A time machine that could actually deliver and tell us who would win mythical match ups... Is it more of a reality than we think?!

[billeau2]

Quantum computers will revolutionize artificial intelligence. It will do so by maxing out howmuch data we can crunch, making present computers glorified calculators. We often speak of matching fighters to see who would have won, we lament that we cannot do so... BUT we may have an option to do this after all!

because of the sheer number of data we will be able to crunch we can literally create the fighters and use predictive programming/probability to have them make decisions that they would make if they were still fighting... with quantum computers we can have a volume of data including everything from what age they were toilet trained, to when they would feint an opponent. And we have the means of proving this is so, easily... before ever putting on a mythical match up...

Its simple: create the program, and back test it using its predictive powers on fights we know the result of. When the program is right everytime, there is no reason to assume it would not be right on a hypothetical fight! So we use the program to predict every fight Floyd has had so far... we do this with 20 other fighters... and when it bats 100... Then we can ask things like "pep against mayweather" etc.

Thoughts?

[Willie Pep 229]

Quantum computers will revolutionize artificial intelligence. It will do so by maxing out howmuch data we can crunch, making present computers glorified calculators. We often speak of matching fighters to see who would have won, we lament that we cannot do so... BUT we may have an option to do this after all!

because of the sheer number of data we will be able to crunch we can literally create the fighters and use predictive programming/probability to have them make decisions that they would make if they were still fighting... with quantum computers we can have a volume of data including everything from what age they were toilet trained, to when they would feint an opponent. And we have the means of proving this is so, easily... before ever putting on a mythical match up...

Its simple: create the program, and back test it using its predictive powers on fights we know the result of. When the program is right everytime, there is no reason to assume it would not be right on a hypothetical fight! So we use the program to predict every fight Floyd has had so far... we do this with 20 other fighters... and when it bats 100... Then we can ask things like "pep against mayweather" etc.

Thoughts?

Need to think on it - but for now a cheeky reply.

When the machines take over they will want to preserve an allusion of consistent authority and will insist that the 1968 computer already gave us the correct answer, we were just too ****** to realize it - Marciano by toupee.

When the machines get that good they're going to lie to us and get away with it.

But I really do want to think through your post.

[Willie Pep 229]

How would it measure a fighter's strength, wouldn't that end up being subjective? Punching power, speed? The data going in has to be good data, free from bias and conjectures.

[Willie Pep 229]

The manner you suggest in testing the program sounds plausible.

[Willie Pep 229]

What if we give the wrong data but it balances out and gives us what appeaers to be a correct answer?

E.g. we give Dempsey too much power but not enough speed. The computer runs 10, 000 simulations ( 538 style) and the results keep coming back Tunney by decision with an occasional KD by Dempsey.

We then think OK the computer is batting close enough to be called 'batting a 1000.' But it turns out that our errors balanced themselves out regarding Tunney but we will never know if it affected a Dempsey-Frazier fight.

[HOUDINI563]

How Dempsey fought Tunney was very different fight one vs fight two. How Dempsey would fight Frazier would differ as well. I can’t see how any computer would be able to determine the specific strategy that each fighter would chose going into a fantasy fight nor how that strategy would change during a fantasy fight. These decisions many times are dictated by trainers not the fighter themselves.

[billeau2]

How would it measure a fighter's strength, wouldn't that end up being subjective? Punching power, speed? The data going in has to be good data, free from bias and conjectures.

heres the change. A lot of what a computer can figure out is limited by com*****g power... that means that given the resources it can take in more and more data and eventually literally create a fictional person. For example, those virtual reality games can recreate an environment. So the computer could extrapolate strength from physical stats, performance etc... we would know because it would predict all of Marciano's bouts we know about.

computers can obtain data in the reverse as well... so taking an "effect" of Marciano...like a KO and obtaining the "cause" (the strength of his punch). All this can be tested! So I go to a vegas punching machine and hit it and it records my punch strength... then the computer is given data about me (size, etc) then the computer is asked my punch strength... does it hit the mark? Its kind of scary actually lol.

[GhostofDempsey]

It would only be as reliable as the data entered by humans. We all know how unreliable that can be. Especially if math nerds who aren’t boxing fans are entering said data. The intangibles and nuances of each fighter could never be fully captured.

[billeau2]

What if we give the wrong data but it balances out and gives us what appeaers to be a correct answer?

E.g. we give Dempsey too much power but not enough speed. The computer runs 10, 000 simulations ( 538 style) and the results keep coming back Tunney by decision with an occasional KD by Dempsey.

We then think OK the computer is batting close enough to be called 'batting a 1000.' But it turns out that our errors balanced themselves out regarding Tunney but we will never know if it affected a Dempsey-Frazier fight.

Thats the thing about "Probability" and Quantum method, vis a vis game theory... there is no wrong data in a sense. All data represents possabilities that have an average... So if one programmed in "Tunney has three arms" the computer could very quickly ascertain the probability of that being so... So, we give tunney a chance, based on characteristics, that he KO's Dempsey... the computer would assume that Tuney and Dempsey fought every day, every hour and that eventually that would happen. yet it would look at the possibility that dempsey Ko's (the good data) as having a much greater probability of occuring in calculations.

The computer would be looking hollistically at all the data given to make these conclusions... so in a very real sense one would hve to feed totally bogus data from start to finish to stop the computer from working... it would be quite difficlt to do so because somewhere down the line the computer would see the real data as more probable.

In other words there would be no bad data... giving a fighter too much, too little, would show up as improbable when compared to other data committed.

[billeau2]

How Dempsey fought Tunney was very different fight one vs fight two. How Dempsey would fight Frazier would differ as well. I can’t see how any computer would be able to determine the specific strategy that each fighter would chose going into a fantasy fight nor how that strategy would change during a fantasy fight. These decisions many times are dictated by trainers not the fighter themselves.

Presently? impossible. BUT Quantum com*****g works on probability. There is always a possibility that a fighter would fight very differently, it simply is a possibility that eventually would occur... Your a high IQ guy so if you think this through you will see it. In other words... If I am going to visit you in New jersey and am coming from california and a quantum computer wants to predict how I will travel there. it will see how I generally travel, it will see how i make decisions about things, like when I married, etc, after a certain baseline it will have a general sense of the probability of me deciding anything in particular... then it will look at the probability of me deciding to swim across the world to get to you... somethng that ad infinitum could eventually be if we have the data to run enough possible outcomes... no doubt the computer will realize that I will fly to meet you.

Strategy as a quantum function is wholly a determination of many different possible outcomes and how they drive a specific outcome based on the relative outcomes in the decision. In ways we do not even know! maybe Frazier went at Foreman the way he did because he had a bad breakfast that morning!