All your stats at each observation should be up to that event, not including it. Then you split the data into train/test. Build your model on the train and see how well it does on the test. Repeat this randomly using k-fold. There will still be some bias in the train data, the data has to represent as close as possible the conditions at the time you intend to place a bet, just before the game or in the morning etc. It's a tough game but fun.
Thanks for the vid, super informative. Quick question: How do you keep all your sheets up to date with stats, bets, etc? I've seen people having data scraping scripts running, but is there an easier way to do that for a beginner?
you hit on some excellent points. there are bettors who share, hindsight data or trends constantly. 'if you had bet X for the past 25 games you would be up X'...implying that you ought to continue doing this. past results dont indicate future returns. models need to adjust each offseason, including rule changes and not go off a dataset when you have all the information. i've shared this video with a few ppl. on the matter of testing if you have access to software that can simulate your data you can alleviate inherent flaw you mentioned and get a forecast instead. i know thats not the goal of your series, which focuses on excel. had a laugh because your clickbait video [the TRUTH about] received 20x more attention than the quality instructional videos you've posted. you can lead a horse to water...you know where im going with this. anyway keep up the good work!
Good stuff I am a pro poker player that is getting in to betting as it's a lot more scalable. What math do you base the 400-500 sample you need on? Also I must say I learned a lot and thanks for the vid you seam genuine to me.
Let me give some insights from an experienced Sports Trader's (20 years) perspective. 1. Almost all sports have handicaps and more importantly handicap opening odds. If the bookmaker things a team is going to win (E.g. for Soccer/Football), they will offer 1.85 for one team and 2.00 for the other. Of course they are thinking that the team with 1.85 odds is going to win and hence winning a 15% spread, or what we term as water money in Asia. 2. Market Consensus. Without knowing what is the opening odds and tracking the odds movement, one is clueless about the marketing movement. Only by knowing how to track opening odd, odds movement transition and market consensus from multiple sources can a person estimate what is really happening behind the scenes. 3. Trading Signals. Most people like to use machine learning or deep learning to try to use past results to figure out what is a good trading signal. Unfortunately things doesn't work out that way for trading. What you do in paper trading and actual trading are totally different. In paper trading, one would take on more risk or less risk compared to actual trading or vice versa. In real trading, it has got to do with the information that was present at that moment, the execution of the trade, trading psychology, capital (bankroll) management and much more. So in short, successful traders are only able to derive strong golden trading signals from real successful trading executions from thousands of trades. Not by running probability tests using machine learning or deep learning. These are only tools that are meant to enhance the probabilities of trades rather than trying to find out trading signals from the scratch. 4. Trading psychology. Without understanding the limitations of human trading, e.g. Emotional trading, Trading Euphoria (God Mode), it is hard to factor in the different aspects of risk taking into an automated A.I. trading bot for any kinds of trading instruments.
Not practical advice at all. This is why statistic majors dont make any money in sports betting. Unrealistic expectations. How do you backtest the nfl with even a 1,000 sample size. Like cmon.... the sportsbook is doing around the same.he is doing. Limited trials of available data.
Do popular training program like Episoketren System really work and if so, how effective are they? We've noticed many awesome things about this popular training course.
What on earth is this? Makes absolutely no sense. This is not how you test. At the very least, please use the previous season's data to test on current season games. Even that is 500 times better than using data you didn't have available at the time.
All your stats at each observation should be up to that event, not including it. Then you split the data into train/test. Build your model on the train and see how well it does on the test. Repeat this randomly using k-fold.
There will still be some bias in the train data, the data has to represent as close as possible the conditions at the time you intend to place a bet, just before the game or in the morning etc. It's a tough game but fun.
This is what I do. Pinning the comment.
How do you backtest against a player slamming his gf against a wall, being benched and the team losing the game.
Thanks for the vid, super informative.
