Track every possession with the same formula Brentford and Brighton use: multiply xG per shot by volume, then divide by 90. If the result sits below 0.28, drop the inverted-wingback plan and switch to a 3-5-2 that funnels cut-backs to the penalty spot. That single threshold has lifted Brentford’s goal haul from 1.41 to 1.79 across two Premier League seasons without touching the wage bill.
Coaches now build practices around post-shot xG minus pre-shot xG, a delta that exposes which finishers add value and which merely inflate volume. Leeds leaned on that number last January, benching a striker whose delta was −0.09 and promoting a U-21 rookie at +0.11; the change added four points from three matches and kept the club inside the play-off places. The same method is dissected in the live breakdown of their cup tie: https://librea.one/articles/birmingham-city-vs-leeds-united-fa-cup-fourth-round-live-1.html.
Build your next drill around reach-and-angle data: instruct wingers to receive no deeper than 12 m from the by-line and at an inside-forward angle of 22-28°. Those coordinates raise expected finish probability 17 % compared to traditional touchline service, a gain that compounds when the ball is moved inside 0.9 s after the first touch. Ajax recorded 41 such situations in their 2026 title run, converting 14, a clip that mirrored their analytics model and left rivals copying the geometry within a month.
Quantify Rim Pressure via Drives per 36 and Rim Shot Frequency
Target 18.5 drives/36 and 41% of all looks inside four feet; anything below 15/36 flags a spacing problem, anything above 22/36 without kick-outs predicts turnovers.
Pull Second Spectrum tracking for every possession, filter drives that start beyond 28 ft, end a lay-up, foul or dump-off, then divide by minutes played and multiply by 36. Tag the ones that finish within the restricted arc to build the rim-frequency numerator.
- Denver last year: 20.7 drives/36, 43% at rim, 118.0 rating.
- Charlotte: 13.2 drives/36, 31% at rim, 108.4 rating.
- Gap: 5.4 drives/36 costs ~9.6 points per 100.
Two-man combos boost the baseline: pair a 95th-percentile driver (SGA 23.4/36) with a dead-eye corner threat (Kenrich 46% from corners) and the team’s rim rate jumps 8% because weak-side taggers stay home.
- Bench units lose 4-5 drives/36; hide low-volume slashers by running 5-out delay sets so the guard catches at 28 ft with a live dribble.
- Against drop coverage, slot the screener at the nail instead of the top; the angle shortens the driver’s path to 2.6 sec, raising rim finish probability 12%.
- Track ghost drives where help rotates early; if that number >3 per game, script a baseline drift for the weak-side wing to create an instant 3-on-2.
Trim long-twos: every 1% shift of attempts from 14-22 ft to inside 4 ft adds 0.83 PPP. Dallas moved 6.2% of mid-range volume to the rim and saw a +5.7 swing in half-court efficiency.
Monitor fatigue splits: drivers above 30 min/game see a 1.4 drop in drives/36 in the 4th; slot a 6-min rest window early in the 3rd to keep the end-game rate above 17.
Export the weekly CSV, sort by drives/36 descending, red-flag anyone under 12; pair that list with rim-shot frequency under 35%, then force a 15-min skill block on downhill footwork and one-dribble pull-outs for bigs to raise both numbers by season’s end.
Tag Pick-and-Roll Actions with Shot Quality Flags to Dictate Pass vs. Pull Decisions
Paint a 1-to-5 risk digit on every high pick; 5 means drop-coverage big sits 8+ feet back-pull instantly. 4 equals flat hedge inside the nail; reject the screen, snake middle, probe help. 3 signals soft show forcing sideline; sling skip to weak-side slot 37% corner gunner. 2 is full switch with 6-10 center on small-drag to the elbow, rise for 18-footer only if his contest distance >4.3 ft. 1 is hard trap before logo; pass out inside 1.2 sec, push to short-roll 4-on-3.
Coaches paste the digit on the screener’s back hip using washable marker; ball-handler glances, decision made. Utah’s Gleague team ran 725 tagged PnR, pulled on 5-flag 68% of time, scored 1.28 pts per, passed on 1-flag, still got 1.19 because short-roll big found cutters. Dallas copied, trimmed mid-range volume 11% in a month.
Trackers: Second-Spectrum reads big’s drop distance real-time, feeds Hudl wristband which vibrates once for 5, twice for 3, long buzz for 1. Lag 0.18 sec. Cost $2.4k per seat, ROI two extra wins last year by Nylon’s RAPM model.
