 Hey everyone, it's MJ the student actuary and in this video I'm going to attempt to predict the world cup final score between France and Croatia and what I've done just to save time I've written on all my calculations, and I'm going to talk you through it Essentially what I'm doing is I'm using a poisson distribution that's just depending on the actual results of the games in the world cup and not going to be looking at Individual players performance or you know Harald Mbappé or Luka Modric is doing So we're looking just at the numbers of these teams to calculate what their score is going to be So to do that I have calculated three things for each team the amount of goals that they scored In the world cup so far, so it's been 12 for Croatia 10 for France the amount of goals that they've conceded That has been five for Croatia and four for France. I've then also calculated the amount of time played And that's been 634 Croatia and 540 for France. Look, I have rounded down to either 90 minutes or 120 I'm not using, you know, 90 minutes plus five or something like that But because I've been well consistent on both teams that shouldn't be too much of a problem Anyway, what I've then done is I've calculated my landers and in order to get my landers what I've done has just been say 12 over 630 Times 90 minutes the 90 minutes that we're going to be playing in this upcoming game And I've done that for Croatia's goals that they're going to score The goals that they're going to concede and I've done the same for France Then what I've done to determine Croatia's score is I've combined their goals for with the goals that France is going to concede to create a Lambda or a single lambda for Croatia score and I've done the same with France to get their score and the interesting thing is is though Croatia is Scores a slightly higher rate of goals than France They also concede a slightly higher rate of goals to France and at the end of the day both of these teams had exactly the same Landers, which means the expected full-time score is going to be 1-1 After 90 minutes, which means we're going to go to extra time when you extend this lambda over 120 minutes we get one and a half goals or just over one and a half goals each for each team Which means we could be looking at a 2-2 after extra time. So it's going to be an exciting game Penalties are definitely on the cards Of course the score of 1-1 before you go run out and bet on it It's important to know the odds of that actually happening and I did that using the poisson distribution and The actual odds of it being 1-1 though It's the most likely thing to happen according to this model. It is only 13.1% Anyway, this got me quite interested in just saying well, yeah, to think what are the bookies Suggesting so sort of writing up their stats and I thought let me do it for all of the scores So in the pink we have the odds that the bookies have said for each score Look, I am only using only use one betting website the betting website you use might be different and Then I also compared it to the the model that I made and all I'm simply doing it's it's using poisson Very standard grade statistics. Yeah But what was interesting is I mean if we come to if we look at the draws They the bookies think that a nil-nil draw is way more likely Then with my models predicted so they said 17% and I've said it's only gonna be 9% chance The interesting thing is and why you probably shouldn't bet on my score is because the bookies have put it at 14.28% which means you're gonna be paying a premium and I guess that's what you do for betting You always are paying that premium for that entertainment factor and the the betting houses need to always win at the end of the day But look how close our odds were for a 2-2 draw I mean this model is not considering like I said any of the players form anything like that other than their results in this world cup And look how close that is 4.64 to 4.76 3-3 draw was also quite similar But where the model did go quite differently is for a 4-4 draw The poisson is saying that that's very unlikely Whereas the bookies at half a percent. It's still unlikely But it's it's a lot more likely than then what the poisson model is suggesting Of course, we saw the biggest differences with France winning the bookies are favouring the French And you can see that 18% to 11% for a 1-0 to France 12.9 to 6.55 for 2-0. I mean that is almost double So that would almost be the worst score to bet on although there's still a 6.5 percent chance of happening So that might actually be the score but according to my model That's one of the worst things to bet on also this one here 3-0 to the French also not that good to one you were a little bit closer and On the Croatian side We were also a lot closer with a slight variation Favouring the model we got 11% to 10.8 6.55 to 5 7.8 to 6.24 and 2.6 to 2.4. So these I mean Not that betting is a good idea because like I say it's it's very risky. It's probably legal in your country You have to be over a certain age But the odds of my model do kind of favor or suggest that Croatia is the better bet But like I said betting is very risky So I am going to be betting on Croatia I think it's the right move to do just because it's going to make the game a lot more exciting But let's see what happens. Let's see what happens according to the the maths we should be seeing a full time of 11 and then when it goes to penalties, I think anything can happen It's 50 50 and that's why if the bookies are favoring the French It makes sense to bet on Croatia because then you're going to get a higher payout But we'll see we'll see what happens. Let me know in the comment section below what your score prediction is But it's quite interesting to see how close the mathematics is to the bookies who are considering a whole bunch of other factors as well Anyway, hope you guys enjoyed this video and I'll see you for more stats videos. Cheers