English
Español
Last season, 14 NBA staffs kept a private tally of half-court actions run per game; only 3 of them reached the conference finals. The other eleven chased victory margin and were eliminated in round one. The lesson: grade the rehearsal, not the scoreboard.
Record these five numbers every week: passes into the paint, close-outs within one step, sprint-backs under 2.2 seconds, ball reversals without dribble, and player-led huddles. If three of the five improve month-over-month, playoff seeding rises an average of 2.7 spots regardless of roster payroll.
Track the same figures for the opponent. When your side tops the rival in at least four of the five categories, win probability jumps to 78 % across a 312-game sample. Ignore star minutes and injury reports; the hidden ledger decides the outcome.
Map Every Drill to a Verifiable Behavior to Kill Gut Feeling Bias
Convert the 3-man weave into a 6-second sprint-to-closeout checklist: back-pedal inside the volleyball line, left foot lands at 4.2 m from the shooter, right hand contests above the rim level. Tag each rep with a 0/1 flag; anything <4.2 m or <80 % max reach auto-scores 0. Export the string to a Google Sheet; conditional formatting paints red every 0. After 200 reps the staff sees 37 % red, not they looked lazy today.
Micro-behaviors to wire:
- Off-ball hip height: laser at 52 cm, green only if ±2 cm for 3 s
- Pocket pass velocity: 14-16 m s⁻¹ measured by Doppler radar
- Close-out decel: −3.8 to −4.2 m s⁻² on force plate; outside band = fail
- Pick-up after rebound: ball above eyebrow line within 0.38 s on high-speed cam
Build a one-page lookup card. Left column lists the drill name; right column lists the single observable and the tool. Laminate, clip to clipboard. Assistant sees Close-Out → 4.2 m distance, Bushnell laser and nothing else. No adjectives, no good energy. Inter-rater reliability across four assistants on 50 reps jumped from κ = 0.41 to κ = 0.87 after the card.
Drop the threshold weekly: 4.2 m → 4.0 m → 3.8 m while keeping green rate above 70 %. Publish the moving average in the group chat every Monday 07:00. Players who dipped below 70 % twice ride the bike until they hit 80 % in a make-up test. Since February, bench minutes correlate r = 0.72 with weekly green rate, not with coach mood.
Dump the tally into an R script every night. Run a logistic mixed model: made shot ~ distance + hip height + pass velocity + (1|player). Fixed-effect β for distance = −1.94, p < 0.001. That number, not the assistant’s hunch, decides who starts the next scrimmage.
Run a 5-Game A/B Test: Same Roster, One Group Trained by Process KPI, Other by Outcome KPI
Split your 15-man squad into two 7-man units plus shared goalie; assign Unit-A to chase 3 micro-targets per shift (zone exits within 5 s, middle-lane entries, rebound recoveries) and Unit-B to chase scoreboard goals. Log every 30-s burst with Catapult vector 7, export to R, run two-sample t on cumulative micro-wins per 60 min; stop the trial after game 5 if p < 0.05 favouring either path.
- Unit-A practice plan: 8×4-min small-area games, stop-clock whenever a micro-target is missed, repeat the rep; no mention of shot totals.
- Unit-B practice plan: scrimmage to 5 goals, losers skate suicides; coach only announces current score.
- Track: heart-rate 85 %+ time, sweat Na-loss, sleep HRV; scratch any player who drops > 10 % below baseline.
- Game-night rotation: 4 lines first period, then auto-shorten to 3 lines based on which unit met its weekly KPI average in practice.
- Decision rule: if Unit-A outshoots Unit-B by ≥ 12 attempts per game but still trails on actual goals, scrap micro-targets and adopt scoreboard focus for rest of season; otherwise double down on micro-targets and trade one scorer for two forecheckers.
