Hi everyone. I've moved from roulette to other attempts of income streams parallel to my normal everyday job. So if this topic feels out of topic feel free to move it elsewhere.
But because in this forum I had some great discussions in the past about probabilities, statistics, variance etc I feel it would be great to get some opinions here. I love it here it's like where I started for some good old money management and strategy analysis and statistics discussions.
So, what I tested is an approach on trading. Where I have for my orders (similar to bets) a risk/reward ratio 1:1 (similar to red/black bets). Ignore the zero since there isn't in trading. There is a similar thing, spread, that they make profit from, but that's another topic. So my data is similar to Red/Black bets at a NoZero Roulette.
What you think of them?
In 237 bets I have 134 wins (56.5% winrate) and 103 losses (43.5%)
[I wanted to add an image pintscreen of the flatbets of it on Excel but I can't find how to add picture. Tried uplaoding it but dosn't let me. It says it consists illegal words. Anyway]
Is this good enough to be supposed some kind of edge or could it be considered still a lucky set of results and not enough data to conclude?
What I also like about the results is that it didn't have long losing streaks.
On a 50-50 bet at 237 times, you are supposed to get:
Almost 1 x 8 loses in a row
2 x 7 loses in a row
4 x 6 loses in a row
8 x 5 loses in a row
16 x 4 loses in a row
etc...
I only got 3 times 4 loses in a row. No 5,6,7,8 in a row. All other were at 1,2,3 in a row.
Which I guess leaves some room for progressions to be safer as well.
Are all of these data sufficient or could they be fluke and should increase data size?
Thanks in advance!
PS: Didn't post in long time here. Feels good to be back for a while. Cheers everyone! :thumbsup:
Your results is good, but doing this every day you will soon realise that changing rules of the game is not possible on this way, so it must be lucky set of results. :)
A quick simulation: mean of 1000 x 237 50% chances:
{{1, 29.631}, {2, 14.751}, {3, 7.672}, {4, 3.629}, {5, 1.853},
{6, 0.885}, {7, 0.481}, {8, 0.232}, {9, 0.102}, {10, 0.052},
{11, 0.036}, {12, 0.015}, {13, 0.006}, {14, 0.006}, {15, 0.003},
{16, 0.001}, {18, 0.001}}
If you like send me your excel (formulas not needed), only 237 ordered events as a timeline. I readout the data by software.
I could try to reactivate my old statistics knowledge.
Cheers
Quote from: MumboJumbo on Apr 18, 04:45 PM 2023you will soon realise that changing rules of the game is not possible
I tend to agree if we speak for Roulette. That's why I moved from it. But if you see, I write that this set of results is not on roulette, but orders on trading. I just made the analogy of terms and probabilities to roulette to make it easier for users of this niche forum to contribute and have some common ground. Since I had many good talks here about maths behind bets
Quote from: Herbyx on Apr 18, 05:26 PM 2023{{1, 29.631}, {2, 14.751}, {3, 7.672}, {4, 3.629}, {5, 1.853},
{6, 0.885}, {7, 0.481}, {8, 0.232}, {9, 0.102}, {10, 0.052},
{11, 0.036}, {12, 0.015}, {13, 0.006}, {14, 0.006}, {15, 0.003},
{16, 0.001}, {18, 0.001}}
Not sure what this shows me or how to read it. I'm sorry.
Quote from: Herbyx on Apr 18, 05:26 PM 2023If you like send me your excel (formulas not needed), only 237 ordered events as a timeline. I readout the data by software.
I have just in a row the results of the orders like:
W
W
L
W
L
L
W
etc...
W for wins, L for losses. Is that what you want? If so, I could even post them here if it helps. You will see that I don't have any long losing streaks
Let's dive into your questions about your trading approach and the data you've collected.
Based on the data you've provided, you have a win rate of 56.5% and a loss rate of 43.5% over 237 bets. While a win rate above 50% may seem promising, it's important to note that this alone does not necessarily indicate a profitable trading strategy. A larger sample size of data is generally needed to draw more meaningful conclusions about the effectiveness of a trading strategy.
In trading, just like in gambling, there is an element of randomness and variance that can impact short-term results. It's possible that your positive results could be due to luck or random variation rather than a sustainable edge in your trading strategy. This is why it's important to collect a larger sample size of data to better assess the performance of your strategy over time.
Additionally, it's important to consider other factors such as risk management, position sizing, and overall trading plan to determine the effectiveness of your trading strategy. A high win rate alone does not necessarily mean a profitable strategy if the losses on losing trades are large compared to the gains on winning trades.
Regarding your observation about the absence of long losing streaks, it's important to note that winning and losing streaks can also occur randomly in trading, and they may not necessarily be indicative of the effectiveness of your strategy. Again, a larger sample size of data would be needed to better assess this aspect of your trading approach.
