Flip of a coin..but

[hilmy83]

Question, so how does one play the mean-reversion game with scale-in on a one-minute chart?

Seems even with code it is an awful lot to get right to make it work.

It's just using initial range multiple. You would have to do it in high volatility like the cash open so you can have wider range.

I would say it's less about mean reversion, but more about "retracement play". In a 50/50 world, you will get a series of consecutive head/tails that represents a move.

Trend traders are betting on low probable high consecutive hits (i.e move in one direction). Mean reversion is just betting on when the flip happens (but still need that consecutive moves to profit after, just less)

Look at table and just pretend consecutive loss as a "one direction candle/bar".

Technically this would be considered "price action" trading. But instead of your fancy patterns or trendlines, it's based on simple fact that price will retrace at some point...most of the time.

The otehr "less of the time", you just need to control with size and b/e stops. Getting out when you get caught is key, not to get married to a position.

 

[SunTrader]

It's just using initial range multiple. You would have to do it in high volatility like the cash open so you can have wider range.
...

Yes but mean-reversion means back and forth/chop/trend-less. Scale-in means trend for the most part.

 

[spy]

You can't generate profits from purely random price series, not even in the absence of execution related cost. That is mathematically simply not possible.

@rb7 never mentioned a "price series", just

your 3 random graphs.

Which makes me think you've never played poker against people for money; random series but lots of money to be made nonetheless.

 

[hilmy83]

Yes but mean-reversion means back and forth/chop/trend-less. Scale-in means trend for the most part.

Each scale in is a 50/50 shot it will work out.

Most trend traders i see over the years use b/e stop and scale out. They're basically scalping a move most of the time and just leaving couple of positions to run.

Nobody really goes all in and out on a trend bet.

 

[farmerjohn1324]

@rb7 never mentioned a "price series", just



Which makes me think you've never played poker against people for money; random series but lots of money to be made nonetheless.

Playing poker against other people isn't random because it depends on the skill level of all players involved. There's more to it than just getting lucky with which cards you are dealt.

 

[spy]

Playing poker against other people isn't random because it depends on the skill level of all players involved. There's more to it than just getting lucky with which cards you are dealt.

True, it seems we agree then... because the order of the cards is random; just like the graphs. And, the person/people you trade/play against may not have comparable skillz

 

[rb7]

@rb7 never mentioned a "price series", just



Which makes me think you've never played poker against people for money; random series but lots of money to be made nonetheless.

FWIW, I'm playing poker against people for money on a regular basis, and since many years (and not for dimes and nickels!!).

Playing poker and day trading are closer than what people usually think.
In both cases, we're putting money based on probabilities of forecast of an outcome.

 

[M.W.]

What?

For starters, a graph with temporal axis is a visual representation of a time series or price series. You can't reliably predict a purely random generated time series with any statistical significance, meaning you can't make money from many forecasts or predictions based on a randomly generated price series.

Next, while there are random components in poker (such as the card shuffle) poker is anything but random at all, the comparison makes absolutely no sense here. You are playing against humans (or computers that have been trained based on human imposed rules and policies) and human emotions, most basic fear and greed. Nothing of this relates to purely randomly generated time series.

@rb7 never mentioned a "price series", just



Which makes me think you've never played poker against people for money; random series but lots of money to be made nonetheless.

 

[hilmy83]

FWIW, I'm playing poker against people for money on a regular basis, and since many years (and not for dimes and nickels!!).

Playing poker and day trading are closer than what people usually think.
In both cases, we're putting money based on probabilities of forecast of an outcome.

I think there's more variance in poker then it is in trading though. Your hand odds are not 50/50. Even a pair is like what 40% odds?

It's mostly a game of telling a story, maybe 10% of it is actual game.

 

[M.W.]

For fucks sake, how old are you dude? First of all there was nothing agreeable between your two statements, @farmerjohn1324 correctly pointed out that the random component in winning poker matches is minimal over many games. You claimed poker to be a great analogy for a randomly generated time series. Shrugs...

True, it seems we agree then... because the order of the cards is random; just like the graphs.