Worked session three, and the review
Lesson 24 · about 11 min
The third worked session is a trend day, the type that punishes the habits that worked in the first two. It is followed by the end-of-day review, which is where the playbook actually improves. Same hypothetical product, 0.25 ticks; numbers invented to be realistic.
Session three: the trend day
Prep. Yesterday (session two): POC 5007.00, VAH 5008.25, VAL 5006.00, high 5008.75, low 5005.50, closed 5006.00, b-shape (liquidation built value low). Overnight: high 5006.75, low 5001.50, POC 5003.00, currently 5002.25. IB median 9.5.
Overnight traded entirely below yesterday's value and made a new multi-day low. Levels ranked: 5003.00 (overnight POC, rank 1), 5005.50 (two-day low, now resistance, rank 2), 5001.50 (overnight low, stops below, rank 3), 5000.00 (round number, prior-week HVN edge, rank 4), 5007.00 (yesterday POC, rank 5).
Scenarios: A, continuation lower (opens below 5003, IB narrow, acceptance below 5001.50; short pullbacks to 5003 with responsive selling, target 5000 then lower). B, failure of the overnight low (absorption below 5001.50, long the swing failure to 5005.50). C, open above 5003 and reclaim of 5005.50 (long, but low probability given the b-shape and overnight).
Expected values from the log: A +0.45R, B +0.50R, C +0.10R (skip unless five yeses).
The open. 5002.00. Bars 1 to 3: deltas −540, −610, −480; lows 5001.50, 5001.00, 5000.25. Open-drive down. The overnight low at 5001.50 was passed in bar 1 with 1,900 contracts on the bid and only 620 on the ask across the level, price stepping down four ticks: acceptance, not absorption. Scenario B is dead at bar 2. Scenario A is live.
Bar 5 at 5000.00. The ranked-4 level. Checklist for a long (the fade temptation on a round number): aggression in, yes. DOM bids at 5000.00 were 420, pulled to 90 as price arrived (no). Volume at 5000.00 is 680 with price stepping through (acceptance, no). Low row unfinished. Delta not flipped. One yes. Pass. The trend is accepting each level.
The IB. 4997.50 to 5002.75, 5.25 points, 55% of median. Narrow, one-directional, on an open-drive. The plan said: on a trend day, join pullbacks; never fade.
Bar 14, the pullback to 5000.00. After the IB, price retraced from 4997.50 to test 5000.00 from below, which is now the ranked-4 level acting as resistance.
Price | Bid x Ask
5000.50 | 45 x 6 <- finished high
5000.25 | 380 x 210
5000.00 | 740 x 520
4999.75 | 310 x 190
4999.50 | 120 x 60
Delta: −735 Close: 4999.50
Checklist for a short: aggression in (the pullback lifted with +410 delta in bar 13, yes). DOM offers at 5000.00 added from 180 to 510 as price approached (yes). 1,260 contracts at 5000.00 with one tick of progress, delta −735 (absorption, yes). Top row 45 × 6 (finished, yes). Close 4999.50, below the level (flip, yes). Five yeses.
Trade 1. Short 4999.50. Stop 5000.75 (5 ticks). Target 4995.00 (18 ticks; the prior-week LVN edge). 3.6R.
| Bar | High | Low | Close | Delta | Action |
|---|---|---|---|---|---|
| 15 | 4999.75 | 4998.00 | 4998.25 | −480 | Working |
| 16 | 4998.50 | 4997.00 | 4997.25 | −390 | Through the IB low; one third off at 1R (4998.25) |
| 17 | 4997.50 | 4996.00 | 4996.25 | −610 | Acceptance below IB |
| 18 | 4996.75 | 4995.25 | 4995.50 | −280 | Approaching target; delta thinning |
| 19 | 4995.75 | 4994.75 | 4995.25 | +90 | At 4995.00: 1,600 contracts, no progress; absorption |
Exit two thirds at 4995.25. Blended: one third at +1.0R, two thirds at (4999.50 − 4995.25) ÷ 1.25 = +3.4R. +2.6R.
