BoxRec

OK, I'll do it. To make the results as accurate as possible I will sim it 1,000 times and post the results. BTW the new database is from June 2006. My bad.



Cube

The computer says:
(1,000 fights. Results converted to %)

Both fighters in top shape
Peter 40.9%, 7.5% KO
Toney 51.2%, 9.4% KO
Draw 7.9%

Toney undertrained
Peter 49.4%, 11.1% KO
Toney 41.1%, 10.1% KO
Draw 9.5%

So basically if both fighters are in shape, Toney takes a decision, or possibly a late KO. If Toney comes in over 230 or so, Peter by a very slight edge, with a slightly larger chance of a KO either way, and with an 8-10% chance of a draw. The pair matches up very close on paper. Of course much more data is available on Toney, so Peter has the chance to continue to evolve quicker than his Titlebout ratings. I'd say the Toney undertrained scenario is the one we'll see. I may run this again after the fight, because it will be more accurate with the proper referee and judges and ruleset which I can't really verify yet.

BTW Rufus the new database is available for download for free here http://www.ootpdevfiles.com/tb/fighters_2006-06-26.zip

and the directions to install are here http://www.ootpdevelopments.com/board/s ... p?t=125223

My own fictional Titlebout universe is here http://www.angelfire.com/games4/cubedrum0/

And finally the game itself http://www.ootpdevelopments.com/tb/

Hell thanks Cubedrum @ the new download info and all


Okies....so from 1,000 simulations it picks Toney over all to win if both men are in shape. So we will go with that. Toney to win via 12 round decision being the most likely and/or by late KO.

We'll have to see come fight night

Couple of points to point out-

1. Toney is 38 and has gone 3-0-1 (1), 1 NC as a HW

2. Peter is 25 and 26-1-0 (22) as a HW

3. Toney has fought from MW to HW winning titles in all those divisions

4. Peter's 'best' opponent thus far has been iffy-chinned Wlad Klitschko

5. Toney has out-classed and out-muscled guys like Holyfield and Ruiz and Rahman as a HW

6. Peter is arguably the hardest puncher currently in the division

7. Toney has been injured in training/fights last few times

8. Peter is still youthful and gets better each time out, he is not yet at his peak, nor is he a novice

The verdict? On paper alone, with Toney's credentials he should have an easy night with Samuel Peter, but Toney has never fought a man who hits as hard as 'The Nigerian Nightmare' either, and with his advanced age and worn parts...he could get caught.

In my opinion, if the computer is right [Toney wins by decision or late KO] then it should show that the computer is very accurate, because there is so many questions than answers about this fight that it is hard to give a clear-cut idea as to who would win.

Anyone's opinions on this fight? Or better yet your views on how accurate the computer has been so far or if you feel the computer simulation for the Toney/Peter bout is accurate?

So far the computer has predicted the winners of:

1. Maskaev vs Rahman- Maskaev KO12th

The computer on a best two out of three basis picked Maskaev to win by knockout.

2. Evander Holyfield vs Jeremy Bates- Holyfield KO2nd

The computer picked 3 out of 3 on this one for Holyfield to win by KO in three rounds or less [some call this a gimme but o well].

I have mentioned before in the MODERN DAY ATG HEAVYWEIGHT tournament that I have used this very same simulator to predict the winner of the Vitali Klitschko vs Kirk Johnson match-up ages ago with great success, but since nobody was there to witness the prediction made then there is no point to adding it in here.

The goal in mind is that this computer can pick at least seven out of ten winners before the actual fights happen, making it 75% accurate. I would love to see the computer do 10 out of 10, but I will remain realistic.

Hopefully I can get ahold of Cubedrum again and see if he can do the Nicolay Valuev vs Monte Barrett fight as well and get this project rolling.

IrishRufusMurphy wrote:Anyone's opinions on this fight? Or better yet your views on how accurate the computer has been so far or if you feel the computer simulation for the Toney/Peter bout is accurate?

So far the computer has predicted the winners of:

1. Maskaev vs Rahman- Maskaev KO12th

The computer on a best two out of three basis picked Maskaev to win by knockout.

