Game Ops

The projection is the start. Game Ops turns it into a plan, an adjustment, and a lesson.

Simulation shows what happens and why; Game Ops does something with it, a scouting plan before, live counters during, and an audit after that sharpens the next plan. And it remembers more than the roster: the coach across programs, the crew profiled like a coach.

Case 01 · the plan, before the game
Phase 01 · Pre-game

The trace becomes a plan you can actually run.

A scout is more than a plan. It is what the opponent runs and how often, the shots they live on, the film to watch, and only then the calls. Game Ops assembles all of it off the simulated matchup. Here is the scout on Duke.

What they run · top actions by frequency and value
Hub elbow post-up / kick21% freq+1.14
Drag screen into pick-and-pop17% freq+1.06
Movement-shooter pindown14% freq+1.09
Freshman-guard ball screen12% freq+0.88
Where they score · shot profile
RIM 44%
3PT 38%
MID 18%
Film KVision has already built the clip list: every rep of the top four actions, cut and tagged, ready before the staff opens the scout.
The plan · what to do about it, in priority order
Attack the pick-and-pop seamDuke's drop-oriented big is pulled out by a stretch five; run it early and often, it is the live seam the sim flagged+1.6 EPP
Hunt the interior anchor's foulsforce Ngongba II into rotations at the rim; his fouls are the fastest way to move Duke off its defensive identityfoul target
Pressure the freshman guardsball pressure on the young creators is where Duke's floor dips, per the trace+2.3 edge
Do not switch the hubswitching hands Boozer the mismatch he wants; stay home and live with the tough shotavoid
Shot diet to forcepush Duke into contested mid-range off the movement shooter, deny the corner threes that bend the help-3% eFG

Every call traces back to a specific clash in the simulation, with the projected value attached. This is the scout the sim produces: not "you win 29%," but the four things to do about it, in priority order.

Real Duke engine read turned into an illustrative plan on the real Game Ops scout structure. Demonstration figures.

Case 02 · the adjustment, during the game
Phase 02 · Live

The plan updates on the fly, powered by a live simulation.

KVision tags actions and coverages live and tracks matchups, so the plan is not frozen at tip-off. When a trigger fires, Game Ops surfaces the counter, and prices it against a live re-simulation.

2Q 6:14
Ngongba II picks up his 3rd foul.
Hunt his fouls: run the pick-and-pop at him now, before the sub.
win-prob swing +4.1% if he sits
2Q 2:40
Duke has run the same elbow action 4 times for +1.5 EPP.
Counter: switch the entry and top-lock the shooter off it.
recommend counter · blunts the action
2Q 1:05
Live shot-diet read: Duke is at 41% of shots from three, above their season 38%.
Your coverage is giving up the corner; tighten the nail help off the weak-side shooter.
shot diet drifting off plan
2Q 0:40
Matchup tracking: your primary defender has taken the hub 6 of 8 possessions and is tiring.
Rotate the fresh switch defender onto him for the last two-minute stretch.
matchup load flagged
Half
Halftime what-if: down 6, what moves the needle most.
Go zone against their spacing for 6 to 8 possessions; re-sim says it is the highest-leverage change.
live re-sim · +3.2% win prob
Film The clip list keeps building live: by the timeout, every possession of the flagged action is already cut, so the staff shows the correction instead of describing it.

The halftime adjustments are powered by the live simulation: the engine re-runs the rest of the game under each option and surfaces the one that moves win probability most. A staff corrects the read rather than building it from scratch, in the minutes it actually has.

Illustrative live feed on the real live-tagging and re-simulation structure. Demonstration figures.

Case 03 · the scout is never finished
Phase 03 · Post-game, and memory

After the game, it audits itself, then remembers.

The postgame packet audits whether the plan was run and whether the flags it called actually played out, and it reads the game grades against the plan, who delivered on the assignment, pulling those from Game Grades rather than grading here. Then it feeds the next scout, so playing Duke again starts sharper. And it tracks the people the roster does not capture.

Did we run our own plan · plan adherence
Run the pick-and-pop early and often. Called 9 times; ran it 11. Adhered.Ran
Do not switch the hub. Switched it 5 times anyway; 4 led to a Boozer bucket.Broke
Did the opponent flags play out · opponent audit
Attack the pick-and-pop seam. Produced 1.14 PPP over 11 possessions.Hit
Hunt the anchor's fouls. He fouled out with 4:20 left; Duke's rim defense fell off.Hit
Pressure the freshman guards. They handled it better than projected; edge did not materialize.Miss
Pre-game
scout and plan
Live
counters, priced
Post-game
audit vs plan
Next scout
starts sharper
The coach, across programs
The opponent's head coach is profiled like a player and followed from job to job. You are not just scouting this roster; you are scouting the tendencies of the person calling it, wherever he has been.
The officiating crew
The crew working the game is profiled on how it calls: foul rates, how tightly it officiates contact, how it swallows the whistle late. Coaching prep the box score never gave you.

A one-off tool gives you a plan and forgets it. This one grades its own plan, reads the game grades against what it called, and carries the coach and the crew forward, so every game you play against an opponent, the engine knows them better than the last time.

Illustrative audit and loop on the real postgame, coach-profiling, and crew-profiling structure. Coach and crew shown generically. Demonstration figures.

The law underneath
The scout is never finished. And it remembers more than the roster.

A projection tells you what should happen. Game Ops is what you do about it: the plan before, the counter during, the audit after. Each pass feeds the next, so the second plan against an opponent is sharper than the first, and the tenth is sharper still. And the engine holds what a roster sheet cannot, the coach followed across programs, the crew profiled like a coach, so you do not just scout the team in front of you. You scout everything that decides the game.