G'day lads, let me know if you have any questions/suggestions 🔥 An explanation of how I built this Linear Regression Model is available here: www.patreon.com/posts/94652720?pr=true
Hi lad, say if Column A is the column of all corner kicks and some rows have blank values. I assume these blank values represent zero corners. So, you can create a new column, e.g. in Column B, and make an IF function. IF the cell ISBLANK, return 0, otherwise return the original value 🔥
HI Excel LADZ I would be interested that such statistical data can be seen at the beginning of the video ( Home coners kicks,home pass ,home shots ,Away coners kicks, away possession,away shot) such where can i find data.
Hi lad, that data can be found in the description of this video. It’s a link to a website called Kaggle, where somebody has kindly contributed the dataset of 2022-23 EPL Matches and the relevant stats. You can download this data, and then copy it into Excel 🔥
@@peterhorvath5372The shooting and possession stats will update by themselves if they are linked to Excel through a Power Query. Then, to refresh the data, you can go into the ‘Data’ tab and select ‘Refresh All’. This will refresh the link to the website Football Reference, and update the stats for you. As for the stats at the very start of the video, this can be considered as just historical data. I’ve just copied these values into the spreadsheet. You can always add more data (matches) into these stats to try and get a more accurate linear equation (due to the sample size increasing). I hope that answered your question lad 👍
Hi lad, the *100 just makes it so that the input for possession is, for example, 60 rather than 60% or 0.6. You could definitely do it without ball possession; just run the regression on shots only 👍
Great thanks. Nonetheless you don't take into account the opponent correct? You assume that brighton is expected to have 6.28 regardless of playing vs fulham or liverpool etc?
Hi lad, there's actually a small change in a team's Expected Corners when you change their opponent, as their Possession input into the regression formula will change. For example, if I've got two teams with an Average Possession of 60%. Using the formula at 6:10, both teams will be adjusted back to 50% (to total 100%). However, suppose Team B was changed to a Team with an Average Possession of 30%. Then, Team A's Possession input into their regression formula becomes 60/(60+30) = 66.7%. So there's a slight change when you change the opponent. As you say though, sometimes this change might not be significant enough when changing the matchup due to the simplicity of the model 👍
Good video lads. Is there a way i can have access to video that explains what each function is for and does and when to use it? Like Linear regressions, Poison Dist. ect. Im new to data science but im invested.
Hi lad, I’m glad you liked it 🔥 I recommend you join the Excel LADZ Patreon (first link in the description). Along with explanations of all the videos (e.g. why and how to use correct probability distributions), you’ll get access to the Excel LADZ Discord where you can ask as many questions as you want 😃