ArticlesBlog

Predicting the Winning Team with Machine Learning

Predicting the Winning Team with Machine Learning

Comments (100)

  1. Hey does anyone know how the prediction function works as siraj did in 2:12 , and why did he use XGDB

  2. do more of this Siraj

  3. do more of this Siraj

  4. Siraj, SVM tries to MAXimize the margin between hyperplane and closest points of different classes – not "MINinize" as you said. It was correctly written on your slide but you said the opposite. (22:53–23:15). Thank you for the video anyway. Very usefull.

  5. Nice tutorial indeed. very informative. What software do you use for making the video siraj? If i may know

  6. But can it do it on a rainy day in stoke?

  7. This video is so informative, I hope you can make more videos like this.

  8. GET RID OF YOUR DEBTS TODAY,* PAY YOUR HOUSE RENT,* SETTLE YOUR CHILDRENS SCHOOL FEES,* CHANGE YOU SHOP,* GET NEW CAR/JEEP,* COMPLETE YOUR BUILDING PROJECT,* GET YOURSELF A PLOT OF LAND AND ETC » YOU CAN'T SETTLE ALL THIS ABOVE WITH YOUR LITTLE INCOME AT ONCE. WE ARE READY TO TAKE YOU TO THE NEXT LEVEL TODAY WITH OUR SPECIAL AFFIXED MATCHES OF 250ODDS. TODAY SPECIAL 100% AFFIXED TIPS: HF/FT ••••••••••••••••• ★ Time: 6:00pm★ 3 games.★ Odd: 250★ Stake: $5,000.★ Potential winning $1,250,000.00.(1) ODDS = 5.00~(100% FIXED) HT/FT(2) ODDS = 5.00~(100% FIXED) HT/FT(3) ODDS = 10.00~(100% FIXED) CS░░░░ 100% FIXED ░░░░░☆-:Sure :100% Guaranteed☆☆☆( % SUCCESS ASSURED.)░░░░ 100% AFFIXED ░░░░░░100% AFFIXED KEY GAMES DON'T COME ALL THE TIME SO GRAB > THIS OPPORTUNITY TO BE A WINNER TODAY. SERIOUS MINDED PERSON!!! •★☆» INTERESTED >>> should email me on [email protected]

  9. Please upload the deep learning application of this.

  10. Awesome video, but svm tries to maximize the margin between support vectors)

  11. Hey Siraj, please do continue with sports analytics – its so interesting!

  12. All the datasets.
    https://github.com/RudrakshTuwani/Football-Data-Analysis-and-Prediction/blob/master/Datasets/final_dataset.csv

  13. I'm trying to build the same for a tennis game from where I can get the relevant data ?

    Thankyou.

  14. Hi Siraj,

    Can you please do a video on choosing the correct predictive algorithm/ model based on our data and other factors please?

    I guess it will be helpfull for many

  15. please am a newbie how do i create this with my excel and start wining to. i dont have any formular to input or how to start can someone help me?

  16. or if possible send me the file for me to download please friends

  17. not sure where he gets EPLStandings.csv from

  18. Why do u have 5600 matches in the file while I download the file from football-data-co.uk and only have 290 matches? My file seems to be missing a lot of content……

  19. I am coming across many errors plus there are no features such as HTGD,ATGD,HTP,ATP,DiffLP

  20. Guys I swear I just got free $5641_real money from this_amazing website>>>>freepplmoney.win/?iX0zEg Try once.

  21. This was Awesome. Can you make one using Deep Learning ? Will be really helpful.

  22. For all of you that can't find the right data file, here it is:
    https://github.com/RudrakshTuwani/Football-Data-Analysis-and-Prediction/blob/master/Datasets/final_dataset.csv

  23. absolutely love your lecs but please minimise the talking.. and focus more on the topic practically!! thanks

  24. When you will make video on Indian election Prediction using neural network

  25. is there a way to train the model on the full data set, with variables that we could only know for a past game, and then predict off the model using only known variables for a future game?

  26. it looks like he has a peacock sitting on his hea

  27. Hey Siraj love your stuff, but either some of your coding seems to be wrong (or some has been deprecated). Without first converting the Date column to a datetime64 from an object, the Preprocess_Features function creates the Date column into individual dummy variable. This problem occurs to other columns too; HomeTeam and AwayTeam, for instance, is another object columns that goes into the dummies function. I've also noticed that the function cannot filter correctly, such as the HTForm column. I think you should add a drop function in the for loop to manually get rid of particular columns you don't want. Otherwise, you'll end up with over 700 columns through the get_dummies.

