution, and, when the result screen is missing, outcome reconstruction from the kill feed and score ticker.