Thanks for shedding some light on the logics.

Since you meant “changing patterns of placement on the Book” to detect the MM behaviors, do you use machine learning for pattern recognition?

From my understanding, you would look throughout the entire DOM and detect for patterns of limit (could be small or large) laid out on the book and how those limit blocks pop up in certain formation according to the changes of the price?

And regarding today (August 1, 2023), NQ made some brutal moves up and then down 300 points. If you have been caught in a long mega position from the get go, and then it makes a move like that, would you have to cut loss at least 50% when NQ rejected some level, turned down then break some good support and shift the structure to bearish?

I could link to the scenario like today could happen again and again because the mega position is kinda bulky to change the direction quickly. How would you handle days like today on NQ? (Assume with a 300K APEX account is in play)

hyperpro123456 

Thanks for shedding some light on the logics. Since you meant “changing patterns of placement on the Book” to detect the MM behaviors, do you use machine learning for pattern recognition? From my understanding, you would look throughout the entire DOM and detect for patterns of limit (could be small or large) laid out on the book and how those limit blocks pop up in certain formation according to the changes of the price? And regarding today (August 1, 2023), NQ made some brutal moves up and then down 300 points. If you have been caught in a long mega position from the get go, and then it makes a move like that, would you have to cut loss at least 50% when NQ rejected some level, turned down then break some good support and shift the structure to bearish? I could link to the scenario like today could happen again and again because the mega position is kinda bulky to change the direction quickly. How would you handle days like today on NQ? (Assume with a 300K APEX account is in play)

Interesting questions.
No, I haven't used any machine learning concepts.
In general, try the simplest things first. We're interested to find when MM
is showing "interest" in a range of Price levels; and we can hypothesize that
some simple measurements might serve as a "proxy" for that level of
interest.

A given Price level on Market Depth is associated with an aggregated size
from multiple sources. One of these sources is assumed to be Market Maker.
If it is true that MM wants to establish quoted size in such a way that She is
the first in the execution queue at that Price level, then perhaps we can
look at the Minimum Size quoted at the Price level as a proxy. Others may
come and go with their quotes at a Price level; but is MM is serious about
maintaining "best price" positioning at the Price level, then She will hold
her quotes. So, perhaps looking for the Minimum observed size over some
suitable period of time, e.g. 10 seconds or so; could be a good proxy for
a level of MM Size indicating an "interest" in that Price level.

If we do this for a corresponding Range of offsets from the current Market
Price; translating relative Price distance from the Market into fetches of
the relevant Price-specific Analyzer object, and finding the result of this
query; then we can take the Summation of corresponding Market Relative
Tier Ranges on both sides of the market and compare them. Our assumption
would be that Price is predicted to move in the direction of Larger Quoted
Size (presumably by MM) for Tiers which are more distant from the current
Market price.

Very near the market, we may see a Push Size effect tightly against the
Inside so that Price would be predicted to be "pushed" away from the
Size.... So the interpretation of these quoted sizes over a range of Market
relative Tiers changes as we approach the "hot zone" just a few ticks outside
of the Market where Trading transactions are actually taking place.

Quotes are not transactions until the Market reaches its Price Level; but MM
will be actively working to position Quotes early, and somewhat distant
from the market; in Preparation for moving the Market to those Price
levels. This is WHY a measure of this sort is a Leading Indicator. MM first
places Quotes, and then Later moves the Market Price in effect to "hit" those
quotes and generate Trade transactions.....

IN GENERAL, you must consider that Technical measurements can be
"swamped" by News Events, or by the Market Open process. So it may be
best to avoid the first half hour of activity, and News Events; in your
trading, for the highest level of Control through Technical Analysis such
as this....... Caveat Emptor?.... Micro Scalping works best in quiescent
markets when others have no idea how to trade..... that's the Edge.

IN GENERAL, operate as though you will never need to Stop Out. This
involves being able to tolerate Price Adversity, perhaps using price-staggered
Micro positions. However, to avoid these situations of uncontrolled price
adversity, WE SHOULD NOT BE GUESSING what the Trend is..... In theory,
our system tells us what the Trend is; regardless of whether Price "seems"
to have an opposite direction. If our Analytics are correct; then we can
have a reasonable level of confidence that we are not "throwing Good money
after Bad" by continuing to bet against a Price movement that may never
turn in our favor !!! So, DON'T GUESS THE TREND; rather, measure it
using information which underlies Price movements; but don't try to
predict price movements solely from price movements themselves... ?? !!