2. Evander Holyfield vs Jeremy Bates- Holyfield KO2nd

The computer picked 3 out of 3 on this one for Holyfield to win by KO in three rounds or less [some call this a gimme but o well].

I have mentioned before in the MODERN DAY ATG HEAVYWEIGHT tournament that I have used this very same simulator to predict the winner of the Vitali Klitschko vs Kirk Johnson match-up ages ago with great success, but since nobody was there to witness the prediction made then there is no point to adding it in here.

The goal in mind is that this computer can pick at least seven out of ten winners before the actual fights happen, making it 75% accurate. I would love to see the computer do 10 out of 10, but I will remain realistic.

Hopefully I can get ahold of Cubedrum again and see if he can do the Nicolay Valuev vs Monte Barrett fight as well and get this project rolling.

You seem a bit shaky on percentages, Rufus. 7 out of 10 = 70%.

The Holyfield - Bates bout was not only a gimme but you didn't put the 'prediction' up until after the bout was over.

So far it has picked 1 from 1 (Maskaev vs Rahman) so at the moment is 100% accurate...

Shouldn't there be some kind of control? Or some kind of comparison? I mean I suspect I could have called the Holyfield fight. Also in boxing draws a generally a bit of a freak result (almost un-callable), so it's a 50-50 split at worst. Maybe we should compare the computer's prediction with some of Decagon's odds/predictions???

Ezzard wrote:Shouldn't there be some kind of control? Or some kind of comparison? I mean I suspect I could have called the Holyfield fight. Also in boxing draws a generally a bit of a freak result (almost un-callable), so it's a 50-50 split at worst. Maybe we should compare the computer's prediction with some of Decagon's odds/predictions???

Irish asked if the alleged experts and self-proclaimed historians amongst us would be willing to put in their predictions for Rahman - Maskaev. I warned him they wouldn't be brave enough to risk their 'reputations' by actually making verifiable statements. Surprise, surprise, only me and him made picks. He got it right (as did the computer) and I (and the many 'viewers' of this thread) got it wrong.

He's also asked for predictions on Toney - Peter. I have yet to see anyone step forward...

I look at this as not just an experiment but as an element of fun and debate. Everybody loved the Modern-Day ATG HW Tournament until the computer picked Marciano over Louis to put him in the finals with Ali.

Everybody asked me to back up how accurate the computer is, and I did prove that it could pick a winner in an up-coming fight. I am somewhat disappointed that nobody is willing to ever make a prediction on any of the fights that I am having the computer do...but that's ok.

I am running this computer program until it has done 10 fights, irregardless if it gets all 10 of them right or just a few right and wrong.

Myself if I was asked who would win between Toney and Peter, it is hard for me to call it than it was the Rahman/Maskaev fight. I do think Toney could win, but I do think the possibility of him losing is very great. I will, as before, stick by the computer prediction because I do think it is [based on probability] as accurate as it can get, but I do think peter has a great chance to pulling the upset.

Pater in my opinion is possibly the hardest hitting man since Earnie Shavers...Toney is no doubt the most skilled and entertaining of the HW's in the top ten. It's a toss up. But I will say Toney, though my instincts somewhat feels peter will do it.

IrishRufusMurphy wrote:I look at this as not just an experiment but as an element of fun and debate. Everybody loved the Modern-Day ATG HW Tournament until the computer picked Marciano over Louis to put him in the finals with Ali.

Everybody asked me to back up how accurate the computer is, and I did prove that it could pick a winner in an up-coming fight. I am somewhat disappointed that nobody is willing to ever make a prediction on any of the fights that I am having the computer do...but that's ok.

I am running this computer program until it has done 10 fights, irregardless if it gets all 10 of them right or just a few right and wrong.

Myself if I was asked who would win between Toney and Peter, it is hard for me to call it than it was the Rahman/Maskaev fight. I do think Toney could win, but I do think the possibility of him losing is very great. I will, as before, stick by the computer prediction because I do think it is [based on probability] as accurate as it can get, but I do think peter has a great chance to pulling the upset.

Pater in my opinion is possibly the hardest hitting man since Earnie Shavers...Toney is no doubt the most skilled and entertaining of the HW's in the top ten. It's a toss up. But I will say Toney, though my instincts somewhat feels peter will do it.