  28. so, can I use the model built ti predict a match if those two teams have never played each other again?

  29. Great video..been working on various football models, you learn the thing is to try and optimize the model to specific niche not seen by the bookie. The goal is to beat the odds, not just correct predictions. Say you predict 80% chance for a team to win,bookies know this too and reflect it in their odds plus sum vig. It seems easy but not quite

  30. This video is just awesome, I don't know what else I should say about it 😜

  31. where i can find newest dataset for Champion league football 2018?

  32. Hey Siraj, they https://betmagician.com/ stole your idea and now making millions! I am also going to steal your idea and do it for tennis!

  33. Is this really possible ?

  34. there is nothing similar in Italy ?

  35. there is no dataset for Italian matches?

  36. this was brilliant! can you do an advanced video for
    this topic ?

  37. my predict for J league which using the same xgboost model only has accuracy 0.45…..

  38. Where do i get the same dataset ? Because my dataset dont have ['HTGD','ATGD','HTP','ATP','DiffLP'] and i have a lot more features then 12

  39. You should try Harrogate towns food

  40. it is hard for me to predict which one will win or lose .but your video helped me i was following https://bit.l y/2J7M2vp

  41. Hey siraj, can I use this as my machine learning project ?

  42. Can i do this in azure machine learning?

  43. Please also use R in some of the videos…

  44. No ML algorithm in the world is or will be able to predict stochastic processes such as soccer or the stock market.

  45. Your training data includes how many goals each team scored. That's clearly data leakage.
    If you would know how many goals each team will score, you don't need a machine learning model to tell the result.

    28:43 Here you can see that the data includes Home Team Goals and Away Team Goals (FTHG, FTAG)

  46. Andrew Groll and his colleagues at the Technical University of Dortmund have paired machine learning and statistical data methods at arriving at the most likely winner of the 2018 FIFA World Cup.
    Knonw more – http://www.estufs.com/fifa-world-cup-2018-winner-predicts-machine-learning/

  47. Love the videos, Siraj!

    I'm hoping this comment will get a couple of likes so that it gets noticed, but I also noticed that someone else had the same issue.

    Trying to run Scraping and Cleaning in Jupyter, I get the following error:

    NameError Traceback (most recent call last)
    <ipython-input-9-e208d65ce9b6> in <module>()
    —-> 1 playing_stat = pd.concat([playing_statistics_1,
    2 playing_statistics_2,
    3 playing_statistics_3,
    4 playing_statistics_4,
    5 playing_statistics_5,

    NameError: name 'playing_statistics_1' is not defined

    …Would anyone be able shed some light on why I'm getting this? I'm running Windows 10, but something must be off with the playing_stat variable under the FINAL DATAFRAME section. Any ideas? Thanks!

  48. How is the F1 score and accuracy score higher for the test set compared to the training set in all the 3 classifiers A, B and C?

  49. Okay 1 question though, we say that xg boost has the best accuracy. But what is accuracy here, i mean how does the system determine if the result is accurate or no?

  50. Does anyone have a copy of the EPLStandings.csv & the final_dataset.csv? When I go through the Scraping and Cleaning.ipynb, I get so many errors and my Python skills are not really good enough yet to cure them, that I have almost given up

  51. Your model shouldn’t just be better than just random chance to bet. You should compare it to the bookmakers odds. I bet if you always take the bookmakers favorite you’d get at least 75% correct.

    Furthermore there are multiple statistical mistakes in this video.

  52. Great video. For the next update you can replace 1 X 2 model to handicaping. I think this is the more popular way of betting.

  53. Great content as always. Thank you!! Would love too see more soccer ml takes.

  54. Where does the set EPLStandings.csv come from?

  55. There are not data enough to achieve that a machine learns something about football. When you know the nature of the game, the nature of football, you don't need any date but watching the match for e few minutes. It's like chess: a master chess just needs a simple view of a chess game to know who is winning. No data, no computer, no algorithms, but just a simple view over the green to process what is happening in the match. S2.

  56. I downloaded the dataset from the website but there are some columns which are missing for example half time goal difference

  57. I want to ask, you're using python right? Where can we get those libraries you are importing?

  58. I couldn't see where you were typing around 15:00

  59. @Igor agree with you. Along with this, there is problem of Data Leaking. Before match does start, how can one know no. of corner kicks, on shots, red cards etc. Very popular data mining problem.
    Great video otherwise BTW @Siraj Raval.
    Love from India 🙂

  60. If you want to make money with such an approach it would be better to build a model that outputs the probability of each outcome such that you can calculate whether you're going to win against a bookmaker. It's often better to bet on the draw or the underdog because the odds are better.

