Train Like It's a Game: The Transfer Hierarchy for Goalies

Here is the only question that matters about any drill you run: does what the goalie gets good at in practice show up in the game? Everything else is secondary. If it does not transfer, it was exercise, not training.

In 1901 Edward Thorndike gave us the sharpest tool we have for answering that question. His "identical elements" theory says something almost crude in its simplicity: skill transfers from one context to another in proportion to the number of identical stimulus-response elements the two contexts share. The more practice looks, feels, and decides like the game — the same visual reads, the same time pressure, the same body positions — the more it transfers. The further the two drift apart, the less survives the trip.

Apply that to the crease and you get a hierarchy. Rank every goalie rep you run by how many of the game's real elements it actually contains.

Level 1 — Maximum transfer (≥80% identical elements).

The goalie is reading a live play that could go several ways. A 3-on-3 in a small area: real shooters, passing lanes opening and closing, a screen that actually blocks sight, a rebound that goes where physics sends it. A 2-on-1 where the goalie has to respect the pass and still be square to the shot. Spend the back half of most sessions here.

Level 2 — High transfer (60–80%).

A shot that arrives after a read — off a pass across the slot, after a controlled rush — with light, passive pressure. The sequence is real; the chaos is dialed down.

Level 3 — Moderate transfer (40–60%).

A shooter passing from a fixed spot, the goalie tracking and reacting. The mechanics are partly live, but nothing moves, reads, or disguises intent the way a human does.

Level 4 — Low transfer (under 40%).

The classic shooting line: pucks fed from a cone or a stationary coach, a predictable sequence from a predictable angle, no read and no pressure. Technique can be grooved here — but in conditions that barely resemble a game.

Level 5 — Zero or negative transfer.

A flawed pattern repeated until it's automatic. A young goalie who drops into the butterfly early every rep because the drill never punishes it is not neutral training — it's anti-training. In the game, under speed, that grooved habit fires before the correct read can. Schmidt and Lee call this proactive interference: the wrong pattern actively crowds out the right one.

The trap is obvious once you name it. Most goalie practice lives at Levels 3 and 4, because those drills are easy to run, easy to control, and produce a satisfying number of saves. They feel productive. But a goalie who only ever faces Level-4 reps builds a false sense of competence: flawless in the drill, a step late in the game, because the game was never in the drill.

This doesn't mean Level 4 is useless. When you're installing a brand-new technique or rebuilding a broken one, you want the quiet, controlled environment — it's where the pattern gets built.

The mistake is staying there. Within a single session the progression should climb: build the piece in isolation, stress it with movement and variability, then put it inside a live read. Start at Level 3, finish at Level 1 (as already talked the 1-2-3 concept) . Don't get stuck at the bottom of the ladder and call it a workout.

Next time you plan a goalie session, do one thing: write the Thorndike level next to each block. If most of your minutes sit at 3 and 4, you've just found why the work isn't showing up on Saturday.