Models aren't about being right on individual matches, they're about being 1 or 2% more accurate than the market over the year. And pricing models tell you the range of probability of an event, so there's no 'getting it wrong'.ija34 wrote: ↑Sat Sep 16, 2023 12:16 pmHow would you produce accurate probabilities though? This is problem with Football, it's a sport where literally anything can happen. So your model could get everything right, predict that something is going to happen, tell you it's the right time to enter a trade because of the current price so you enter and yet the completely opposite happens. Hence having appropriate exit plans, there's absolutely no avoiding losses, we all know that but I think the long-term profit comes from getting the entry and exit points right but is it possible to set and follow rules for entry/exits and be profitable long-term without needing to adjust the rules inplay? Because this is the time emotion sets in and you can make the wrong decision.ShaunWhite wrote: ↑Sat Sep 16, 2023 11:48 amFirst sentence good, 2nd sentence you fell back into the same trap. Ie picking a single selection or scenario first rather than simply looking for opportunity anywhere and any time. Your model needs to produce probabilities which you then compare with the market.
"How would you produce accurate probabilities?" Hard work, hard study, rigorous critical analysis, dedicating every waking moment to contemplation.... etc etc It's not something people do in a fortnight. Equally, statistics is a mature subject so there's no need to reinvent the wheel.