@fii_dii_data17 · post date 21/08/2026 (2026-08-21T15:19:41)
Account's Own CallBullish
This IS the live bias
Read directly off @fii_dii_data17's own "OVERALL TREND" line in the image — drives the hero bias and strategy tab below
Our Domain-Prior WeightingBearish
7 bullish / 5 bearish rows
FII 40% / PRO 30% / DII 15% / RETAIL -15% — for comparison only, not what drives the live bias. Agrees with the account's own call 85.4% of the time historically; both score ~52% hard-day accuracy (coin-flip) when backtested independently.
⚠️ The account's own call and our weighting disagree today — the live bias uses the account's own call (left).
Participant
Instrument
OI Change
Trend
FII
Future
+2,258
Bullish
FII
CE
+15,680
Bullish
FII
PE
-10,969
Bullish
PRO
Future
-806
Bearish
PRO
CE
+30,843
Bullish
PRO
PE
+17,843
Bearish
DII
Future
+837
Bullish
DII
CE
+50
Bearish
DII
PE
Sold Puts
Bullish
RETAIL
Future
-2,289
Bearish
RETAIL
CE
-46,473
Bearish
RETAIL
PE
-6,874
Bullish
Options Data Intelligence
Fyers Option Chain + PE-CE OI Dynamics
Total PE-CE OI DiffCall Resistance Active
-14,384
-1.65% imbalance · Max Pain strike target at 24300
PCR (Put-Call Ratio)Neutral Balance
0.97
PE OI: 427,919 vs CE OI: 442,303
Max Pain StrikeExpiry Target
24300
Option writers payout minimized at this level
Futures StructureLong Build-Up
Long Build-Up
Price & OI co-movement classification
Total PE-CE OI Diff Table (15-Min Intervals)
Intraday 15m Futures Price, VWAP, PCR & Options OI Shift
Time
Fut Price
Fut Chg (Day)
Fut Chg
VWAP
IV
Fut Vol
Tot Fut OI
Fut OI Chg (Day)
Fut OI Chg
PCR-OI
Tot CE OI
Tot PE OI
PCR-OI Chg
Diff (PE-CE OI Chg)
Diff (%)
CE OI Chg
PE OI Chg
OI Chg Trend
09:30
24,284.90
-6.00 (-0.02%)
+2.50
24,284.90
7.7
800
166,724
-470
+675
0.75
503,822
377,477
4.2
-126,345
-14.34%
+2,500
+10,500
+8,000
09:45
24,294.03
+3.13 (+0.01%)
+9.13
24,291.48
7.82
2,062
169,839
-290
+3,115
0.88
479,659
423,014
1.88
-56,645
-6.28%
-24,163
+45,537
+69,700
10:00
24,301.33
+10.43 (+0.04%)
+7.30
24,295.72
7.93
2,163
172,434
-110
+2,595
1.02
458,403
466,296
2.04
+7,893
+0.85%
-21,256
+43,282
+64,538
10:15
24,305.71
+14.81 (+0.06%)
+4.38
24,297.39
8.01
1,011
174,127
+70
+1,693
1.13
442,608
499,654
2.11
+57,046
+6.05%
-15,795
+33,358
+49,153
10:30
24,306.45
+15.55 (+0.06%)
+0.74
24,299.59
8.07
1,935
174,666
+250
+539
1.19
434,803
516,401
2.15
+81,598
+8.58%
-7,805
+16,747
+24,552
10:45
24,303.39
+12.49 (+0.05%)
-3.06
24,300.42
8.1
2,238
173,995
+430
-671
1.17
436,817
512,527
-1.92
+75,710
+7.97%
+2,014
-3,874
-5,888
11:00
24,296.91
+6.01 (+0.02%)
-6.48
24,300.05
8.09
1,219
172,247
+610
-1,748
1.08
449,561
487,428
-1.97
+37,867
+4.04%
+12,744
-25,099
-37,843
11:15
24,287.91
-2.99 (-0.01%)
-9.00
24,298.41
8.05
1,785
169,737
+790
-2,510
0.94
472,523
444,367
-1.88