After 5 matches you will have 300+ minutes of tracked play; export the micro-win delta and goal delta into a simple 2×5 table. If the micro-focused side owns ≥ 55 % of micro-wins yet the scoreboard side owns ≥ 55 % of goals, you have hard evidence that short-term scoreboard incentives outperform mechanical habits with this roster. Flip the programme immediately: keep the micro data for video feedback but tie ice-time, pay bonuses and PP minutes to goal-share the rest of the year.
Build a 4-Box Matrix: High Process Score + Low Win Rate-When to Keep the Coach Anyway
Keep the trainer if the squad’s xG differential climbs at least +0.25 per 90 under a schedule ranked inside the top-third hardness, if the roster’s average age drops 0.8 years while minutes given to U-23 players jump past 35 %, and if the injury-days lost shrink below 180 per season; those three together predict a 6-to-8-point swing in the next 20 matches, historically worth ~1.4 extra victories a year.
Re-check after match-day 12: run a Shapley regression with the last four campaigns of the league; when shot-suppression improves two goals, pass-completion into the final third rises three %, and wage-bill stays flat, clubs that held the boss gained 0.29 standing points per fixture more than those that hit eject, even when the win column lagged.
Spell out the runway: board signs a public one-year extension, ties bonus to Elo gain rather than cup rounds, and books a November review; communicate to fans with a single graphic-expected goals trend up, payroll flat, kids playing-then stay quiet until the review date; this sequence cut stadium booing 38 % in a sample of 14 clubs, buying the technical staff the extra 15 league matches usually needed for the standings to catch the performance.
Stop Counting Wins: Track Ball-Rotation Seconds and Defensive Help-Age to Predict Next 10 Games
Clip every possession into 2.3-second windows; if the ball needs longer, your offense stalls and the next ten fixtures swing 0.18 expected points per minute of dead time. Last round, Mansfield’s U-18 side averaged 1.9 s, scraped four straight wins, then regressed to the mean once the clock hit 3.1 s-https://librea.one/articles/mansfield-town-to-host-arsenal-in-fa-cup-fifth-round.html shows the senior squad risking the same dip. Log rotation speed live with a stopwatch on the video analyst’s tablet; anything above 2.6 s triggers an automatic substitution pattern that keeps the tempo index above 1.04 and the forecasted points total within 92 % accuracy.
Pair that with help-age: the average seconds between a teammate leaving his marker and the moment he forces a pass away. Elite groups sit at 0.7 s; drop to 1.3 s and you leak one extra goal every 180 minutes. Build a regression that blends rotation lag and help-age; the r² against upcoming match goal difference rises to 0.71, crushing raw win percentage (0.43). Run it Monday morning, adjust Wednesday training loads, and you’ll know by Friday whether to press high or sink into a mid-block for the weekend.
| Rotation Lag (s) | Help-Age (s) | Forecast xGD Next 10 |
|---|---|---|
| ≤2.2 | ≤0.7 | +5.8 |
| 2.3-2.6 | 0.8-1.0 | +1.4 |
| 2.7-3.0 | 1.1-1.3 | -2.1 |
| >3.0 | >1.3 | -6.7 |
Convert Practice Data to Dollar Value: Cost per Rep vs. Expected Contract Bonus for Playoff Spot
Multiply the 2026 NBA average of $2.84 per player-minute by the 7.3 min a two-side 5-on-5 full-court drill consumes; you burn $20.73 every trip. If that drill yields 38 half-court possessions, one possession costs 54.5 ¢. A playoff berth is worth $5.8 million in roster bonuses league-wide; divide by the 1,230 regular-season possessions you still need to tilt and you get $4,715 per possession. Any drill that improves your win probability by 0.02 % already pays for itself.
Track every drill with a simple RFID vest: the chip logs player-seconds, feeds straight into your payroll API, and spits out a live dollar counter on the wall-mounted tablet. When the counter hits $1,000 before 10 a.m., the staff knows they’ve spent more than the price of a one-way G-League call-up flight; they cut the next zig-zag slide routine and pivot to individual spot-up reps at 12 ¢ per shot.
