Seeds · scanning a position

Seeds of tactical destruction

A tactic is not remembered. It is found — by sweeping a position for the features that make one possible, and by knowing when those features are telling you nothing.

Not started
guess “quiet” — 29%ceiling — 78%
clean = right, and inside the cap
Finish board vision first

This course assumes you can sweep a position for loose material and count attackers and defenders under a time cap, batteries included. The entry condition is the board vision final check at 18 of 20 within cap.

Naming loose pieces drop off without reliably finding a loose piece is a slogan, and it is the failure this whole sequence was built to avoid.

What you will be able to do

Given a position you have not seen: name every seed present, find the square where they converge, and read what that convergence is and is not telling you.

Note what that does not say. It does not say you will decide reliably whether a position holds a tactic. You will not, and neither can anything else built on these features Inference. Why, and what to do about it, is the second half of this course.

Not in this course
Before you startRead · ~6 min

What the evidence is worth

Sourced S1 Inference Contested Single study Vendor framing

There is no experimental literature on teaching tactical vision. What exists is coaching craft: Heisman's twelve seeds Sourced S1, and Neiman's near-identical list of signals, arrived at independently Sourced S2. Two practitioners overlapping is the strongest evidence available here, and it is still not evidence that any of this transfers to playing strength Contested.

The numbers are different in kind. The 29% you start at, the 78% ceiling, the band thresholds — all measured on 2,100 positions from real games Sourced S3, and every one can be recomputed.

Time

BlockModeEstimate
1 · Three seedsDrill~1 hr
2 · Six seedsDrill~1 hr
3 · All twelveDrill~1.5 hrs
4 · From memoryDrill~1 hr
5 · ConvergenceRead, then drill~1.5 hrs
6 · The ceilingCalibrate~1 hr
7 · Mixed checkGate~40 min

About 7 hours of blocks. People take roughly two and a half times their estimate Sourced S4, so plan for around 17. Each block closes on its own; none needs a desk.

Where you are meant to feel stuck

Block 3 is the hard one. Twelve seeds is more than working memory holds, and for a while you will find eight and miss two, then find those two and miss three others. That stretch is the crossing, not a sign you are doing it wrong.

Block 6 is uncomfortable differently: it is where the scan stops paying off and you have to accept a number lower than you want.

PacketFor the tutor

Take a miss to the tutor

Any item you have answered can be added. Build it, copy it, paste it into the tutor project — the position, the question, the right answer, yours, the time and your confidence, so the conversation starts at the disagreement instead of at reconstruction.

0 items