-28,156
-3.07%
+22,962
-43,061
-66,023
11:30
24,277.67
-13.23 (-0.05%)
-10.24
24,295.35
7.97
2,284
166,913
+970
-2,824
0.77
503,591
390,007
-1.75
-113,584
-12.71%
+31,068
-54,360
-85,428
11:45
24,267.64
-23.26 (-0.10%)
-10.03
24,293.03
7.87
1,418
164,283
+1,150
-2,630
0.62
539,175
333,219
-1.6
-205,956
-23.61%
+35,584
-56,788
-92,372
12:00
24,259.26
-31.64 (-0.13%)
-8.38
24,290.08
7.76
1,616
162,350
+1,330
-1,933
0.49
574,422
283,640
-1.41
-290,782
-33.89%
+35,247
-49,579
-84,826
12:15
24,253.70
-37.20 (-0.15%)
-5.56
24,286.07
7.64
2,299
161,525
+1,510
-825
0.41
604,054
249,705
-1.15
-354,349
-41.50%
+29,632
-33,935
-63,567
12:30
24,251.74
-39.16 (-0.16%)
-1.96
24,283.61
7.52
1,604
162,079
+1,690
+554
0.38
623,024
237,282
-0.65
-385,742
-44.84%
+18,970
-12,423
-31,393
12:45
24,253.58
-37.32 (-0.15%)
+1.84
24,281.81
7.42
1,430
164,083
+1,870
+2,004
0.4
627,751
248,236
2.32
-379,515
-43.32%
+4,727
+10,954
+6,227
13:00
24,258.91
-31.99 (-0.13%)
+5.33
24,279.81
7.35
2,285
167,428
+2,050
+3,345
0.45
616,544
280,476
2.88
-336,068
-37.46%
-11,207
+32,240
+43,447
13:15
24,266.87
-24.03 (-0.10%)
+7.96
24,278.99
7.31
1,775
171,814
+2,230
+4,386
0.56
590,456
327,977
1.82
-262,479
-28.58%
-26,088
+47,501
+73,589
13:30
24,276.28
-14.62 (-0.06%)
+9.41
24,278.88
7.3
1,231
176,827
+2,410
+5,013
0.69
552,952
382,339
1.45
-170,613
-18.24%
-37,504
+54,362
+91,866
13:45
24,285.77
-5.13 (-0.02%)
+9.49
24,279.37
7.33
2,242
181,988
+2,590
+5,161
0.85
509,666
434,174
1.2
-75,492
-8.00%
-43,286
+51,835
+95,121
14:00
24,294.00
+3.10 (+0.01%)
+8.23
24,280.21
7.39
1,926
186,828
+2,770
+4,840
1.02
467,459
474,921
0.97
+7,462
+0.79%
-42,207
+40,747
+82,954
14:15
24,299.90
+9.00 (+0.04%)
+5.90
24,280.80
7.48
1,024
190,973
+2,950
+4,145
1.15
433,073
498,573
0.69
+65,500
+7.03%
-34,386
+23,652
+58,038
14:30
24,302.82
+11.92 (+0.05%)
+2.92
24,282.11
7.59
2,169
194,195
+3,130
+3,222
1.22
411,909
502,676
0.19
+90,767
+9.92%
-21,164
+4,103
+25,267
14:45
24,302.62
+11.72 (+0.05%)
-0.20
24,283.20
7.71
2,054
196,445
+3,310
+2,250
1.2
406,963
488,641
-2.84
+81,678
+9.12%
-4,946
-14,035
-9,089
15:00
24,299.68
+8.78 (+0.04%)
-2.94
24,283.54
7.82
813
197,856
+3,490
+1,411
1.1
418,188
461,294
-2.44
+43,106
+4.90%
+11,225
-27,347
-38,572
15:15
24,294.86
+3.96 (+0.02%)
-4.82
24,284.11
7.93
2,069
198,729
+3,670
+873
0.97
442,303
427,919
-1.38
-14,384
-1.65%
+24,115
-33,375
-57,490
Technical Indicators
Daily candles, ~120-day lookback
RSI (14)No data
—
03070100
Not enough history yet.
MACD (12,26,9)No data
—
Not enough history yet.
SMA Trend (5/20/50)No data
—
Not enough history yet.
Bollinger Bands (20, 2σ)No data
—
Not enough history yet.
Classic PivotNo data
—
Not enough history yet.
Opening Range (30m)Not formed yet
—
Market just opened, or data unavailable.
