Experiment failed (Gambling on the basis of AI)

[Churchillvv]

Not to create long story! however I did experiment one of my long waited biases yesterday and it did turn out to be some sort of good guess but later failed.

On the basis of AI, I took a chance at Deepseek promoting it to give me a close guess or tips on what team to win football matches yesterday and somehow it gave a good tips with statistics that gave me a chance to be rational in decision making but because of greed and the payout was quite small so I added my own guess to increase the chance of high return and it turnout that my own guess ruined the bet.

Curious to ask, did anyone ever try this out? it could work and also fall.

[Agbamoni]

Since the predictions giving by deepseek was successful until you ruin it with your own bet, also know that it is not always like that. Sports betting demands two things to win one is skill and the other is luck. Your skill is as result of your ability to make informed research on the teams you are betting on. If you watch sports always it will be a lot easier. What deepseek ws able to do for you was to do quick research by surfing the web to know possible outcomes. Its result might be a shuffle from other prediction website and sports market analysis. Don't think next time deepseek will be a successful game again. Which is why you should bet with an amount you would have bet with if you predicted the game yourself to stay financially safe.

[Tungbulu]

Curious to ask, did anyone ever try this out? it could work and also fall.

I did try out something like this before, I made a few picks from different leagues and asked AI to give me a possible outcome to the games. AI ran a quick analysis on the teams, which included their previous performances, H2H, current position on the league table and a few other tools and then came up with a potential outcome for the games.
The result? All the games went accordingly as predicted except one, the result turned out to be quite the opposite and it made me realize just how unpredictable gambling is, even for the professional gamblers, and that luck is all that winning is actually dependent on luck.

[swogerino]

Not to create long story! however I did experiment one of my long waited biases yesterday and it did turn out to be some sort of good guess but later failed.

On the basis of AI, I took a chance at Deepseek promoting it to give me a close guess or tips on what team to win football matches yesterday and somehow it gave a good tips with statistics that gave me a chance to be rational in decision making but because of greed and the payout was quite small so I added my own guess to increase the chance of high return and it turnout that my own guess ruined the bet.

Curious to ask, did anyone ever try this out? it could work and also fall.

I have never tried Deepseek as it asks me to register while I have tried anonymously ChatGPT and the results were catastrophic in a bad way of speaking as I told it to give me the best games on a Saturday from the 5 major leagues, it kept giving me favorites which anyone knows that they should win and it did not do a thorough analysis at first. I asked it again to give me insights on why it thinks these are going to be winning tips and the analysis was not good either. This shows we are a very long way from having AI influence us in the choices we make in sport betting to give us an upper edge. It needs much more training than it already has right now, I acknowledge though we are in early stages of AI so these experiments we make are not mature at all.

[Inwestour]

Not to create long story! however I did experiment one of my long waited biases yesterday and it did turn out to be some sort of good guess but later failed.

On the basis of AI, I took a chance at Deepseek promoting it to give me a close guess or tips on what team to win football matches yesterday and somehow it gave a good tips with statistics that gave me a chance to be rational in decision making but because of greed and the payout was quite small so I added my own guess to increase the chance of high return and it turnout that my own guess ruined the bet.

Curious to ask, did anyone ever try this out? it could work and also fall.

I have never tried asking AI for betting advice, I think it is capable of making an average forecast based on past results, and will also rely on the favorite in relation to the bookmaker's odds. As for your experiment, here you also cannot rely on just one game day, the statistics on the results need to be assessed at least for several weeks or months, only over a distance will you be able to understand how good a bettor the AI ​​can be.

[acroman08]

Curious to ask, did anyone ever try this out? it could work and also fall.

I remember Ognasty posted a thread where he tried using AI to pick the winning team for him, and the last time I saw the thread, he was on a win streak, but I don't know if those were his last bets or if he continued betting after that.

