Can AI really help with sports betting, or are we overrating it?

[bitcoindusts]

And do you think AI can create a real edge in sports betting, or is it mostly hype for now?

I don't think it is a hype for now, since AI help us gather data more faster and more efficiently, it can help us with team analysis, AI can give suggestion and reasons on its suggestion but at the end of the day , it is our own analysis that will prevail.

Having someone to gather important information about competing teams, their statistics, recent performance, team injuries, team roster updates and so on, this data alone if we are good in data analysis can help us increase our chance of predicting correct outcome of the game.  But obviously, in gambling, success of winning works by percentage and probability, not by absolute things.  So it is normal for a gambler to lose even though they find out that the one they choose and bet on have higher chance of winning.

[Jubilee58]

Sometimes I use an Aiscore to make make my predictions, it shows you stats, and gives you all necessary information to enable win your game. It is actually helpful to make an informed decision, but does not guarantee winning. I have used Aiscore several times to make my gambling decision, but eventually one or two games will spoil the bet. In essence using Ai to predict bets gives you no full guarantee of winning your bet.
While making your sports bet prediction, don't wholly rely on Ai, your personal research, and ideas can still help you win a bet instead of fully resorting on the ideas of Ai.
Gamblers do win matches without the use of Ai, there are actually some tools, and stats that can aid you on that, and sometimes, you may not play your bet as predicted on those platforms, you can decide to make some informed changes that can result to your own favour.

[puloweh555]

Sometimes I use ChatGPT to get statistical information from both teams. If you ask about winning chances, it will also provide results based on statistics. 
I do not trust predictions given by AI. But surely there are some gamblers who use it to place bets. But I will not use it. Such things might not be any different from betting based on someone else's predictions.

Yes, I also don't use AI as a reference when betting. Because what AI provides is actually based on statistics and the history of the two teams' meetings. This analysis is also quite general; sometimes I need to look for specific information, for example matches where coaches rest players to prepare for the UCL semifinals, such as PSG vs Loirent and München vs Heidenheim.

If we rely on AI as a reference for betting the AI ​​will certainly suggest betting on PSG and München winning. However, if we analyze it ourselves we will use logic it's impossible for coaches Kompany and Enrique to field their best squads for a domestic match when they face a crucial UCL match in a few days. This is just a small example.

Furthermore, I don't use AI for betting because I don't get the satisfaction of gambling if I rely solely on AI but rather on my own analysis. But this comes back to every gambler, in essence the goal in gambling is not only to seek luck but also to seek pleasure.

[moneystery]

Since AI has gotten smarter, I think it can help us predict matches. However, I don't think it's very accurate because AI is limited in its ability to analyze data, not to mention other factors like intuition, luck, and other factors, which make AI predictions imperfect and potentially wrong.
So, if someone wants to rely on AI for predictions, they shouldn't be overconfident that the AI ​​can make predictions and then just accept them. They need to conduct their own analysis to be confident in the results, and only bet with cold money so that even if they lose, they won't regret it too much.

[sompitonov]

Since AI has gotten smarter, I think it can help us predict matches. However, I don't think it's very accurate because AI is limited in its ability to analyze data, not to mention other factors like intuition, luck, and other factors, which make AI predictions imperfect and potentially wrong.
So, if someone wants to rely on AI for predictions, they shouldn't be overconfident that the AI ​​can make predictions and then just accept them. They need to conduct their own analysis to be confident in the results, and only bet with cold money so that even if they lose, they won't regret it too much.

I think players are currently overwhelmingly thinking that AI will help them, but the fact is that many players are using it to try to improve their winnings. It's not just that, but also that the employees on gambling platforms see all this and will adjust the odds so that they, rather than all players, profit first. Because if this doesn't happen, the business won't survive, that's all. Time will pass, and players will realize that AI isn't very helpful, or that it's only temporary, and they'll decide that they can win permanently with AI, but that's certainly naive.

[Hazink]

Sometimes I use ChatGPT to get statistical information from both teams. If you ask about winning chances, it will also provide results based on statistics. 
I do not trust predictions given by AI. But surely there are some gamblers who use it to place bets. But I will not use it. Such things might not be any different from betting based on someone else's predictions.

