Topic: UFC Fight Night: Gane vs. Spivac - AI prediction (Read 83 times)

With the rate of 8/11, it's already enough to make money even if you just pick what the AI gives. No betting on which round or winning method. If in the next UFC 293 prediction of this AI will still have high rate, I will really believe this is one heck of a tool for non-MMA fans.

Just test it with just $110 and bet $10 each. If the fighter is known as KO artist I guess you can try the winning method just to get more from the odds. I will closely watch it this weekend.

UFC 293: Adesanya vs. Strickland
MAIN EVENT
Israel Adesanya VS Sean Strickland

CO-MAIN EVENT
Tai Tuivasa VS Alexander Volkov

MAIN CARD
Manel Kape VS Felipe dos Santos
Justin Tafa VS Austen Lane
Tyson Pedro VS Anton Turkalj

PRELIMINARY CARD
Carlos Ulberg VS Da Un Jung
Jack Jenkins VS Chepe Mariscal
Jamie Mullarkey VS John Makdessi
Nasrat Haqparast VS Landon Quiñones
Blood Diamond VS Charlie Radtke
Shane Young VS Gabriel Miranda
Kevin Jousset VS Kiefer Crosbie

this is what i am seeing here also about the results. if you won't consider the round or the method of winning, AI has high chance of getting it correct. so maybe if you want to use the AI predictions, just bet on who will be the winner and don't bet on other betting lines such as over/under, winning method and what round. i can say, that's high percentage of getting it right about the overall winner.
and don't do multiple bets, maybe you can gain more profit via single bets here.

8/11 picked winners eh? That's a good deal I guess. I would have never had that many if I did it on my own. My take away on this is to consider the prediction on who would win but not put too much faith on how they win or what round the fight finishes. It's like "safer" to put your money on ML when using AI.

It's understandable if this is the first time that you've used AI but if you can continuously do this you can come out with a near-perfect prediction, There's a big possibility that many sports bettors are already using AI, and a few years this is going to be the new trend, this is why they create these AI as a tool for in-dept analysis. 

I still don't think that you should rely everything on AI, it's good for comparison but there's nothing like a human analysis because we have what we call intuition and it has a bigger impact than AI analysis, I have proven that intuition sometimes plays a major role in our decision making.

Conclusion: I can't make any conclusions from this. I would need extraoridinary large sample size and a bot to make enough predictions.
But it was a fun experiment and i lost money i experimented with, as i did multi bets with that i would have lost with only one error.
Anyway, i'll add this data to my spreadsheet for the future and might do some other game next.

Yes, and we gamblers never thought that AI can predict the fight as what humans do, I mean we look at everything when we do our analysis, like previous fights, the weight and others. So not sure how AI predicted this fight.
But as what you have said, it's very interesting to see how the predictions will be or what will be the accuracy of AI.


Most likely there are already some sports bettors that have tried this before.
But first time that we have heard this one here because the OP posted his experiment and then we will see how it goes to him and others who are going to look at this fight prediction and bet.

This Bing has access to a lot of MMA data making the AI compare from sherdog to tapology.
Cool to see an AI has the idea of which one will win over a bout and it suggests which round the fighter may win. Since this is a search engine, I was expecting g to see an ad related to Bookmaker but there seems none.

OpenAI didn't have such a feat when I tried it months ago. I couldn't agree with a few of its predictions but the majority is okay.

I laughed at some of the fighter and match up info because some of them are wrong.  But interesting experiment...  All in all not a bad idea.   

And if you can, keep records of wins and losses made by the bot if it made bets of 1 unit at every fight and see how it performs after 6 months or so.  Anything above 55% will show that it's a profitable strat.

This is very interesting I never believe in the accuracy of AI when it comes to sports betting, but since you posted it here we'll see if these AI can give better results than humans.

This could be the new trend here in gambling discussions posting AI predictions against human predictions, if AI offer a good alternatives many gamblers will turn to using AI for their bets.

Well i used small bets and did several combinations of multi bets with all of them.
I even used exact rounds and winning methods with some of them. I can't recommend doing it as the winning change is less then slim, but it was for fun anyway and to keep my interest up on this.

I have no expectations to win, but if i do, i would get my money back with pretty sweet multipliers.
If i get decent results i will try this again, maybe with some other sport.

Very interesting, we will see who will get the win and find out if AI can really predict games or fights.

I check one sports bookies and this are the odds,

Farid Basharat vs Kleydson Rodrigues   *   Basharat will win by submission in the third round   -   Basharat favorite, ML @1.25
Loosa Ange vs Rhys McKee   *   McKee will win by knockout in the second round   -   McKee underdog @2.30
Nora Cornolle vs Joselyne Edwards   *   Edwards will win by knockout in the first round   -   Edwards favorite ML @1.79 (but it's very close)
Zarah Fairn vs Jacqueline Cavalcanti   *   Cavalcanti will win with a TKO or a submission.   - Cavalcanti favorite ML @1.25
Taylor Lapilus vs Caolán Loughran   *   Lapilus will win by decision   -   Lapilus favorite ML @1.55
 Morgan Charriere vs Manolo Zecchini   *   Charriere will win by KO/TKO in the second or third round   -   Charriere favorite ML @1.26
William Gomis vs Yanis Ghemmouri   *   Gomis wins by knockout on 2nd round or decision.   -   Gomis favorite ML @1.43
Volkan Oezdemir vs Bogdan Guskov   *   Oezdemir wins by KO/TKO in the first or second round.   -   Oezdemir favorite ML @1.51
Benoit Saint-Denis vs Thiago Moisés   *   Moisés would win this fight by decision or late stoppage.   -   Moisés underdog @2.26
Manon Fiorot vs Rose Namajunas   *   Fiorot will win by decision   -   Fiorot favorite ML @1.49
Ciryl Gane vs Serghei Spivac   *   Ciryl Gane wins by knockout in 4th or 5th round or decision   -   Ciryl Gane favorite Ml @1.52

---

So based on the predictions by AI, there are 2 underdogs that we need to watch out. So that is the fight that we should look as well and the underdogs are well over 2.3x. Maybe throw some money on them?

It might be a good idea to do some parlays as well? What do you think?

After ChatGPT came out, i've seen some threads about AI and if it could help in gambling.
Back then, AI:s couln't access to internet so they couldn't find latest stats to compare, but now it's a whole new ball game.
While i am highly skeptical about results i thought i might give it a go to try it.

I admit that this is a fraction of any sample size that would be needed to come to any conclusion if it's working but i thought it would be fun little experiment.

I chose Bing AI chat creative mode as my tool, as it seems to be most efficient free tool for this kind of stuff. I tried also google bard, while it had some insightful opinions about some matches, it didn't have any data about most of the fighters so it couln't make any predictions. Basically my prompt for ai was "if these two would fight, who would win and how, and can you give me details how you came that conclusion"

I haven't check if the data is legit, as AI's have an issue of inventing things that doesn't exist (lying), you should take the results with the grain of salt. Prediictions doubly so. I myself did some small multi bets from these predicions. Not too much because i don't want to trust my money on this silly experiment, but some to have some skin in the game.

Full details were too long and specific to post in here, they contained links to articles, records and statistics, fighting styles, advantages and disadvantages, etc, so i asked it to summarize it for me afterwards. But i put links to original long answers and they are most interesting part of this and definitely worth checking out.

Links to full, detailed answer won't work immediately, you have to wait after a click a bit for them to appear in bing search page.

I will update the results next to prediction recap at the end to see if any of these were even remotely correct. I'll do it Sunday, as i am offline whole Saturday.

But here we go; summarized predictions about that fight night ->