No video :(

The PROBLEM with using POISSON distribution to PREDICT football matches... [Tutorial PART 2]

Подписаться 900

Просмотров 3,5 тыс.

50% 1

In this video i will be showing you the problem with the using poisson distribution following on from the last video in which i showed how to make a spreadsheet model for poisson didstribution that helps with football betting, all from scratch. This is just a tester with games as an example and many many games would need to be tested before being implemented. Remember this is just for fun, please gamble/trade responsibly and carefully. This is not financial advice, rather it is just me having some fun finding and testing new trading strategies on the betfair exchange.
For custom spreadsheet requests: www.fiverr.com...

Опубликовано:

23 авг 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 23

@mattmid7012 5 месяцев назад

Great video. This is why I'm a firm believer in using the data from at best similar games or at least due to lack of data top/middle/bottom 'halves' For instance there is little point using data from say Man City at home to Sheffield Utd in a model for an upcoming game of let's say Man City v Arsenal.

@Predictitfootball 5 месяцев назад

Exactly! That’s a great way of approaching it🔥

@Jimt4728 Месяц назад

I had been looking at the correct score markets but abandoned the idea after the new rules came into effect with additional time added for goal scored celebrations etc which I felt affected any analysis.

@Predictitfootball Месяц назад

I agree. I think getting involved in correct score markets can be very difficulty especially as it’s so volatile and as you say with the new added time rules it does make it a lot harder with a lot of late goals

@lollipopboientertainment4792 8 месяцев назад

Can you use a Monte Carlo simulation to improve poission distribution model?

@Predictitfootball 8 месяцев назад

Yeah you can! It’s a really good way of getting simulations (thousands) and ultimately getting convergence

@AlanH123 9 месяцев назад

Thanks for sharing. I use Poisson Distribution for trading correct scores but I know it has its limitations. I've been looking for a better way to identify target scores so following this with interest. Do you have any thoughts relating to the correct score market?

@Predictitfootball 9 месяцев назад

So Poisson distribution isn’t the end of the world, it can be good but I was just showing an extreme example of when it can be bad. In the next part of the series I’ll show you a better way of predicting football and target that at the correct score for yourself👍

@Predictitfootball 9 месяцев назад

@viva50s61 hopefully the new video should help you with this

@AlanH123 9 месяцев назад

@@Predictitfootball thanks! I saw it earlier today and watcvhed it once. I'll watch a few more times later tonight when I get some time. Thanks so much for doing this for me.

@sportspriceanalysis416 7 месяцев назад

The problem is that Poisson itself tends to be skewed toward the lower values, typically zero which is why there is another distrubution called 'Zero Inflated Poisson'. That isn't the biggest issue though, the distribution itself assumes that events are independant which of course they are not, if one team scores then its going to change the likelihood of the other team scoring. I havent actually used ZiP so not sure of its effectiveness vs a typical PD.

@Predictitfootball 7 месяцев назад

@@sportspriceanalysis416 exactly. Completely agree. That’s actually the next part of the series using ‘zero inflated Poisson’. This is all part of a series for beginners to build models and learn excel skills

@ChristianRavnkildeAndersen 8 месяцев назад

In this video you mention that the model can be improved and you will show how to do that in your next video. Where is that specific video? Can you link to it? Thx in advance. Thx for great videos and inputs.

@Predictitfootball 8 месяцев назад

Hi. So in terms of how the method can be improved I essentially made a video showing that it’s good to use the spreadsheet in cohesion with other methods. The actual video is linked below. I hope you enjoy and thanks for the support: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-HpOwdk6r3r8.htmlsi=vtX8BuJRdXCPJESE You may also find this video useful: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-mwG5bpoQ69k.htmlsi=kq2c7AdsxmH0PEF3

@mashudumashamba7321 6 месяцев назад

@@Predictitfootballhey man can you make the video public 🙏

@Predictitfootball 6 месяцев назад

@@mashudumashamba7321sorry only just saw this reply. The second video linked should be visible. The first one had a little issue with the audio so had to private it after release sadly. But I’ll make a similar vidoe as soon as I can. But the model video should be public

@mashudumashamba7321 6 месяцев назад

@@Predictitfootball ok thanks

@luigibettelli7840 Месяц назад

@@Predictitfootball hey there. just came across this video, but the link you reported are not public again. Where can I find this new video for improving poisson distribution?