Are there really people who blindly trust predictions made by AI or even people they consider to be pros without doing their own research? I still wonder how since those who are more focused on possible winning can be doing that. I still wonder how they can just trust someone else to make decisions for their own money they are using to take in the game. Ai prediction is also included when I talk about trusting third parties.

[Slow death]

And do you think AI can create a real edge in sports betting, or is it mostly hype for now?

I don't use AI for game analysis; honestly, I don't even know what advantage I'd gain from doing so. I'd lose all the fun. The only thing I use AI for is when I want to make memes from photos, because AI is terrible at photo editing. Aside from that, I don't use AI for anything else, and I think AI, like us humans, won't be much different in predictions. In fact, humans have an advantage because we can watch team videos and know how each player plays, while AI will only analyze data available on the internet. I prefer to continue analyzing games myself.

[o48o]

Since ChatGPT came out in late 2022, AI has become mainstream enough that almost every industry has people trying to plug it into something. But I’m curious about sports betting specifically.

Is anyone here actually building or testing anything serious around AI and sports betting?

I don’t mean those random “AI prediction” websites that just say Team A will win,  Team B will lose, and gate it behind a paid service, and then disappear when the pick fails. I mean proper systems that try to model games, compare bookmaker odds, track closing line value, separate strong picks from watchlist picks, or even handle bankroll/risk decisions better.

For example, I’ve been playing around with the idea of a football betting system that does more than just predict match winners. Something that looks at team strength, expected goals, home/away scoring rates, goal probability models and using all of that data to make a goal grid to predict the probability of a fixture ending in over 1.5, over 2.5, etc.

The more I look at it, the more I feel AI may be more useful as a research assistant than as a magic prediction machine. Maybe it can help organize data, spot weak markets, compare lines across bookmakers, explain why a bet is risky, or separate “looks good” picks from bets that actually have some value behind them.

So far, I have had 2 wins and 1 loss and each bet was an acca of 2-4 odds.

Has anyone here tried building something like this, even as a personal project? And do you think AI can create a real edge in sports betting, or is it mostly hype for now?

You do realize that sport books in most games already give you prediction of the outcome with their odds? That doesn't mean they are correct predictions either.

I tried different kinds of LLM's in the past to get me predictions for matches, and asked them to explain their reasoning. What i didn't understood back then, was that these free language models aren't "intelligent" nor they are designed for that. It's basically giving you words that is likely to be next to each other. LLM doesn't understand the content.

And even if you insert data to compare, at some point they started to hallucinate and might "forget" or invent new sections of it.

But by all means, use it, as it's most likely giving as good results as trusting to your gut.

[hedgeh0g]

And do you think AI can create a real edge in sports betting, or is it mostly hype for now?

I don't use AI for game analysis; honestly, I don't even know what advantage I'd gain from doing so. I'd lose all the fun. The only thing I use AI for is when I want to make memes from photos, because AI is terrible at photo editing. Aside from that, I don't use AI for anything else, and I think AI, like us humans, won't be much different in predictions. In fact, humans have an advantage because we can watch team videos and know how each player plays, while AI will only analyze data available on the internet. I prefer to continue analyzing games myself.

I also continue to analyze games myself, because I haven't yet figured out how to best other players and brilliantly ask AI questions related to sports betting. Perhaps, at the beginning of the AI ​​era, some players are currently making money from it, while many others underestimate its importance in betting. But in any case, we'll only find out about this later, because right now no one is willing to share this information if it's truly useful. It's also important to understand that different versions of AI can be used for betting in different ways, and whoever gets there first may be rewarded.

[Davidvictorson]

Yup. I get that. The reason I am bringing it back is because current AI models are continuously being improved. So the idea is that as these models get improvements and get smarter, they would get better and smarter right? From my understanding so far, just allowing AI to make all the decisions is wrong but I'm actually talking about actually using it to create a deterministic model that does 99% of the hardwork. Before now the bar to creating such models were ridiculously high but I reckon it has dropped considerably in the last few years.

I would say thank you for adding this and I think you are correct as well. What I will add to it are two things. AI models though they are improving would need large data to be able to topple any casino system and unfortunately no single individual can do that at that scale. The person would need a lot of money to train the model to make predictions. It cost money and requires expertise. Secondly, I believe that online casinos have their own tools in place that would flag any suspicious account. Yes, they are that some people would attempt it and they will be ready for them.