Quick question: How do you keep all your sheets up to date with stats, bets, etc? I've seen people having data scraping scripts running, but is there an easier way to do that for a beginner?
you hit on some excellent points. there are bettors who share, hindsight data or trends constantly. 'if you had bet X for the past 25 games you would be up X'...implying that you ought to continue doing this. past results dont indicate future returns. models need to adjust each offseason, including rule changes and not go off a dataset when you have all the information. i've shared this video with a few ppl.
on the matter of testing if you have access to software that can simulate your data you can alleviate inherent flaw you mentioned and get a forecast instead. i know thats not the goal of your series, which focuses on excel.
had a laugh because your clickbait video [the TRUTH about] received 20x more attention than the quality instructional videos you've posted. you can lead a horse to water...you know where im going with this. anyway keep up the good work!
Good stuff I am a pro poker player that is getting in to betting as it's a lot more scalable. What math do you base the 400-500 sample you need on? Also I must say I learned a lot and thanks for the vid you seam genuine to me.
400 to 500 is generally a lower end of a 95% confidence interval
@@williamleiss4 cool thx I am new to this and eager to learn. I find it fascinating
whats your strike?ratio? 10 games what is your accouracy, mine is i win 9 or 10
Let me give some insights from an experienced Sports Trader's (20 years) perspective.
1. Almost all sports have handicaps and more importantly handicap opening odds. If the bookmaker things a team is going to win (E.g. for Soccer/Football), they will offer 1.85 for one team and 2.00 for the other. Of course they are thinking that the team with 1.85 odds is going to win and hence winning a 15% spread, or what we term as water money in Asia.
2. Market Consensus. Without knowing what is the opening odds and tracking the odds movement, one is clueless about the marketing movement. Only by knowing how to track opening odd, odds movement transition and market consensus from multiple sources can a person estimate what is really happening behind the scenes.
3. Trading Signals. Most people like to use machine learning or deep learning to try to use past results to figure out what is a good trading signal. Unfortunately things doesn't work out that way for trading. What you do in paper trading and actual trading are totally different. In paper trading, one would take on more risk or less risk compared to actual trading or vice versa. In real trading, it has got to do with the information that was present at that moment, the execution of the trade, trading psychology, capital (bankroll) management and much more. So in short, successful traders are only able to derive strong golden trading signals from real successful trading executions from thousands of trades. Not by running probability tests using machine learning or deep learning. These are only tools that are meant to enhance the probabilities of trades rather than trying to find out trading signals from the scratch.
4. Trading psychology. Without understanding the limitations of human trading, e.g. Emotional trading, Trading Euphoria (God Mode), it is hard to factor in the different aspects of risk taking into an automated A.I. trading bot for any kinds of trading instruments.
Hello
Can you give a detailed app that can predict well?
Sir, your videos are awesome could you please do video on cricket player analysis
Where do you find the data for your backtesting? ie lines, spreads. Do you run your backtests in Excel or a more robust language? Thanks
Excel and Python. Data is data I have mined over the past 8 years or so.
Yes, back testing is absolutely necessary .... especially when looking for value.
What do you think of the Action Network app/service?
Tout Service
This is why you don't bet any league until a month or so into it...I don't touch baseball until early May...
Why subtitle?
100k samples minimum, 400 way too low
Not practical advice at all. This is why statistic majors dont make any money in sports betting. Unrealistic expectations. How do you backtest the nfl with even a 1,000 sample size.
Like cmon.... the sportsbook is doing around the same.he is doing. Limited trials of available data.
@@jaco9286 you just don’t model around nfl brother, there are other sports 😉
your mic is to big :P
Any one tried the Episoketren System (search on google)? We have heard many awesome things about this popular training program.
Do popular training program like Episoketren System really work and if so, how effective are they? We've noticed many awesome things about this popular training course.
Your method is not the what bookmakers do have.im preety sure this is make you lose up and down people will say finally sports impoosible hgahgah
i dont bet with models way to complicated for me fuk that
Yeah, I avoid the NFL... not enough info, it's a crapshoot.
What on earth is this? Makes absolutely no sense. This is not how you test.
At the very least, please use the previous season's data to test on current season games. Even that is 500 times better than using data you didn't have available at the time.
👎