So I looked for the thread Ognasty posted, and from the looks of it, the total result of his bets is 9/16.
Artificial Intelligence Gambling Using Grok! (Results: 9/16)

[mindrust]

I used deepseek and chatgpt for other purposes and i think deepseek is way too simple to be called “ai”. Compared to chatgpt (free version), deepseek looks like a glorified search engine, nothing else.

Chatgpt can do deep analysis and come up with its own ideas on any subject. I didn’t see deepseek doing the same thing.

I am not sure if chatgpt will be a more successful tipster. I think it is worth a try but I definitely know deepseek is no good.

[BABY SHOES]

Never tried experimenting betting with AI (Deepseek or ChatGPT) it has never been done... because often in sports betting always rely on their own skills to find better opportunities in the bet.

YEAH although some say it can succeed or fail I myself have not tried it.
This in my opinion lacks adrenaline if it only relies on deepseek or other AI.
Even if there is a winning streak maybe it won't last long, right?

[Stalker22]

OP, Why do you call it a "failed experiment"? It sounds to me that you blew the experiment with the introduction of your predictions, so you couldnt get conclusive results. You had something interesting going with the AI's insights, right? It gave you statistical backing, a rational basis for your bet. Thats the experiment! Then, if Im understanding correctly, you couldnt resist throwing your own gut feeling into the mix because the payout wasnt big enough. And thats where things went south. So, it wasnt a failure of the AI experiment, it was a failure to stick to the experiment, right?

[Hispo]

Honestly, this is not the first time I have read or heard about people trying to use Artificial intelligence to enhance their chances to get money from casinos, specially from sportbetting, so far, all experiments I have seen in this community have never given positive results back to their gamblers.
Artificial intelligence is useful to analyze and give a prediction based on publicly available information, information which can be used by anyone to do their own analysis, but with AI is faster. But AI cannot tell the future or filter out entropy one could easily find in football matches, for example, where there are so many factors which can affect the final result of the game.

If someone managed to score some predictions through the use of AI, I would be more inclined to call those coincidences and have nothing to do with a hidden or specially ability on an algorithm to predict the future in sport betting.

[rachael9385]

Not to create long story! however I did experiment one of my long waited biases yesterday and it did turn out to be some sort of good guess but later failed.

On the basis of AI, I took a chance at Deepseek promoting it to give me a close guess or tips on what team to win football matches yesterday and somehow it gave a good tips with statistics that gave me a chance to be rational in decision making but because of greed and the payout was quite small so I added my own guess to increase the chance of high return and it turnout that my own guess ruined the bet.

Curious to ask, did anyone ever try this out? it could work and also fall.

The same to what @Agbamoni have said, I won't add much to it because he's right. Yes Deepseek gave you a nice predictions with through statistics but you added more games that ruined the whole bet. However I think that's the cause of greed because if not for greed, you would have bet the game the way you received it. Don't reply on Deepseek predictions because you will not get the same results.

[Slow death]

I have never used Deepsee and to be honest I wouldn't. I prefer to rely on my own analysis, which gives me greater security and fun. If I were to use tips from AI or other people to be able to place a bet, then it wouldn't make sense to say that I'm having fun with sports betting. AI in sports betting won't be able to replace analysis done by people any time soon. Finally, be careful with greed and manage your bankroll well, sports betting should not be seen as a source of income.

[Fortify]

Not to create long story! however I did experiment one of my long waited biases yesterday and it did turn out to be some sort of good guess but later failed.

On the basis of AI, I took a chance at Deepseek promoting it to give me a close guess or tips on what team to win football matches yesterday and somehow it gave a good tips with statistics that gave me a chance to be rational in decision making but because of greed and the payout was quite small so I added my own guess to increase the chance of high return and it turnout that my own guess ruined the bet.

Curious to ask, did anyone ever try this out? it could work and also fall.

There is definitely potential for AI to disrupt the equilibrium and advantage that sport bookmakers have over the standard punter in future. Up until recently the main stumbling block was how out of date the information was that fed AI systems - it was a year or two behind, but that gap has shrunk. There are most likely people out there who have found and are using strategies to data mine using AI. Particularly in areas with high odds involved and a wide spread that can change over a few days, like horse racing. It will be just as adapt at finding patterns and mispricing as the sophisticated tech that bookmakers are using, but it's how to feed it the latest information to extract value that is probably the remaining hurdle.