[fredericktaylor]

AI can help a lot in sports betting but the help is mostly kinited to the provision of information such as the latest and historic information regarding the teams and players, it can give predictions according based on the data available but the prediction is not 100% successful. I like to see AI as a tool for information alone, so that the final decision still rest with the player.

As it is with any new technology, people are forcing AI into every area of human endeavors and gambling is not exempted. This is why many people believe that AI help them in making accurate prediction, a claim I know is not true. So yes I believe that AI is overrated when it has to do with gambling.

AI can collect various information very easily and quickly and solve various complex issues very easily. AI can announce an approximate result based on various information, there is no confirm that the result will be completely correct. Gambling based on the results of AI does not increase the chances of winning, as a result, there will be a higher chance of losing money. It is better to gamble through research without greed or emotion, the chances of winning depend more on luck, so if luck is good, it is possible to win, but if you gamble through proper research in a systematic manner, you can protect yourself from big dangers.

[Akbarkoe]

Friend, AI only calculates probabilities from the data they have, the more data they have the better they are at making calculations, that's how AI works.
 
I have tried using AI this year and in the previous year because I think AI this year is much better at data problems than the previous years which are still lacking.
However, the results are still fairly random, there are wins and losses, so it's the same as when we bet without using AI, and AI friends just like humans cannot predict the future.

The results are indeed random. I have not given the AI predicting thing much time, but those who have been using it can only say it helped them reduce the stress of sourcing data from different sources and compiling them together, but at the end we have access to almost the same data, and our predictions are based on history and updated team data. Even with that, we still can't get it accurate all the time, likewise the AI.

Yes, it only helps us look for data in one place because the sources are also the ones we usually use, so we only need to double-check each reference of the data source, it saves a little time.
I need to say repeatedly that there will be nothing accurate with a 100% guarantee in predicting even using the most sophisticated AI, because they will not know what the future will be like, they only talk about probabilities with the conclusion of the data they have.

[Ronsbit]

The more I look at it, the more I feel AI may be more useful as a research assistant than as a magic prediction machine. Maybe it can help organize data, spot weak markets, compare lines across bookmakers, explain why a bet is risky, or separate “looks good” picks from bets that actually have some value behind them.

You have said it all here. AI can possibly be used as a research assistant, as you have opined, to set up information, data, analysis, and the rest of the other requirements that one would deem fit and necessary to use in setting up a debate or an argument for a proper verdict.
For sports betting, you could get data about previous games and also a forecast, but a good analysis should come from you, void of AI conclusions, because AI can not give you an accurate prediction based on its analysis of previous matches, but only a forecast, which would likely not be accurate.

[Mr Reporter]

Since AI has gotten smarter, I think it can help us predict matches. However, I don't think it's very accurate because AI is limited in its ability to analyze data, not to mention other factors like intuition, luck, and other factors, which make AI predictions imperfect and potentially wrong.
So, if someone wants to rely on AI for predictions, they shouldn't be overconfident that the AI ​​can make predictions and then just accept them. They need to conduct their own analysis to be confident in the results, and only bet with cold money so that even if they lose, they won't regret it too much.

I think players are currently overwhelmingly thinking that AI will help them, but the fact is that many players are using it to try to improve their winnings. It's not just that, but also that the employees on gambling platforms see all this and will adjust the odds so that they, rather than all players, profit first. Because if this doesn't happen, the business won't survive, that's all. Time will pass, and players will realize that AI isn't very helpful, or that it's only temporary, and they'll decide that they can win permanently with AI, but that's certainly naive.

Well it sounds much this world would come to think this that Ai more feel that Ai is a very one..well to me it also one but just check how your region leaders who are to control the church how a pastor used Ai the to preach in the church…well that not even the point am hitting back to gambling sector well having to use Ai to predict games it more cheating and the business might close down if the business operations system keeps quiet to watch out how the game which is might to change the game…

Things like this should not even happen’s a again, I well I will wat permission to do thing ohhh…most it even not to be allowed to do such..

[Cointxz]

Yes, it only helps us look for data in one place because the sources are also the ones we usually use, so we only need to double-check each reference of the data source, it saves a little time.
I need to say repeatedly that there will be nothing accurate with a 100% guarantee in predicting even using the most sophisticated AI, because they will not know what the future will be like, they only talk about probabilities with the conclusion of the data they have.