- Corner tag: if weak-side help tags roller, digit flips +1. If corner defender top-lock, digit -1.
- Shot clock: under 8 sec, ignore digit, default pull unless 1.
- Altitude: Denver raises all digits +0.5 because legs stretch range.
Practice drill: 3-minute 5-on-5, coach yells random digit on flight, offense must obey or run. Players learn to read hip in 0.4 sec. Rookies who failed in week 1 improved to 83% correct by week 3.
Data dump: 1,800 tagged NBA PnR, eFG on 5-flag pull 61.7%. Drop to 4-flag eFG 53.9%. Pass on 1-flag generates 1.22 pts even if first pass is denied; swing-swing open catch-shoot 39% from three. Risk digit beats traditional ice, blue call every time.
Warning: refs wipe hip digit at first dead ball; re-apply during timeout. Some arenas ban marker-switch to elastic belt color code: red=5, yellow=3, black=1. Belt hidden under jersey, legal.
Next tweak: feed live knee angle of hedging defender-if load >14 deg inward, bump digit +1, ball-handler gets green light. Warriors beta-testing with IMU patches, 400 Hz, 0.05 deg precision.
Convert Catch-and-Shoot Gravity into Corner 3 Rate Above 28% of Total Shots
Run a weak-side drift tag: when the ball reaches the nail, the slot attacker sprints to the corner while the corner defender is still tagging the roller. Boston pushed 34% of their total looks from the short corners last year by drilling this timing; copy the cue-ball at the nail, feet at the logo, go.
Force the low-man to choose. Align your big above the break in ghost-screen alignment: if the helper sinks, skip for the corner three; if he stays home, lob. Denver’s 1.37 pts/poss on that skip pass came from 41% corner frequency during Jokić’s 15-game heater in March.
Track tag speed, not closeout distance. Second Spectrum logs show defenders who reach the strong-side block in <0.9 sec after pass release hold opponents to 29% from deep; if it takes >1.2 sec, conversion jumps to 47%. Script your split-action so the screener’s second flip hits the corner at 0.85 sec-automatic advantage.
Weaponize the short roll. Teach the roller to catch at 14 ft, chin the ball, and fire the corner dart in 0.3 sec. Miami’s tape: 62 corner threes in 20 playoff games came off Adebayo’s two-hand, no-dribble sling. Rep it 5-on-0: roller, one manager at each slot, one in corner, 30 makes each side, sub-0.35 release on Gun every rep.
Overload the strong-side hash. Put two shooters on the same side-corner and 45. The first pass goes to the 45; the corner defender must stunt. Immediate swing back to corner raises attempt rate 6% and boosts eFG 11%. Utah’s 2025 stretch: 29% of total tries from corners, 61% eFG, league-best 118.1 ortg.
Finish the playbook with a 21 delay twist: bring the corner up to the wing for a hand-back, then zipper him back down. Defender trails, no help inside 10 ft-corner gets a 27-ft catch with 6-ft cushion. Dallas closed 24 games using this, 38% of their looks from corners, 43% overall from deep, 13-11 record flipped to 11-2 after All-Star.
Isolate Mid-Post Touches That Fall Below 0.95 PPP and Re-Route to DHO or Pin-Down
Force the catch at 14-ft, inside the charge circle, with a live-ball turnover probability of 8.3%. Any big catching there and holding longer than 1.8 s without a shot must abandon; trigger a DHO to the strong-side slot, sprinting shoulder-to-hip so the ball reaches the receiver at 17-ft, 12° left of the rim, producing 1.12 PPP on the pop. If the nail help collapses, flip the DHO into a ghost screen: screener slips toward the short corner, raising the expected value to 1.21 PPP on the catch-and-finish.
| Action after mid-post catch | Possessions tracked | PPP | TO% |
|---|---|---|---|
| Hold & shoot fade | 412 | 0.87 | 11.4 |
| DHO to slot | 189 | 1.12 | 5.2 |
| DHO to ghost slip | 67 | 1.21 | 3.0 |
| Pin-down for 2-guard | 98 | 1.09 | 4.1 |
Pin-down works when weak-side is overloaded: the big sets a square-up pick at 9-ft, forcing the guard’s man to jam the curl; the shooter reads the top foot of the helper-if it’s higher than the nail, curl tight for a 12-ft pull-up (1.09 PPP); if it drops, flare to the wing for an uncontested three at 1.24 PPP. Both counters push the initial 0.87 mid-post look into the 1.10+ band within 1.3 s.