Feck, Rufus. Talk about having an each way bet...

I agree though that like Rahman - Maskaev it's not an easy fight to pick. You can make a case for each man. That's the fun of boxing. I can understand our resident 'experts' not wanting to put their alleged reputations on the line here.

I'll go for Toney to win on points. Sure Peter can bang a bit but you need a bit more than that to beat Toney even if he is 50 pounds over his natural fighting weight.

I pm'd Cubedrum and also sent him an e-mail for him to do a run on the up-coming VALUEV vs BARRETT fight. Hopefully he will respond back in time for the fight and all.

At the end of this tournament, depending on how accurate the computer is after 10 fights [7 fights to be our goal] I will have the main event of the MODERN DAY ALL-TIME HEAVYWEIGHT tournament run the same way that Cubedrum has done for the Peter-Toney fight, and maybe even get it saved a have the fight to be viewed round by round on here, if that is at all possible

The main event determined by the computer ended up being ALI vs MARCIANO, after both men bested through the ATG PUNCHER and ATG BOXER tournaments.

Then when that is accomplished I will have a friend of mine make an ATG HEAVYWEIGHT plaque of whoever wins and have it posted as that man being the 'Greatest' of all time.

I plan on having similar tournaments done at LHW and MW and make similar ATG plaques for them, and hopefully manage to talk the moderators in this forum to make a sticky thread showcasing the ATG HW, LHW and MW champions, as determined by the computer; which depending if the computer simulator can do at least 7 out of 10 or 10 out of 10 on upcoming matches, will show that this is the accurate result.

The Peter-Toney fight is just two days away...anyone else wanna make any bets on the fight? or u all gonna stick to ur guns and not participate?

Btw I looked up the votes on the Peter-Toney fight on FIGHTNEWS and it seems to be close on the numbers, with 45% of Toney winning on points, 37% Peter by KO, 12% Toney by KO and 6% Peter on points.

Almost the same as the computer's predictions with numbers of:

Both fighters in top shape
Peter 40.9%, 7.5% KO
Toney 51.2%, 9.4% KO
Draw 7.9%

Toney undertrained
Peter 49.4%, 11.1% KO
Toney 41.1%, 10.1% KO
Draw 9.5%

Whatever happened to just using a Magic 8-Ball?

I would think a computer based program based around the variables used in the original computer tournament in 1967 and probability methods through the simulator would be more advanced than a 'Magic 8 Ball'.

Anyways...Cube just replied back to me and he should shortly post either the computer's predictions on Valuev-Barrett in here or through via e-mail.

Every time a fight approaches on the horizon I get so excited to see if technology has come along well enough to pick winners in fights, and considering it proved to predict Maskaev over Rahman as well as Holyfield over Bates [gimme or not] it gives a good idea that MAYBE this isn't so far off.

On another note, my friend is currently in the works of making the ATG HEAVYWEIGHT title plaque for the winner of the MARCIANO-ALI fight that will be held after the testing of this computer program. I have already wrote up, as well, 12 different names for both the MW and LHW tournaments, if the computer program can be proven to be accurate [7 out of 10 is the goal].

And there will also be plaques made for those winners of those computerised tournaments. I plan to have tournaments made for all the divisions [at least the original divisions] and eventually talk the mods of this forum to make a sticky showcasing the ATG champions.



Anyone still want to make predictions on the Toney-Peter fight that happens tommorow? If not, whose going to be watching the fight?

generic screen name wrote:Whatever happened to just using a Magic 8-Ball?

Come on, don't be shy. make a pick. It doesn't matter if you are wrong. There are no experts here.

Ok, I will pick Toney by decision. (though i'm not too confident in this pick)
Enough of whether Toney will be in shape or not, all, who is the computer picking? Or this is this going to be like the Holyfield fight and we will be told after the fight who the computer picked ? :)

No the computer picked Toney to win ^^^^ as said in the replies above.

Ambling Alp wrote:Ok, I will pick Toney by decision. (though i'm not too confident in this pick)
Enough of whether Toney will be in shape or not, all, who is the computer picking? Or this is this going to be like the Holyfield fight and we will be told after the fight who the computer picked ? :)

The computer is picking Toney on points.