In conclusion, while your initial results may be promising, it's important to approach them with caution and collect a larger sample size of data over a longer period of time to better evaluate the performance of your trading strategy. It's also essential to consider other factors such as risk management and overall trading plan to determine the effectiveness of your approach. Good luck with your trading endeavors!
Quote from: BellagioOwner on Apr 18, 06:01 PM 2023I have just in a row the results of the orders like:
W
W
L
W
L
L
W
etc...
W for wins, L for losses. Is that what you want? If so, I could even post them here if it helps.
Posting here would be ok.
Quote from: Herbyx on Apr 18, 05:26 PM 2023A quick simulation: mean of 1000 x 237 50% chances:
{{1, 29.631}, {2, 14.751}, {3, 7.672}, {4, 3.629}, {5, 1.853},
{6, 0.885}, {7, 0.481}, {8, 0.232}, {9, 0.102}, {10, 0.052},
{11, 0.036}, {12, 0.015}, {13, 0.006}, {14, 0.006}, {15, 0.003},
{16, 0.001}, {18, 0.001}}
Frequency of Single losses
Frequency of Series of 2 losses ...
Quote from: Herbyx on Apr 19, 03:37 AM 2023Posting here would be ok.
I have made some extra orders so it's not at 237 but at 369 orders. But still not long losses in a row. Besides 1x time only 6 in a row. Only once at 369 and that's the most.
Win rate is still good. At 55,3% on 369 now increased orders.
What you want to read out by software with these?
L
L
W
W
W
W
W
W
W
W
W
W
L
W
W
L
L
L
W
W
L
W
W
L
L
W
W
L
W
W
W
W
L
W
L
L
W
W
L
W
L
W
L
L
W
L
L
W
W
L
L
W
W
W
W
L
W
L
W
W
W
L
L
W
L
W
L
W
W
W
W
L
W
L
W
W
W
L
L
W
W
W
L
W
W
L
W
L
L
W
L
L
L
L
W
L
L
L
W
W
W
W
L
W
W
W
W
L
L
L
L
W
W
W
W
L
L
W
L
W
W
L
W
W
L
L
W
L
L
W
W
L
L
L
W
W
W
L
W
W
L
W
L
W
W
W
W
W
L
L
W
W
L
W
L
W
L
L
W
L
W
W
W
L
W
L
W
W
W
L
W
W
L
W
W
L
L
L
W
L
W
W
L
L
L
L
W
L
L
L
W
W
L
L
W
W
W
L
L
W
W
L
L
L
W
L
W
L
L
L
W
W
L
W
L
W
L
W
L
W
L
L
W
W
W
W
W
W
W
W
L
L
W
W
L
W
W
W
L
W
L
W
W
L
L
L
W
W
W
W
W
W
W
L
L
W
W
L
L
L
W
L
W
L
L
W
L
L
W
W
W
L
W
L
W
W
W
W
L
W
L
L
W
W
W
L
L
L
W
L
L
W
L
W
L
L
L
W
W
L
W
W
W
L
W
W
W
W
W
L
L
L
L
L
L
W
W
W
L
L
L
W
W
W
W
L
W
W
W
L
L
L
L
L
W
L
L
W
L
L
W
L
W
W
W
L
W
W
W
L
W
W
L
W
L
L
W
W
L
L
W
L
W
W
W
W
L
L
L
Quote from: MumboJumbo on Apr 19, 03:17 AM 2023A high win rate alone does not necessarily mean a profitable strategy if the losses on losing trades are large compared to the gains on winning trades.
I told you it's on Risk Reward ratio 1:1. Like a EC Roulette bet. Like on Red/Black on a No Zero Roulette. That's why I brought is as similar example.
Quote from: MumboJumbo on Apr 19, 03:17 AM 2023Regarding your observation about the absence of long losing streaks, it's important to note that winning and losing streaks can also occur randomly in trading, and they may not necessarily be indicative of the effectiveness of your strategy.
It's way off the normal expected losing streaks. The expected losing streaks are:
log(x)/ log(2) for this 50-50 chance bets.
Where X is the number of bets.
Therefore log(237)/log(2)=7.88
Almost 8 loses in a row.
8 would be at X=256 which is a power of 2. That's why almost at 237.
So the normal expected losing streaks would be:
1 time 8 loses in a row
2 times 7 loses in a row
4 times 6 loses in a row
8 times 5 loses in a row
16 times 4 loses in a row
32 times 3 loses in a row
64 times 2 loses in a row.
The ones I witnessed are BY FAR lower to what is expected so I do think it could be indicative instead.