You’re right on it just makes our life easier to gather data and analyze it for us in the same manner. The help which many gambler expecting is the high accuracy prediction based on the same data analysis that the AI currently doing which we all know that it’s hard to have an accurate prediction using recorded stats alone because there’s still other factor that can affect the game which is not recorded on the usual stats such as physical records and so on.

AI can only help to provide a logical prediction based on stats but it can’t guarantee a high accuracy prediction because of other factors involved.

[Ayers]

Are there really people who blindly trust predictions made by AI or even people they consider to be pros without doing their own research?

not only prediction, i've seen people even are now borrowing ideas from AI to carry out major terrorist acts,,,

artificial intelligence may now be like an assistant who has reduced a huge amount of work. maybe in second you can get all the stats, injury update, how the player is in form or old history at your fingertip. but AI still struggle to understand the most vivid thing on the field. also it has no emotion. such as changing tactical move in the middle of the match, the players stubbornness or mental state, the coach's last minute decision, the weather or the referee's mood, these thing are still beyond the intelligence of AI

so i would say the wisest thing to do would be to first collect all the hard data and calculations with AI. then make decision based on your own intelligence and current situation on the field. so those who place all their bet based on AI thinking that it is a magic machine for predicting the future will ultimately be in trouble

[Z_MBFM]

Ai can help us to win on sports betting but not guaranteed on every bet. Ai can do analysis and give us a approximate answer. some time it will work and some time it will fail. so i think we shouldn't use Ai for gambling prediction we can just ask to Ai for collect some data and then we can make a prediction ourself based on that data. because Ai can collect data in few minutes so it will save our time.

Gambling is actually depends on our luck in any type of gambling it can be sportsbook or casino games. so always we have to accept the risk and use small amount of money that we can afford to loss and that loss we can easily forget so only then we can protect ourselves from gambling addiction and emotion

[Odogwu-Blockchain]

AI could only compare data, compute previous results and h2h analysis on each club and bring out possible options that could best fit the outcome, where a gambler can pick the best option best known to him depending on his intuitive perception and trust on the AI tools used. From what I know, AI cannot give you the best predictions but can give you options to choose from it.

If I may ask, what AI tools did you used for your predictions?

[nelson4lov]

AI could only compare data, compute previous results and h2h analysis on each club and bring out possible options that could best fit the outcome, where a gambler can pick the best option best known to him depending on his intuitive perception and trust on the AI tools used. From what I know, AI cannot give you the best predictions but can give you options to choose from it.

If I may ask, what AI tools did you used for your predictions?

Okay, you're the first to ask about details of my experiment haha.

I currently use codex by openai as the main tool to build what I call the system. So right now, it comprises of:

• A model
• A Selection layer

Both of these are deterministic btw.  The model is a simple mathematical tool that takes data like xG, npxG (non penalty goals  incase a team just got lucky with penalties), along with other goal profile like home and away goal expectancy if  xG data is not available.

Then we take all of that data to compute a goal grid with possible scorelines and their probabilities.

Finally we have the selection layer that go over the model probabilities for an event and decide if the fixture deserves 2+goals, over 2.5, home ML, away ML as well as double chance. This selection layer also tries to look for legs with decent edges but the problem is that most of the top European leagues have efficient markets. You'll likely find super tiny or almost no edges when bookmarkers have priced out the mstch6.

There are a couple of more things that go into the model too like fittings etc. After computing probabilities, all of the data including the probabilities data, goal grid, the stats that went into the model, etc andit is sent via an API request to openai models to get second opinions and also do it's acca construction to compete with the selection layer. For context, it doesn't get the selection layer data so it doesn't go into its own review with biased data.

Last week the selection layer produced 2.5 odd acca and the ai part gotten from openai produced 4.35 odds acca. The first acca lost and the ai bit won.

Currently 3 wins, 1 loss with this experiment and each acca is above 2+ odd and would likely be around 2.5 odd average.

[Makus]

At some point I think that this was overrated but now a lot of people are beginning to see that Ai doesn't really give anyone a guaranteed prediction. you can ask AI for a guide, analysis or idea but you cannot really get a game that one hundred percent sure because AI cannot read the future so that's the reason why I said it was overrated. People are now beginning to use AI according to what it was designed for and not a tool to give a guaranteed prediction. A lot of bettors get caught up and addicted when they start thinking that this is a sure way of winning.