[Cantsay]

From what you wrote and what I understand from the OP, I won't call the experiment a failure because the AI predictions were correct; it was the few other bets that you included in your bet that made everything not come out well -- if you had stuck with the ones the AI gave you, you wouldn't have had any issues.

In addition, I have done some analytics research with AI in the past, and I have even asked them for teams that they believe have more advantage of winning a game, and the AI response was based on information that was found online - so if anything happens urgently that was exposed to the internet, you won't be able to get a good prediction from the AI, and also if the AI uses outdated info, it might also lead to your bet failing, and these are things that you should have experimented with.

[Obim34]

OP, Why do you call it a "failed experiment"? It sounds to me that you blew the experiment with the introduction of your predictions, so you couldnt get conclusive results. You had something interesting going with the AI's insights, right? It gave you statistical backing, a rational basis for your bet. Thats the experiment! Then, if Im understanding correctly, you couldnt resist throwing your own gut feeling into the mix because the payout wasnt big enough. And thats where things went south. So, it wasnt a failure of the AI experiment, it was a failure to stick to the experiment, right?

We can talk about greed from his side later, let's first talk about using AI to make predictions which is something diabolical to do when we don't trust our analysis on events, just a run test i know to see how efficient it is so i think OP score may be right this first time but won't for the majority of times AI will give her predictions. I believe that for gambling to be interesting, predictions must come from the gambler, not from a punter or AI.

OP, how many events were included on the list given to AI prediction? A single event cannot be used to judge on this, multiple events accumulated should be submitted to AI for more credibility of predicting correctly.

[BitMaxz]

Why use AI if you can analyze the game and the team you choose?
I tried the AI today. It seems that this DeepSeek has old data. I asked the AI about the current crypto market condition, but the result seems to the last year's data. It said the current price is $63k and the resistance is under $65k.

So if you are going to use it to analyze your bets, I think it will only give you outdated statistics, not good if you usually follow its predictions.

[ryzaadit]

Sometimes tried this.

But, sometimes the data they're provided is outdated. Example, using the data performance from last years rather than recent matches or a few recent matches. I once tried, and ask the AI to input the resource too.

Suprising for me, the data being used as reference 2023 ~LOL

[bitbollo]

... it could work and also fall.

There is 50% of the chance that guess the winning team in a choice A and B.
Of course it can be profitable in some way, but I didn't expect a super-technology that could outperform the market. At least AI must be trained
Actually, there are already alghorythm and professional tools able to get good performances.
Probably AI could be just an "interface" and is not adding nothing new.

[Alphakilo]

On the basis of AI, I took a chance at Deepseek promoting it to give me a close guess or tips on what team to win football matches yesterday and somehow it gave a good tips with statistics that gave me a chance to be rational in decision making but because of greed and the payout was quite small so I added my own guess to increase the chance of high return and it turnout that my own guess ruined the bet.

The experiment failed because you introduced your own bias. Or maybe it didn't. To know the true outcome of this, you will have to repeat the experiment. This time do not introduce your own bias. Also repeat the prompt of several games for 5 times. Use only money you know you can afford to lose. Of course you already know this. Then report back to use.

Do not be eager to report..This could take a couple of weeks that is only when you can conclude that it is a truly failed experiment.

[Woodie]

This whole AI thing isn't so much different from a coin toss in my opinion , but I must say... depending on what historical number's the AI was using to predict an outcome, I think 6 out of 10 times it will arrive at the correct team to win unless it's an upset or other team has nothing to play for. Also fixture will matter, because if it's a game which has bookies grade a team for saying 1.2 odds then expect AI to be correct on this one, but where it gets to say 2.5 odds Vs 2.8 odds these will be a 50:50 thing even for punter's.

Otherwise Man wins over AI, anyday!