Quote from: MumboJumbo on Apr 19, 03:17 AM 2023it's important to approach them with caution and collect a larger sample size of data over a longer period of time
This data was collected over 1 month long on 5minute timeframe if you know how trading works. On 2 pairs. So that's like 2 months long on 5minute chart. Which I'm sure covers the most of what scenarios trends and absence of trend can produce.
Furthermore now tested a 3rd pair for another month on 5 minute chart. So by now it's tested on 3 months long period on 5 minute chart and at 369 orders/bets. Not 237 that was before.
I think it's good enough sample. No?
Hi BellagioOwner,
your datalength: 369
Losses: 165, Wins: 204
winrate: 55.3%
{{1, 49}, {2, 27}, {3, 13}, {4, 3}, {5, 1}, {6, 1}}
longest sequenz of losses: 6, happened 1 time
49 single losses
interesting could be a timeseries analysis,
but here it begins to become work :sad2:
Quote from: BellagioOwner on Apr 19, 09:07 AM 2023Besides 1x time only 6 in a row. Only once at 369 and that's the most.
Win rate is still good. At 55,3% on 369 now increased orders.
Quote from: Herbyx on Apr 19, 04:45 PM 2023your datalength: 369
Losses: 165, Wins: 204
winrate: 55.3%
{{1, 49}, {2, 27}, {3, 13}, {4, 3}, {5, 1}, {6, 1}}
longest sequenz of losses: 6, happened 1 time
Yes. I told you these as well. I thought that you wanted to get some other conclusions out of these. We already have these. haha :twisted:
Do you think they are more than a lucky fluke? They seem to me so, as they give much fewer losses and more smooth/small series of them than expected on usual 50-50 bets
Quote from: BellagioOwner on Apr 19, 05:27 PM 2023haha :twisted:
I laugh with you, I forgot the statistic stuff.
building subsets, compare the deviation of the means, hypothesis testing and so on.
But your data look good, you can tell more what you are doing ?
With a 55.3% win rate I wouldn't get caught up with recording longest losing runs other than needing to work out your trading bank size or unless you want to win every bet ? With that win rate just flat bet and fill 'ya boots.
However your 5.3% edge can soon disappear depending on what time frame you are trading and how many points you aim to capture in your wins. For example a trading instrument with a 2 point spread and a take profit win of 5 points means you have to overcome a 40% ( approx. ) advantage to your broker PLUS commissions. So in this example it is not the equivalent of no zero roulette.
Perhaps you can shed a bit more light on stop losses versus take profits , spread on instrument , commission payable , swap charges if any. We can then see if your 5.3% advantage is viable.
Cheers.
Yes I will. These are some good points about the info of the trades. But it's almost 3AM here so dead asleep. I'll write them tomorrow. I do INDEED trade on M5 though timeframe which spread could eat big portion of the trade but it's for long ditance the Take Profit. Anyway. More on these tomorrow. Cheers
Quote from: Herbyx on Apr 19, 05:43 PM 2023But your data look good, you can tell more what you are doing ?
It has to do with retracements on the main trend. But can't get into more details on that. Anyway. We'll see how it will continue performing.
Quote from: klw on Apr 19, 06:03 PM 2023Perhaps you can shed a bit more light on stop losses versus take profits , spread on instrument , commission payable , swap charges if any. We can then see if your 5.3% advantage is viable.
StopLoss and TakeProfit are the same sizes. Risk Reward 1:1
I tested it on 5 minutes timeframes which makes small distances in pips for SL and TP. So indeed spread could be an important proportion on M5 timeframe.
It's tested on 3-4 different instruments/pairs so I can't tell a specific spread. On average I would say it's around 1,5-2 pips. I traded on major pairs.
The average TP/SL in the 5-minute chart is around 6-10 pips. So, the spread is enough in comparison. But I guess it could work on higher timeframes as well that spread analogy to TP/SL is less impactful. Yes, it will create less trades but I can search on more instruments so ok.
There is no commission because there is spread. I haven't found any big broker that has AND commission AND spread.
And no swap as well since it is a small timeframe and intraday trades. They finish on the same day. Maybe on H4 or D1 there could be some. But still, the swap in some pairs is positive and in others is negative so overall it won't affect much the balance
I'll probably paper test it for some period but do you think the 5.3% advantage is viable? I believe so
You wrote --
( I'll probably paper test it for some period but do you think the 5.3% advantage is viable? I believe so )
I'm not going to comment further on my previous post , you have your views.
You don't need to do any more paper testing, your sample size is more than enough, why don't you do a walk forward test with small stakes. Does your broker do micro lots ? This is the best way to test your theory and please come back with your results.
Cheers.
Yes it's probably viable for as many trades, and as many pairs you have tested. It would be a good basis for a live test with a modest amount of capital. You could then start testing how much spread might affect your results.
What are you using for charting? Tradingview?