TRADING MUST NOT BE GAMBLING, and too many traders are heavily
gambling on outcomes, often with stakes which are too high and are
uncontrolled. Unless you have reliable Technical Analysis based upon
provable Predictive measurements (such as I've suggested using the DOM
or some other non-Price indicator); then really you should not be in
the business of trading, IMHO in my opinion.....

You need computer assistance in Order Management, so that you can
calculate in some LIFO manner, your Break Even (NET of expenses)
and have a method to manage Orders and Positions in a semi-automated
fashion. If you get into trouble "defending" the indefensible; then you
will need to reduce your position quickly or liquidate; rather than try to
fight to break even; but making that determination in the "heat of the
moment" is not going to be easy in any case... Plan never to have
to Stop Out; but realize that you may be pushed to do so; and Cut
your Losses to fight again in a few minutes when things settle down?...

Also in my opinion; non-technical traders, or traders using conventional
indicators based solely on Price; and who think they can "Learn to Trade"
are doomed to have a very depressing and unpredictable career...
It's not for me anyway, and I don't recommend it for anyone.....
Basically, you can't bring a knife to a gun fight; and unless you have
a substantial provable "edge" you should NOT be trading... But all of
you know that already, I hope !!!

Good Trading !
hyperscalper

Understood. Thank you very much.

The price level analyzer module of your app has the ability to ask, "Are there any potential order blocks of MM in the last X seconds?" This goes beyond merely identifying a simple, general big-sized order in the book. By doing that, I can understand why you need to timestamp pieces of information with atomic precision. Now, things are interconnected.

Moreover, that "pattern" which emerges as the MM's plan serves as a guide for the trend, where the price will eventually be "pushed" to a specific location as part of the MM’s strategy. Price indicators are inherently lagging, and MM only uses them to trap retail traders.

The key is to identify the MM's plan from the DOM and piggyback on it using high-speed analysis and bot-assisted order placement.

You mentioned that the micro-trend spotted with this method is useful for the next 3-5 minutes. Is this the duration during which a mega-position position lasts? Additionally, do you enter a new mega-position during the day whenever a micro-trend is detected?

I have no doubt about utilizing technology and automation (mostly semi-automatic) to assist with precise and repetitive tasks. This approach also eliminates the burden of emotion. This is especially true when the market turns against our position; a quick and non-emotional reaction is not only helpful but also required.

Thank you for sharing your ideas and taking the time to address my questions.

“ Very near the market, we may see a Push Size effect tightly against the
Inside so that Price would be predicted to be "pushed" away from the
Size.... So the interpretation of these quoted sizes over a range of Market
relative Tiers changes as we approach the "hot zone" just a few ticks outside
of the Market where Trading transactions are actually taking place.”

Could you please ellaborate a bit more on this part. Are you depicting a pullback against the direction of the final target of MM?

hyperpro123456 

“ Very near the market, we may see a Push Size effect tightly against the Inside so that Price would be predicted to be "pushed" away from the Size.... So the interpretation of these quoted sizes over a range of Market relative Tiers changes as we approach the "hot zone" just a few ticks outside of the Market where Trading transactions are actually taking place.” Could you please ellaborate a bit more on this part. Are you depicting a pullback against the direction of the final target of MM?

OK, well firstly we accept that Market Maker moves Prices; and then
consider that if an UP trend is in effect, then anything which MM Buys
from a Retail seller will be a "no risk" purchase; since MM will inevitably
be pushing the Price higher. So stacking size in this case against the
BID adjacent tiers would be expected, since MM has "decided" the
Price will rise, so any purchases will be "muscled" to a higher Price
as I said earlier.

Elsewhere I suggested that MM's behavior is less obvious than other
large Retail operations. We would not expect to see "obvious" big lot
Limit placements to originate from MM. Instead, we'd see an
increase in MM's distribution of Size near the BID, which is a
"proxy" for "greater interest" in transacting with the Retail
population since, again, anything a Retail player sells is going
to be bought in an environment of virtually guaranteed higher
price, during the Micro Trend UP in Price.