Yesterday's Predicted vs Next-Day Outcome (Last 2 Months)
Day-by-day record of strategy prediction issued on D_(t-1) vs actual Nifty move on D_t
Domain-prior weights (not data-fitted): FII (+40%), PRO (+30%), DII (+15%), RETAIL (-15% contrarian) · threshold 0.10 — see the Weight Optimization tab for why a data-fitted alternative was rejected
FII Weight+40%
+40%
Foreign Institutional Investors — smart-money prior, largest weight
These weights are a domain prior, not fit to this data. A data-driven grid search was tried
and rejected: it initially found FII at -40% (negative), which turned out to be a
collinearity artifact (FII and RETAIL correlate at -0.69 in this sample, so the optimizer could
extract the same predictive value from either sign on either one). Every individual
participant/instrument reading correlates with next-day return at under 0.12 in magnitude —
below the ~0.075 noise floor at this sample size. When re-run constrained to respect
FII/PRO/DII ≥ 0 and RETAIL ≤ 0, out-of-sample accuracy collapsed to
38.9% (worse than a coin flip) — see the Weight Optimization tab for the full breakdown.
Account's own call vs. our weighting — which is actually better?
A separate bug meant this dashboard's live bias was silently using our own weighted computation
even when it disagreed with @fii_dii_data17's own stated "OVERALL TREND" call in the image (now
fixed — the live bias uses the account's own words). The two signals agree
85.4% of the time across
178 posts. Backtested independently, neither
clearly beats the other: the account's own call scores
51.9%
hard-day accuracy and 79.2%
option-seller win rate; our domain-prior weighting scores
51.9%
hard-day accuracy and 77.0%
option-seller win rate — both essentially a coin flip. With no statistical reason to prefer
our derived approximation over the account's own real, verifiable statement, using their words
directly is the more defensible and more transparent choice.
Option-Seller Win Rate79.2% Payout
79.2%
141 Wins out of 178 calls — loses only if Nifty moved >0.6% against the call
Hard-Day AccuracyThe real test
51.9%
Directional accuracy on the 77 days that actually moved >0.6% — a constant bias scores ~50% here. This is honestly ~coin-flip, not a proven edge.
Winning CallsWins
141
Weighted participant prediction matched market move, or adverse move stayed within 0.6%
Losing CallsLosses
37
Market moved >0.6% against the weighted participant bias
Post Date
Weighted Stated Call
Rows (Bull/Bear)
Next-Day Nifty Return
Backtest Result
2026-08-20
Bullish
5B / 7S
+0.08%
WIN
2026-08-19
Bearish
6B / 6S
+0.64%
LOSS
2026-08-18
Bearish
5B / 7S
-0.32%
WIN
2026-08-17
Bearish
5B / 6S
-0.55%
WIN
2026-08-14
Bearish
7B / 5S
-0.32%
WIN
2026-08-13
Neutral
7B / 4S
-0.12%
WIN
2026-08-12
Bearish
6B / 6S
-0.16%
WIN
2026-08-12
Bearish
5B / 7S
-0.16%
WIN
2026-08-07
Bearish
4B / 8S
+0.05%
WIN
2026-08-06
Bullish
6B / 6S
-0.27%
WIN
2026-08-05
Bearish
5B / 7S
+0.05%
WIN
2026-08-04
Bearish
5B / 7S
+0.04%
WIN
2026-08-03
Bullish
7B / 5S
-0.64%
LOSS
2026-07-31
Bearish
4B / 8S
+1.60%
LOSS
2026-07-30
Bearish
4B / 8S
+0.27%
WIN
2026-07-29
Bullish
7B / 5S
+0.28%
WIN
2026-07-28
Bearish
4B / 8S
+1.10%
LOSS
2026-07-27
Bullish
6B / 6S
-0.04%
WIN
2026-07-24
Bullish
6B / 6S
+0.96%
WIN
2026-07-23
Bearish
4B / 8S
-0.43%
WIN
2026-07-22
Bearish
6B / 6S
-0.53%
WIN
2026-07-22
Bearish
4B / 8S
-0.53%
WIN
2026-07-21
Bearish
5B / 7S
-0.79%
WIN
2026-07-18
Bullish
5B / 7S
-0.21%
WIN
2026-07-16
Bullish
5B / 7S
+1.09%
WIN
2026-07-15
Bullish
6B / 6S
-0.02%
WIN
2026-07-14
Bearish
7B / 5S
+0.11%
WIN
2026-07-13
Bullish
5B / 7S
-0.66%
LOSS
2026-07-10
Bullish
5B / 7S
+0.02%
WIN
2026-07-09
Bullish
7B / 5S
+1.02%
WIN
FII/PRO/DII/RETAIL Weight Optimization Experiment
Option-seller scoring: a call only loses if Nifty moves >0.6% against it · ranked by hard-day accuracy, not raw win rate · target 75%
Target AccuracyGoal
75.0%
Requested target, measured on hard-day accuracy (see below)
Constant-Bias Free BaselineNo signal used
~78%
Always calling Bullish (or Bearish) wins this often on raw win rate alone — big adverse moves are rare
Unconstrained Fit: Test Hard-Day72.2% Test
72.2%
Looked good... until sign-constrained (see next card)
Honest read: why the unconstrained fit was rejected
Scoring matches how an option seller actually gets paid: a directional call only loses if
Nifty moves against it by more than 0.6%. That makes raw win rate misleading (a constant
"always Bullish" call already wins ~78% for free), so
candidates are ranked by hard-day accuracy instead — directional accuracy on just the
days that actually moved >0.6%, where a constant bias scores ~50% (a coin flip).
