OPEN-SOURCE SCRIPT
Trade Coach-Journal
Every trader knows they should journal. Almost nobody does it, and the ones who do
mostly end up with a spreadsheet of numbers they never open again.
The problem isn't discipline. It's that a journal tells you what happened last month,
and the moment you actually need it is right now — with your finger over the button,
about to take a trade at 2pm on a Thursday after two losses, which is exactly the
combination that has cost you money forty times before.
This puts your journal on the chart and reduces it to a verdict that changes as your
day does. At 9:50am it might say:
GOOD WINDOW
this is when you trade best
By 2:15pm, after a loss, the same panel says:
STOP · DONE TODAY
weak window and you're coming off a loss
Nothing in between those two moments came from the market. It came from you, forty
trades ago, doing the same thing.
THE PART THAT ACTUALLY CHANGES THINGS
Timing patterns are useful. Knowing your afternoons win 30% is worth something. But
you can't decide to make afternoons behave differently.
You can decide to stop chasing.
So every trade can carry tags — words you invent, starting with #. Log a trade you
chased as:
8/17 13:35 S 29140 14:10 29168.75 #fomo
Do that twenty times and the panel stops talking about the clock:
TOP FIX
#fomo trades win 18% (n=22)
→ stop taking #fomo entries
That's a different kind of sentence than "your Tuesdays are weak." It names a habit,
it's yours, and you can change it tomorrow morning. Tag findings outrank timing
findings in the panel for exactly that reason.
Use whatever vocabulary fits how you actually trade — #plan, #revenge, #late, #news,
#tired, #a+. The only requirement is honesty. Putting #plan on a trade you chased
makes the whole thing useless, and nobody sees this but you.
LOGGING A TRADE
One line, in the settings:
8/17 9:45 L 29048 10:15 29096.5 #plan
August 17th, 9:45am, long from 29048, out at 10:15 at 29096.50, planned setup.
The parser tries hard not to make you think about formatting. These four lines are
the same trade:
8/17 9:45 L 29048 10:15 29096.5
8/17, 9:45, L, 29048, 10:15, 29096.5
08-17 0945 LONG 29048 1015 29096.5
2026-08-17 09:45 buy 29048 10:15 29096.5
Dates take 8/17, 08-17, 2026-08-17, 0817 or 20260817. Times take 9:45, 09:45 or 0945.
Side takes L/S, LONG/SHORT or BUY/SELL in any case. Spaces and commas both work. The
year is set once in settings so you're not retyping it. Exit time is optional — leave
it out and you keep every statistic, you just lose the line drawn on the chart. An
exit time earlier than the entry is read as an overnight hold.
Times should match the clock on your chart's time axis.
WHAT YOU SEE
Under the verdict, the panel is deliberately short:
now: pm ▰▰▰▱▱▱▱▱▱▱ 30% n=20
form ●●○○●○○○●○
equity █▇▆▄▅▃▂▁
today 3 trades · -2.1R
Four lines, and the third one is the one that hurts. Your form dots can look fine
while the equity sparkline slides down the page — that's the shape of winning often
and losing big, and it's the most common way a trader who looks profitable isn't.
The footer says the same thing in numbers:
40 trades · 67% win · -0.12R avg
A green win rate sitting next to a red R average is worth more than any entry signal
you'll read this year.
Every statistic carries its sample count, and rows stay grey until they've earned an
opinion. A grey row means the script doesn't know yet — more honest than a confident
percentage built on six trades.
Turn on "show full detail" for the full breakdown: morning against afternoon, long
against short, after-loss, average winner against average loser, worst losing streak,
and every tag ranked by how often you use it.
On the chart itself each trade draws where it happened — a triangle at the entry, a
line to the exit, the R result labeled, tags in the tooltip.
ALERTS
The panel only helps if you're looking at it, and the moments you most need it are
the moments you're not.
Entering a weak window
Daily stop hit
STOP for today
Set them once. Then the coach speaks first, and you don't have to remember to ask.