Now, there will be a need to reduce "noise" in the data, since we
are talking here about the "chaotic" zone near the market, where
"the rubber hits the road" and actual Trades are being made.

There will also be a "scramble" for BID adjacent Limit Price placements
so perhaps just looking at overall Activity (updates per second)
at the Inside Tier would also show MM "jockeying" for Best Price
during this UP move in Price, in competition with the few Retail
players who may also be placing limits at or near the BID.

Specifically on that last point, MM is raising the Price; but How?
Well, obviously by BIDding Higher (pushing the BID) and if She
Bids higher then She'll automatically have "best execution price"
at that new higher BID price level. (Also She'll be easing off
and allowing the ASK side of the market to rise; without placing
as much size on that side of the market during the UP trend,
since Her interest is Buying from; and less so in Selling to the
Retail population. This should be true until the Micro Trend
comes near its end.)

Remember you're capturing Quotes using Price specific analyzers,
but you taking "snapshots" which are Market Tier Relative, by
querying the corresponding Absolute Price capture objects.
At least that's the way I do things.

Hope this helps ! Good Trading !
hyperscalper

Thank you very much for taking the time to explain the ideas and share your experience. I’ve gained a clear understanding of the overall picture.

I’ve begun by working on ‘accounting and risk management’ through a C# application that connects to Sierra Chart in real-time. This is where I currently stand.

For the entry part, I’ve been trying it manually, and so far, it’s been going well. There have been three instances where the market moved unexpectedly since I started, but having a systematic approach to risk management has made a significant difference!

My next step involves testing hypotheses and attempting to identify those order blocks in the DOM using the characteristics you mentioned (order blocks - not in SMC term)

Once again, thank you, hyperscalper!

@hyperscalper,

Does Ninja Trader (NT) still hand you real time MBO from its DOM? Sierra Chart does not provide real real-time data from DOM with MBO because of their agreement with the exchanges.

I'm thinking if NT allows it, it would be better to take data from NT. Or the worst case, I will need to use R | Protocol API directly with Rithmic.

Thanks

hyperscalper 

OK, well firstly we accept that Market Maker moves Prices; and then consider that if an UP trend is in effect, then anything which MM Buys from a Retail seller will be a "no risk" purchase; since MM will inevitably be pushing the Price higher. So stacking size in this case against the BID adjacent tiers would be expected, since MM has "decided" the Price will rise, so any purchases will be "muscled" to a higher Price as I said earlier. Elsewhere I suggested that MM's behavior is less obvious than other large Retail operations. We would not expect to see "obvious" big lot Limit placements to originate from MM. Instead, we'd see an increase in MM's distribution of Size near the BID, which is a "proxy" for "greater interest" in transacting with the Retail population since, again, anything a Retail player sells is going to be bought in an environment of virtually guaranteed higher price, during the Micro Trend UP in Price. Now, there will be a need to reduce "noise" in the data, since we are talking here about the "chaotic" zone near the market, where "the rubber hits the road" and actual Trades are being made. There will also be a "scramble" for BID adjacent Limit Price placements so perhaps just looking at overall Activity (updates per second) at the Inside Tier would also show MM "jockeying" for Best Price during this UP move in Price, in competition with the few Retail players who may also be placing limits at or near the BID. Specifically on that last point, MM is raising the Price; but How? Well, obviously by BIDding Higher (pushing the BID) and if She Bids higher then She'll automatically have "best execution price" at that new higher BID price level. (Also She'll be easing off and allowing the ASK side of the market to rise; without placing as much size on that side of the market during the UP trend, since Her interest is Buying from; and less so in Selling to the Retail population. This should be true until the Micro Trend comes near its end.) Remember you're capturing Quotes using Price specific analyzers, but you taking "snapshots" which are Market Tier Relative, by querying the corresponding Absolute Price capture objects. At least that's the way I do things. Hope this helps ! Good Trading ! hyperscalper

bro keep writing these great posts, i love reading them, also post more videos on your youtube channel, that bot was fantastic