  61. hey ,if anyone else wants to discover soccer scores predictions try Jadonite Top Sports Buddy (Have a quick look on google cant remember the place now ) ? Ive heard some unbelievable things about it and my brother in law got cool results with it.

  62. Do you have a site for current future predictions mate? Id like to try out you're method if possible. Im not this adequate in computer programming. Give me a bell if you can brother.

  63. Hello Mr Siraj Raval, i have been following your videos on machine learning and was wondering if you can help me? I am from Nigeria and here we don,t have much of Data Scientists or Machine AI expert. My company have been looking for a very good and experience Analyst that can give us a Model that will predict football outcome, i.e. HALFTIME/FULLTIME results that are either AWAYWIN/HOMEWIN or HOMEWIN/AWAYWIN. We also have other areas like Stocks, Business Input/Output, and much more. Can you recommend or connect me to anyone that can be perfect for this Job. Salary is very big and will keep increasing for as long as the analysis is on-point. Expecting to hear from you soonest…

  64. This is great. Am now fully committed to learn machine learning next year! thanks for the great work Suraj!🙌👌✔

  65. Please make more videos on this topic Sir! 🙂

  66. Amazing, you are a genious Siraj,

    What if you have a system to capture cast data before the game starts, so know if any important player will play what can affect the outcome.

  67. This video increased my interest in machine learning

  68. He sounds such a bellend even talking about football

  69. will you make a data prediction on this file
    https://drive.google.com/open?id=1JsiNp8Du6WIrrnoSsflKqUxJkLTV9KPR

  70. haha so basically it will work if I use the right numbers?

  71. #Deflategate #IsisKDaesh

  72. can you please, upload the video showing
    overall process of creating models of various variables choosing anyone prediction algorithm sir!!!

  73. WARNING
    Will not get rich with that. This is exactly what bookmakers do. It's easy to predict which team is the most likely to win

  74. Is any one there to help in r programming to calculate the statistics analysis!

  75. what type of software should i use to code for machine learning ?

  76. I have access to an algorithm that is 70.5% accurate across multiple sports, stocks, in fact any future event.

  77. hello suraj, im working on the same project which is been shown by you in this tutorial, but i m confused with data cleaning, the video shows you just started with jupirter and then some commands and boom, you just got the required results. But how to do this ? will you elaborate the procedure for this please, would be really glad to know.

  78. Great video but i'm still not clear on what could be some good features.

  79. From where did you get the trained data set

  80. thoroughbreds pls [horses]

  81. Am doing this as mini project. Do you have source code for this

  82. Hello Siraj! I second Igor's opinion about splitting the data according to time and not randomly. Having said that, I thoroughly enjoyed this video. I would love to see another video on the same topic which includes sentiment analysis from twitter feeds as well on top of the baseline models. Perhaps we can use LSTM with a time distributed dense layer to capture the time information?

  83. Thumbs up if you guys want a part 2 of this video (using deep learning + twitter sentiments)!

  84. How this trained data can predict a team….it is just a upto fit the data and showing accuracy…nothing predicting

  85. Hi Siraj,

    could you explain a bit, what "Difference in last years prediction" is and how do you calculate it?

  86. AWESOMENESS SHOULD BE YOUR NAME

  87. Where did you get these variables: 'HTGD', 'ATGD', 'HTP', 'ATP', 'DiffFormPTS', 'DiffLP'? They are not in the csv file.

  88. no ones can predicted soccer games

  89. i want to talk to you for my project

  90. can i get link for the dataset

  91. Sports are fake and scripted you can't beat the bookies they already know who will win they only want to lure you to the losing side, is impossible to win with machine learning for sports is a waste of time. All the numbers you see the line the money is fake you need inside info like billy walters is the only way.

  92. so i copy this code edit it a little and start gambling. Then me run to the bank..

  93. Abey chutad, in the whole program its 'X_test' and not 'x_test'. Yet youre able to run this program. WTF?
    Also, you fuckin idiot, dont you know how to input a row data? I'll tell you, you use a method called .iloc()!

    My problem isnt that youre a stupid piece of shit. My problem is how the fuck are you running wrong code dude?!

  94. Does it give VALUE BET ???

  95. Beating the bookmaker will take you more than that : )) 75 accuracy ? ok what about the odds lol
    the only guy who did it using ML : https://youtu.be/2BhWvdIHQYM

Comment here