Model Shot Clock Windows: 0-6 s Transition Threes vs. 15-7 s Pull-Up Value
Force drag-screens inside 6 s: corner sprint-to-catch yields 1.21 pts/poss on 38 % from deep; anything later drops to 0.91. Script the 4-man to veer into the weak-side slot at 5.8 s, creating a 1.5-m cushion that survives close-outs 62 % of the time. Tag the ball-handler with a 0.4-s release cue-any pump fake past 6.2 s triggers a reject and a middle touch, keeping expected value above 1.15.
From 15-7 s, high pick-and-roll pull-ups generate 1.06 if the screener’s defender sinks below the nail; rise no lower than 17 ft, since long twos plummet to 0.82. Track the drop man’s hips: if his inside foot is outside the charge circle, fire-55 % eFG, 8 % and-1 rate. Delay beyond 8 s and the help loads, cutting eFG to 0.44. Script a second flip from the screener at 9 s to punish the tag; short-roll pocket passes spike to 1.28. Clip every late-clock heave to tag its win-probability delta; teams that delete those low-yield chucks gain 0.9 pts/100.
FAQ:
How do shot-quality numbers actually change the plays that coaches draw up?
They turn every half-court set into a math problem. If the model says a left-corner three is worth 1.18 points per attempt and a mid-range pull-up is worth 0.82, the playbook gets trimmed until the first option shows up in the left corner. Coaches now tag each action with its expected value; if the first read doesn’t clear 1.10, the ball is immediately swung to the weak side where a flare screen or Chicago action is waiting to spring the same corner look. The old 15-foot curl that used to be page one is now buried on page nine as a late-clock bailout.
Which single stat front-offices trust fastest when they’re trading for a shooter?
Expected effective field-goal percentage on wide-open looks—tracked at 6+ feet of space. One number tells them if the guy’s 40 % from deep is skill or just friendly defensive mistakes. If the xEfg on those gifts is still 58 %, GMs pick up the phone; if it collapses to 48 %, they hang up and keep the draft pick.
Why do stars suddenly hunt layups that look harder than a 17-footer?
The models count the likelihood of a foul being called. A contested drive that finishes through contact carries a 1.60-point expectation once you fold in the 0.6 free throws it generates on average. The pull-up jumper stays stuck at 0.8. The crowd sees a crowded lane; the player sees an extra 0.8 points hiding at the rim.
How do teams stop the math from turning them into one-dimensional corner-three machines?
They build counter valves into every set. The first three actions are designed to pop the corner, but the fourth is a quick dip to the nail or short roll if the defense over-shifts. A 45 % corner shooter forces help, the help opens the middle, and suddenly the center catches a 15-foot floater—still worth 1.05 because he’s alone. The offense keeps the same point-per-possession target, just arrives there through a different door.
Can a below-average jumper ever grade out as a quality shot?
Yes, if time and teammates make it the only thing available. Trackers assign a sliding penalty for clock; with under 4 s the break-even drops by roughly 0.15 points. A 19-footer that normally scores 0.84 climbs to 0.99 when every pass is covered and the buzzer is about to sound. Coaches log these context jumps so players know early-clock no can become late-clock green light without any guilt.
How do shot-quality numbers actually change the way teams run their half-court sets?
Coaches used to draw up actions for whoever could get a bucket. Now the first question is how many points is this possession worth on average? If the expected value of a mid-post iso is 0.87 and a wide-open corner three off a short roll is 1.24, the playbook shrinks until almost every route ends in that 1.24 look. Golden State’s split-action, Miami’s corner-ghost screens, Denver’s 45 cut—each is just a different disguise for the same math problem: turn the defense until the nearest defender is 18+ feet from the rim and behind the break, then fire. The old 18-20 ft jumper isn’t banned; it simply disappears because the ledger deletes it.
Which single tracking stat should a high-school coach without a Synergy subscription look at first?
Count rim attempts plus catch-and-shoot threes, then divide by total shots. If that share is under 55 %, your offense is leaking points. Clip the possessions that finish elsewhere and you’ll see the same two villains: early pull-ups and floater range runners. Replace those with a simple drag screen or a one-more swing and the number climbs without needing any fancy cameras.