Moving Averages (Advanced Charting)

[jamesg46]

For those of you who use Advanced Charting, do you have your moving averages set based on personal preference for different strategies or is there some well respected moving averages that apply to pre race markets.

[ShaunWhite]

The only thing I know about MAs is that (some) people use maybe 3 different timescales (say 10s, 30s and 1min) and look for some sort of pattern or correlation. A bit like the way they use vwap in certain financials over periods like 1day, 1week and 1 month. And then it's trial and error I'm afraid. I posted a question on here a couple of years ago asking how to measure the accuracy/effectiveness of various indicators after I developed a new one that looked promising visually, but I didn't get any useful public replies, just the usual try it and see what happens. People on here either don't know, or won't say how to develop and test market indicators empirically.

[jamesg46]

The only thing I know about MAs is that (some) people use maybe 3 different timescales (say 10s, 30s and 1min) and look for some sort of pattern or correlation. A bit like the way they use vwap in certain financials over periods like 1day, 1week and 1 month. And then it's trial and error I'm afraid. I posted a question on here a couple of years ago asking how to measure the accuracy/effectiveness of various indicators after I developed a new one that looked promising visually, but I didn't get any useful public replies, just the usual try it and see what happens. People on here either don't know, or won't say how to develop and test market indicators empirically.

It's rather tricky, people in financials typically have some well respected moving averages (i believe.) I've used a 60 second width and a 5 second interval, using a 60, 30 & 5 second expo ma. I'll just have a play around when the racing starts back up... thanks for the reply Shaun.

[jamesg46]

I wonder if because we have pretty much in the majority of races a 10 min trading window if creating multiple charts would work better. I'm sure I've watched a video where Peter mentions he uses 6 charts, i had always assumed it was for 6 different strategies but like in financials where they have the Daily & intra day charts it could be an idea to have a 10 min chart followed by by 2,3 or more smaller time period charts. My issue is, I dont want to fall into the trap of screen junk.

[ShaunWhite]

The problem with indicators is that they always look great in hindsight because they're based on what's previously happened. I'm yet to find any that tell you much about what's going to happen beyond the righthand edge of the chart.

This is why I was so interested in obtaining an actual measurement of their effectiveness, that's possible but I couldn't do it with my basic maths. It's actually one of the reasons that my Udemy course of choice for this year is on calculus, the mathematics of change. It's a bit of a drag atm though, I'm still on the chapter about the volume of an inflating balloon and the chapter on pricing models seems a long way off.

[jamesg46]

The problem with indicators is that they always look great in hindsight because they're based on what's previously happened. I'm yet to find any that tell you much about what's going to happen beyond the righthand edge of the chart.

This is why I was so interested in obtaining an actual measurement of their effectiveness, that's possible but I couldn't do it with my basic maths. It's actually one of the reasons that my Udemy course of choice for this year is on calculus, the mathematics of change. It's a bit of a drag atm though, I'm still on the chapter about the volume of an inflating balloon and the chapter on pricing models seems a long way off.

For me it's not about having them to predict what will happen, it's more about visually seeing where the psychological points are within the market, moving averages as well as round price points create good points of resistance, I just cant see how there would be a respectable moving average. I've started off by creating a 5 min chart & using a MA for 5, 4, 3, 2 &1 mins. Same for a 3 min chart, using a MA for 3,2 & 1 minutes. And finally a 1 min chart using a 60 second, 30 second & 15 second MA. How useful that'll be I have no idea but I guess time will tell.

I started a coding course on Udemy, great platform but like most other things it goes straight over my head, I'm choosing to digest it in crumb size pieces, good luck with yours, I'm sure you'll do well at it.

[ShaunWhite]

I learnt C# on Udemy.... slowly. Almost too many courses there and finding the right one is tricky. The problem is drifting off through things that seem obvious then realising later it wasn't obvious and you don't know it when you actually need it. Same as school really but fewer detentions. It's boring sometimes too and I really don't want to do it, but every job has times like that so you just have to suck it up.

Good luck with what you're doing, it's an interesting field. I don't rule out pure observation as a measure, your brain is phenomenal at detecting fuzzy patterns even AI struggles with so if can avoid things like confirmation bias it's definatly worth a month or two of investigation.