Adjust the parameters below and watch Squintz score the signal in real time. This is the exact algorithm behind every alert.
The GRIP algorithm scores every moving stock on four dimensions: Growth (velocity of the move), Range (where price closed in its intraday range), Intensity (volume participation), and Probability (ML-scored continuation odds). Each dimension scores 0–25. Combined, they produce a GRIP total of 0–100. Signals above 75 with all 7 conviction gates aligned trigger a high-conviction alert. Read the full methodology at how GRIP works, or see the live results on the Signal Record.
Try a preset:
How much did it move, and how fast?
Price Change8.0%
Minutes to get there30 min
Velocity: 0.27%/min
Where did price close in today's range?
Strong close — momentum holding
How much volume behind this move?
When did this move happen?
9:30 AM10:00 AM ET4:00 PM
Strong window
HIGH
GRIP Total / 100
G — Growth
17/25
Velocity of the move
R — Range
21/25
Where price closed in range
I — Intensity
19/25
Volume participation
P — Probability
23/25
ML-scored continuation odds
🔥 HIGH CONVICTION — All 7 gates pass. This signal would trigger an SMS alert.
✓GRIP total ≥ 75
✓G score ≥ 17
✓I score ≥ 12
✓Move ≥ 8%
✓Velocity ≥ 0.20%/min
✓Time: 9:30 – 11:30 AM ET
✓Move happened in ≤ 45 minutes
Why These Four Dimensions?
G
Growth — velocity, not magnitude
Growth measures velocity, not just magnitude. A stock moving 8% in 20 minutes (0.40%/min) scores dramatically higher than the same 8% move spread over 90 minutes (0.09%/min). Speed tells you a move is still happening — not that it already happened.
R
Range — where it closed matters
Range measures where price closed within today's high-low band. A close at the session high means buyers are still in control at the end of the window — momentum is intact. A close near the low means sellers pushed back.
I
Intensity — volume separates signal from noise
Intensity measures volume. Retail traders move small volume. Institutional money moves markets. An 8% move on 5M shares is structurally different from the same move on 50K shares. Squintz scores volume to separate signal from noise.
P
Probability — the ML layer
Probability uses a machine learning model trained on 22,000+ real trade outcomes to score the likelihood the move continues. Early session moves score higher because they have more time to run. The model retrains every Friday.
Ready to receive signals like this?
Every GRIP-qualified alert lands in your inbox — and on Pro, your phone — the moment all 7 gates pass.