ON THE NEURAL NETWORK IN GROUP 4
There's a small neural network in the advanced settings that trains on your logged
trades and estimates whether a trade taken under current conditions would win. Most
of the time it says this:
verdict not significant
vs control 58% vs 62%
That second line is the whole reason to trust it. Alongside the real network, the
script trains an identical one on deliberately shuffled labels — a model that cannot
possibly know anything. If the real network can't clearly beat that, its opinion is
suppressed and the panel says so.
A network with this many parameters needs several hundred trades before it can
separate a pattern from a coincidence. It will probably read "not significant" for a
long time, and that's the safeguard working rather than the tool failing.
Nothing in the verdict, TOP FIX, or the pattern tables involves the model. That's all
plain counting, which is why it becomes usable around 20 trades and trustworthy around
40 — while the network is still deciding whether it knows anything at all.
IF SOMETHING LOOKS WRONG
The panel tells you which of three things went wrong rather than making you guess:
⚠ 3 lines unreadable (line 12)
Line 12 didn't parse. Usually a missing price, a typo in the date, or a side it
didn't recognize.
⚠ 8 trades off-chart — check timezone
The timestamps don't land on a loaded bar. Either scroll left for more history, or
your times aren't in the exchange's timezone.
⚠ 5 trades too early on this chart
The script needs 220 bars of warmup before it can read market context. Load more
history or move to a higher timeframe.
If it reads "0 of 40," everything was rejected or fell outside the chart. Work
through those three in order.
This is a review tool. It reports patterns in trades you've already taken, generates
no entry signals, and makes no claim about future results. What it shows you is your
own history.
mostly end up with a spreadsheet of numbers they never open again.
The problem isn't discipline. It's that a journal tells you what happened last month,
and the moment you actually need it is right now — with your finger over the button,
about to take a trade at 2pm on a Thursday after two losses, which is exactly the
combination that has cost you money forty times before.
This puts your journal on the chart and reduces it to a verdict that changes as your
day does. At 9:50am it might say:
GOOD WINDOW
this is when you trade best
By 2:15pm, after a loss, the same panel says:
STOP · DONE TODAY
weak window and you're coming off a loss
Nothing in between those two moments came from the market. It came from you, forty
trades ago, doing the same thing.
THE PART THAT ACTUALLY CHANGES THINGS
Timing patterns are useful. Knowing your afternoons win 30% is worth something. But
you can't decide to make afternoons behave differently.
You can decide to stop chasing.
So every trade can carry tags — words you invent, starting with #. Log a trade you
chased as:
8/17 13:35 S 29140 14:10 29168.75 #fomo
Do that twenty times and the panel stops talking about the clock:
TOP FIX
#fomo trades win 18% (n=22)
→ stop taking #fomo entries
That's a different kind of sentence than "your Tuesdays are weak." It names a habit,
it's yours, and you can change it tomorrow morning. Tag findings outrank timing
findings in the panel for exactly that reason.
Use whatever vocabulary fits how you actually trade — #plan, #revenge, #late, #news,
#tired, #a+. The only requirement is honesty. Putting #plan on a trade you chased
makes the whole thing useless, and nobody sees this but you.
LOGGING A TRADE
One line, in the settings:
8/17 9:45 L 29048 10:15 29096.5 #plan
August 17th, 9:45am, long from 29048, out at 10:15 at 29096.50, planned setup.
The parser tries hard not to make you think about formatting. These four lines are
the same trade:
8/17 9:45 L 29048 10:15 29096.5
8/17, 9:45, L, 29048, 10:15, 29096.5
08-17 0945 LONG 29048 1015 29096.5
2026-08-17 09:45 buy 29048 10:15 29096.5
Dates take 8/17, 08-17, 2026-08-17, 0817 or 20260817. Times take 9:45, 09:45 or 0945.
Side takes L/S, LONG/SHORT or BUY/SELL in any case. Spaces and commas both work. The
year is set once in settings so you're not retyping it. Exit time is optional — leave
it out and you keep every statistic, you just lose the line drawn on the chart. An
exit time earlier than the entry is read as an overnight hold.
Times should match the clock on your chart's time axis.