The first (unconstrained) pass found a combo scoring
72.2% out-of-sample — but it did
so by weighting FII negative, which read as backwards given how central FII flow is
supposed to be. Digging in: FII and RETAIL correlate at -0.69 in this 178-day sample, and
every single participant/instrument reading correlates with next-day return at under 0.12 in
magnitude (the noise floor at this sample size is ~0.075). That combination — strong
collinearity, no real univariate signal anywhere — is exactly the setup where an
unconstrained fit finds spurious sign flips instead of real relationships.
Re-running the search constrained to FII/PRO/DII ≥ 0 and RETAIL ≤ 0
(respecting the "smart money helps, retail is contrarian" prior) confirmed it: out-of-sample
accuracy collapsed to 38.9%
— worse than guessing. The live model has been reverted to a domain-prior weighting
(FII 40% / PRO 30% / DII 15% / RETAIL -15%, not fit to this data at all), which scores an
unremarkable but stable ~50-52% hard-day accuracy across train/test/full — consistent
with genuinely having no measurable edge yet on this data, rather than a fit-then-collapse
pattern. 75% is not reachable honestly at this sample size; more history is needed before this
signal can be trusted to override an intuitive prior in either direction.
FII wt
PRO wt
DII wt
RETAIL wt
Threshold
Train Hard-Day %
Test Hard-Day %
Full Hard-Day %
Full Win Rate %
-0.40
+0.80
-0.70
-0.45
0.01
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.70
-0.45
0.03
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.70
-0.45
0.05
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.65
-0.50
0.01
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.65
-0.50
0.03
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.65
-0.50
0.05
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.60
-0.55
0.01
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.60
-0.55
0.03
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.60
-0.50
0.01
64.4%
72.2%
66.2%
85.4%
-0.40
+0.80
-0.60
-0.50
0.03
64.4%
72.2%
66.2%
85.4%
Top candidates ranked by train-set hard-day accuracy (days that moved >0.6%), refined with a finer grid around each coarse-search seed. Win Rate column is the raw option-seller payout rate, shown for reference only. Sorted best-first.
Next-Day Option Selling Strategy
Built from a LIVE Fyers option chain fetch (real strikes & premiums), strikes chosen for ≥1:1 reward-to-risk — meant to be generated around 3:15pm for the next trading session
Plan unavailableNo trade
Live option chain unreachable (Fyers auth expired or market data unavailable).
Opposite trade-off to the spread above — capped, known max loss but UNCAPPED upside. Deployed the trading day after every weekly (Tuesday) expiry, held through the next.
Alternative unavailableNo trade
Could not resolve the weekly expiry cycle (option chain or Nifty history unreachable).
Active Trade Monitor — 15-Min Exit Signal
Refresh this dashboard every ~15 minutes intraday for an updated signal
No active tradeNot tracked
Once you've placed the trade above with your broker, click below to start live exit-signal tracking. Until then, there is nothing to monitor.
Read before using
Strikes and premiums above are pulled live from the Fyers option chain, not the synthetic
placeholder table used elsewhere on this dashboard. Lot size is hardcoded at 65 — verify
against the current NSE Nifty lot size before sizing a real trade. Strikes here are chosen to
hit ≥1:1 reward-to-risk, which is a different (harder, lower-probability) bet than the
wider, higher-win-rate structure the FII/DII backtest tabs describe — see the "implied
cushion" card above for this specific structure's own backtested numbers. This is a systematic
tool output, not investment advice; size and manage risk according to your own judgment.