The afternoon. Price built a new balance between 4995.00 and 4998.00. A retest of 4998.00 produced three yeses (absorption present, extreme unfinished, no flip); passed. Closed 4996.00. Day: trend day down, closing in the lower third. Session result: +2.6R, one trade, two passes.
The trade that would have lost money was the one-yes long at 5000.00 in bar 5, the round-number fade on a trend day. The checklist removed it.
Key idea: On a trend day the checklist protects you from your own fades. Levels get one yes and are accepted; the trade is the pullback that gets five. Same procedure, opposite side.
The end-of-day review
The review takes fifteen minutes and produces the numbers the next day's prep will use. Do it every session, including the ones with no trades.
1. Log every level touch, not just trades
| Session | Level | Direction | Yes count | Traded? | Outcome (R) | Correct pass/take? |
|---|---|---|---|---|---|---|
| 1 | 5013.00 | Short | 5 | Yes | +2.2 | Yes |
| 1 | 5011.00 | Short | 2 | No | Would have lost about −1 | Yes |
| 2 | 5008.00 | Long | 2.5 | No | Would have lost about −1 | Yes |
| 2 | 5007.75 | Long | 5 | Yes | −0.22 | Yes |
| 2 | 5007.50 | Short | 3 | No | Would have won about +2 | Yes (by rule) |
| 3 | 5000.00 | Long | 1 | No | Would have lost about −1 | Yes |
| 3 | 5000.00 | Short | 5 | Yes | +2.6 | Yes |
| 3 | 4998.00 | Short | 3 | No | Would have won about +0.5 | Yes (by rule) |
After fifty rows, this table tells you the win rate and average R by yes count. If fours pay nearly as well as fives on your product, you can loosen the rule; if threes are coin flips, keep passing. Nobody else's numbers apply to your product and your eyes.
2. Score the prep
| Question | Session 1 | Session 2 | Session 3 |
|---|---|---|---|
| Did a written scenario play out? | B | B then A | A |
| Did the rank-1 or rank-2 level matter? | Rank 2 | Rank 2 | Rank 4 (1 and 2 untouched) |
| Was the IB read (narrow/wide) consistent with the day? | Narrow, neutral: no | Normal, normal variation: yes | Narrow, trend: yes |
| Were expectancy estimates close? | Too early to say |
3. One sentence
Finish with a single sentence about what the market did in auction terms: "Rejected the overnight high with absorption and rotated back into yesterday's value." "Accepted below the overnight low after a stop run failed to hold." "Open-drive down, accepted every level, trended to the weekly LVN." If you cannot write the sentence, you did not understand the session, and that is worth knowing before tomorrow.
Three sessions, summed
| Trades | Passes | Net R | Process errors | |
|---|---|---|---|---|
| Total | 3 | 5 | +4.58R | 0 |
Three sessions is not evidence of an edge. It is evidence of a process that generates a log which can, over fifty or a hundred sessions, tell you whether you have one. That is the honest claim this course makes for order flow trading: not that it wins, but that it lets you find out whether you do, at a level of detail that a candle chart cannot provide.
Try it: Set up the level-touch log with the columns above and keep it for twenty sessions, in replay or live at minimal size. At the end, compute win rate and average R for touches with five yeses, four, and three or fewer. That table is your edge, or the proof that you do not yet have one, and either answer is worth the twenty days.
Recap
- Session three: open-drive down, levels accepted one after another, a one-yes fade correctly passed, a five-yes pullback short for +2.6R.
- On trend days the checklist's main job is to keep you off the fades; the trade is the pullback in the trend direction.
- Log every level touch with its yes count and outcome, including passes with what they would have done.
- Score the prep against what happened and write one auction-terms sentence about the day.
- Three sessions prove nothing; fifty rows of level-touch data tell you whether the confirmation rule works on your product.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.