WHAT YOU SEE
Under the verdict, the panel is deliberately short:
now: pm ▰▰▰▱▱▱▱▱▱▱ 30% n=20
form ●●○○●○○○●○
equity █▇▆▄▅▃▂▁
today 3 trades · -2.1R
Four lines, and the third one is the one that hurts. Your form dots can look fine
while the equity sparkline slides down the page — that's the shape of winning often
and losing big, and it's the most common way a trader who looks profitable isn't.
The footer says the same thing in numbers:
40 trades · 67% win · -0.12R avg
A green win rate sitting next to a red R average is worth more than any entry signal
you'll read this year.
Every statistic carries its sample count, and rows stay grey until they've earned an
opinion. A grey row means the script doesn't know yet — more honest than a confident
percentage built on six trades.
Turn on "show full detail" for the full breakdown: morning against afternoon, long
against short, after-loss, average winner against average loser, worst losing streak,
and every tag ranked by how often you use it.
On the chart itself each trade draws where it happened — a triangle at the entry, a
line to the exit, the R result labeled, tags in the tooltip.
ALERTS
The panel only helps if you're looking at it, and the moments you most need it are
the moments you're not.
Entering a weak window
Daily stop hit
STOP for today
Set them once. Then the coach speaks first, and you don't have to remember to ask.
ON THE NEURAL NETWORK IN GROUP 4
There's a small neural network in the advanced settings that trains on your logged
trades and estimates whether a trade taken under current conditions would win. Most
of the time it says this:
verdict not significant
vs control 58% vs 62%
That second line is the whole reason to trust it. Alongside the real network, the
script trains an identical one on deliberately shuffled labels — a model that cannot
possibly know anything. If the real network can't clearly beat that, its opinion is
suppressed and the panel says so.
A network with this many parameters needs several hundred trades before it can
separate a pattern from a coincidence. It will probably read "not significant" for a
long time, and that's the safeguard working rather than the tool failing.
Nothing in the verdict, TOP FIX, or the pattern tables involves the model. That's all
plain counting, which is why it becomes usable around 20 trades and trustworthy around
40 — while the network is still deciding whether it knows anything at all.
IF SOMETHING LOOKS WRONG
The panel tells you which of three things went wrong rather than making you guess:
⚠ 3 lines unreadable (line 12)
Line 12 didn't parse. Usually a missing price, a typo in the date, or a side it
didn't recognize.
⚠ 8 trades off-chart — check timezone
The timestamps don't land on a loaded bar. Either scroll left for more history, or
your times aren't in the exchange's timezone.
⚠ 5 trades too early on this chart
The script needs 220 bars of warmup before it can read market context. Load more
history or move to a higher timeframe.
If it reads "0 of 40," everything was rejected or fell outside the chart. Work
through those three in order.
This is a review tool. It reports patterns in trades you've already taken, generates
no entry signals, and makes no claim about future results. What it shows you is your
own history.
Açık kaynak kodlu komut dosyası
Gerçek TradingView ruhuyla, bu komut dosyasının mimarı, yatırımcıların işlevselliğini inceleyip doğrulayabilmesi için onu açık kaynaklı hale getirdi. Yazarı tebrik ederiz! Ücretsiz olarak kullanabilseniz de, kodu yeniden yayınlamanın Topluluk Kurallarımıza tabi olduğunu unutmayın.
Feragatname
Bilgiler ve yayınlar, TradingView tarafından sağlanan veya onaylanan finansal, yatırım, alım satım veya diğer türden tavsiye veya öneriler anlamına gelmez ve teşkil etmez. Kullanım Koşulları bölümünde daha fazlasını okuyun.
Açık kaynak kodlu komut dosyası
Gerçek TradingView ruhuyla, bu komut dosyasının mimarı, yatırımcıların işlevselliğini inceleyip doğrulayabilmesi için onu açık kaynaklı hale getirdi. Yazarı tebrik ederiz! Ücretsiz olarak kullanabilseniz de, kodu yeniden yayınlamanın Topluluk Kurallarımıza tabi olduğunu unutmayın.
Feragatname
Bilgiler ve yayınlar, TradingView tarafından sağlanan veya onaylanan finansal, yatırım, alım satım veya diğer türden tavsiye veya öneriler anlamına gelmez ve teşkil etmez. Kullanım Koşulları bölümünde daha fazlasını okuyun.