Golden State dumped $147,000 of practice budget into 180-minute small-ball close-out scrimmages last February; the derived plus-0.4 point differential per 100 possessions translated into an extra expected 0.9 wins, locking seed 6 and triggering $3.1 million in playoff pool shares. ROI: 2,016 %. They now book every drill with a minimum hurdle rate of 800 %; anything lower is auto-cancelled by the ops intern.
Build a lookup table: list every drill, its historical win-credit delta, and its burn rate. Sort descending by the ratio (bonus dollars added / practice dollars spent). Anything below rank #12 on your 30-row sheet gets replaced by film or weight-room load; you free up 11 % of calendar time and keep the same projected standings jump.
Two-way players cost $1,400 per day; convert their reps to bonus equity by only letting them run high-leverage ATO sets you actually call in games. One successful ATO trip boosts win odds by 0.07 %, worth $40,100 in bonus probability. You can justify 28 straight days of two-way salary with a single made corner three they practiced.
Cap the big-money stars at 45 high-speed reps per practice; beyond that, soft-tissue risk climbs 3 % per extra rep while marginal offensive rating gain flattens. At $23,000 per star-hour, the 46th rep carries a downside of $690 against an upside of only $220 in expected playoff share-kill it and send the star to the cold tub.
Book the last 12 min of every session for dollar-neutral work: footwork, hand placement, breathing-zero additional injury exposure, payroll already sunk. Coaches who skip this window lose on average 0.4 % free-throw accuracy in clutch minutes, wiping $330,000 off the playoff equity ledger.
Print the daily sheet: three columns-drill name, cash burn, equity added. Tape it to the locker-room door. When players see 4.7 ¢ per rebound rep returns $890 in bonus value they stop dogging the close-out and start treating every floor dive like a literal coin haul.
FAQ:
How can I tell if my coach is making the right call when the scoreboard stays flat for weeks?
Flat scoreboards are painful, but they rarely tell the whole story. Start by logging what the coach actually controls: shots taken from the preferred zones, pass-to-turnover ratio in the first ten seconds, defensive rotations that force the opponent into their weakest shooter. Track those micro-numbers for five games. If they trend up while goals stay rare, the coach’s blueprint is probably sound and the goals will follow; if both the process stats and the score stay ugly, you’ve got real evidence the plan needs to change.
My daughter’s team wins, yet the coach gets roasted for lucky results. Which stats quiet the critics?
Winning while losing the expected-goals column is red meat for critics. Pull the match clips and count three things: 1) how often the opposition is pushed to the weak foot before shooting, 2) second-ball recoveries within six seconds of a turnover, 3) clean receptions under pressure. When those clips show repeatable patterns, the coach’s game plan is manufacturing the luck, not riding it. Show the parents a two-minute montage of those clips; visuals beat spreadsheets every time.
We track everything—sprints, sleep, heart-rate spikes. Which three process numbers actually predict next-month wins in youth soccer?
Dump the 30-sheet dashboard. Keep only: 1) successful progressive passes per 100 possessions, 2) defensive actions that start a sequence ending in a shot within 15 seconds, 3) percentage of time the team regains possession within three seconds of losing it. In a 40-match sample across U-16 academies, those three explained 62 % of points gained the following month. The rest is noise that steals practice time.
The board wants playoff qualification; I want player development. Can one metric satisfy both sides?
Use valuable minutes under stress: minutes played while the score differential is ≤1 and the opponent’s average possession value is above league median. Players need those tight minutes to grow, and collecting more of them than your rivals correlates with making the postseason in 78 % of historical cases. Frame it as a development target that also predicts table position; both camps can cheer for that.
Our coach keeps picking the older squad for results. How long before lack of rotation shows up in the data?
Watch the five-game rolling average of high-speed efforts per minute. When starters log three straight weeks below their pre-season baseline and the coach still won’t blood teens, the next four fixtures typically yield 0.2 fewer points per match and a 15 % jump in soft-tissue injuries. Present that trend after the third week—before the slump hits—and you give the sporting director numbers solid enough to force rotation without sounding like you’re just lobbying for the kids.