Colossl Victory

Week 12 Review - A slight improvement this week and edging closer to the target of half of the results correct a week. This week the predictor got 4 correct results, 3 of which were correct score lines as well which is a shockingly high proportion. Once again we were denied a correct result (& score) in the early kick off by a 90+ minute goal, the early kickoff starting to become unlucky. This takes the tally to 58 results with 11 of them being correct scores. We also lost more money on the betting system this week as the results got were mostly straightforward with quite high odds for success (Man C, Liverpool, Spurs) and are now at a £.120 loss for the season. Week 13 Predictions - I can't get the graphic to work this week so here is a table, unfortunately this means the predicted GD has been excluded and it is just the results. Home Predicted Score Away Most Confident Brighton 1-1 Leicester Everton 2-1 C...

There has been several reports in the media suggesting that despite Town’s impressive first half performance against West Ham they were in fact lucky to get a point in the end due to a number of high quality chances given away. Now when you look at the xG totals from the game it would seem to back this theory up with the xG score being: 1.19-2.04. The only problem here with these totals is that they take all context out of the individual chances when summing them up, it assumes all chances are independent for the relation to hold. Take Hernandez’s double chance just after half time where he takes a shot from close range which is saved before putting a header over the bar. Now he generates an xG of 0.49 and 0.57 for these two chances respectively, which would suggest that he has a 22% chance of not scoring, 50% chance of scoring once and a 28% chance of scoring twice. Although this can’t be true because how could he have scored two goals in that time there was only 1 second between ...

Week 11 Review - So all in all not the best first week to post about with just 3 correct results including 1 correct score. The 7 draws predicted did manage to catch the only draw of the weekend but  unfortunately for the predictor lots of teams who had been going through bad spells decided to show up; Newcastle, West Ham and Huddersfield (a result I am very pleased about though). This takes the tally to 54 results with 8 of them being correct scores. The bad week also means that we lost hypothetical money and are now down to £1.27 in profit over the course of the season - the Arsenal-Liverpool game making most of the difference to keep it in the green. Week 12 Predictions - The information in the central column is the predicted scoreline followed by the predicted goal difference. This week we have a much more even spread of results with 5 home wins, 2 draws and 3 away wins. Unsurpisingly the two most confident results are Liverpool and Chelsea at home to Fulham and E...

Each week I am going to try and remember to post my ML premier league predictor's predictions for the round of fixtures and keep track of how well it is doing in terms of correct results and scores. I will also be including a track of money won and lost if each predicted result was bet on. This will be done using my v2 for now but will transition to v3 when it is complete. I am starting 10 weeks in to the season as most teams have now played 5 home and 5 away games and thus the attributes fed to the machine learning algorithm should be fully populated and give more sensible predictions. Due to the way it is set up in very early weeks attributes like form can behave weirdly if a team has 0 points and another max even if it has only been 1 game. For comparison, so far the predictor has 51 results correct of which 7 of these have been correct scores. Betting £1 on each game's predicted result would have yielded a profit of £5.01 so far not including stake. Week 11 Prediction...

Expected Points Here’s a quick look at the expected points from Huddersfield Town’s first 8 games this season compared to the corresponding games last season (Cardiff for Stoke) as well as our first 8 games last season. (The expected points would be higher if I had not included the expected goal total from the penalty awarded on Danny Rose, as it was never a penalty, but for the sake of objectivity I have left it in). First let's look at this season’s expected points compared to the corresponding fixtures from last year. I think the first thing it shows is that we were right to be worried after the first 3 games, the performances were clearly worse this season (also worse compared to our opening 3 games last season) and we hadn’t even earnt a whole expected point in that time, the results and performances were bad. I also believe that this has had a knock on effect as to how people have viewed the games since then, as a lot of people have seen the results not improve and ...

Problems and things to correct I think the most glaring issue is that the most likely result calculators logic is not wholly sound, as I think it has some dependent variables being treated as independent and thus the probabilities are not quite correct. For example: goal difference and probability a team scores and concede the corresponding values - if GD is 1 then probability of Team A scoring 2 and conceding 0 is clearly 0 which is clearly not the same as P(A scores 2)*P(A concedes 0). The same goes as for one team to score and other team concedes, they are clearly related.  Result and score are very different, it’s much easier to say a team will win a game than to say by how much, and the current system doesn’t really reflect this at all. It gives a predicted GD but if say team A is 0.5 xG better than team B does this mean to expect team A to win or draw. It may be better to predict the result first and followed by a score based on predicted goals.  There is a pretty b...

The prediction algorithm has been through 2 main versions so far. I unfortunately don’t have the original anymore so can’t quantify how effective it was comparatively, all I can say is that it wasn’t as good as V2 but better than just predicting the most common score from the previous season every week. Version 1 A simple calculation of: total form diff * a + position diff * b + location form (home teams home form - away teams away form) * c + home advantage. The values of A, B, C and home advantage were all calculated by searching all values in a range to find the values that produced the highest total number of correct result over all previous games. A very slow process but it gave decent accuracy. Version 2 The predictive element was replaced with a system to predict the goal difference and another to convert that to a likelihood for each score line, so the one with highest probability can be picked. A linear regression algorithm is used for the predictive element, where a...