The extraneous crap around it is the 'excuses' clause in the event of that not being the result. You know, "Well the computer did also pick Peter some of the time...".

Anyway, its pick is Toney W12.

So far, me, you, Rufus and HAL are all picking the fat fella. And all going for a points victory.

Well Toney weighed in at 233 so that's hardly a trained Toney, though he has been bigger.

I'd say going by these numbers....

Both fighters in top shape
Peter 40.9%, 7.5% KO
Toney 51.2%, 9.4% KO
Draw 7.9%

Toney undertrained
Peter 49.4%, 11.1% KO
Toney 41.1%, 10.1% KO
Draw 9.5%

The computer predicts a Peter win moreso than Toney, in what is almost regarded as a 50/50 fight. That's a solid prediction I think although there's no mention of Peter undertrained (which he is)

I'm stuck on the fence with this fight, but I picked Peter by decision in "Pick The Big Fights"

Thanks for those updates Cube

So on the 'betting' predictions of the BoxRec members who did vote, thus far its more or less 4-1 odds that Toney will win, based on votes.

The computer predicts it a much closer fight than 4-1, since its numbers are 51% for Toney and 40% for Peter [esimates], but still an 11 point difference, while if Toney is undertrained Peter is now 49% and Toney is 41%, an 8 point difference.

But you have to figure, even if Toney is undertrained, based on all his experience and skills alone---he would slightly still be the favorite over the hard punching but limited Peter.

Remember Toney has faced champions in various weightclasses and thus far has proven to take on the HW's top men, while Peter could not beat the iffy chinned and flawed Wladimir Klitschko---but he did drop the 245 pound Ukranian giant three times in 12 rounds.

So far Toney has shown he could take a punch against the likes of a faded Holyfield, former 2x champion Rahman and former 2x champion John Ruiz [who floored Rahman and lost controversial decision to Valuev].

But no one can disagree that 'The Nigerian Nightmare' hits harder than any of the men listed above, will that make a difference? It could, considering men like Earnie Shavers, as limited as he was, could drop Ali and Holmes and Norton and Young, guys who were light years better than Shavers was in skill and experience.

50-50 odds? Maybe. It's that close a call. But we will all see come fight night.

Note- Irregardless if the computer is proven right or wrong in it's predictions for this fight, the testing of the TITLE BOUT simulator will continue as planned until the goal of 7 out of 10 is accomplished or proven to be nothing more than a joke.

Here's the upcoming Valuev - Barrett fight simmed on Titlebout Championship Boxing 2

(1,000 fights. Results converted to %)



Both fighters in top shape

Valuev 62.5% KO 34.3%
Barrett 32.4% KO 12.4%
Draw 5.1%


Looks like Valuev's physical tools will just be too much for Barrett. Barrett has a better conditioning rating so I'm assuming that he always comes in trained, but Valuev's is above average too. Only simmed for both men in top condition.

Thanks Cube So the computer, based on top conditioning, picks Valuev to win against Monte Barrett when they square off.

Tonight will be the Peter-Toney fight...and I am sure everybody will be anxious to see if the computer is accurate enough to determine the winner of that match, like it has done in the previous two other matches.

I will post asap the results of that fight when it happens.

Things to consider:

1- Toney is 233, lighter than he was for Hasim Rahman, is 38 and is facing arguably the hardest hitting man since Shavers. No matter how you look at it, this will be possibly Toney's toughest test as a Heavyweight.

2- Peter is much younger, far more stronger, but has shown lack of stamina and a decrease in knockout power as he stepped up into the better part of the division. He is still green, but with two KO's since his loss to Klitschko he should be confident.

What Toney needs to do is use the center of the ring to his advantage, and stay away from those ropes. Toney is elusive enough to get close and duck punches, but that was against a faded Holyfield, Ruiz, Rahman and Williamson, not a thunderous puncher like Peter.

What Peter needs to do is stay on his man and go to the body, no head hunting until he has slowed down Toney---he is facing the fastest man and no doubt his most experienced and skilled opponent to date. He has to use that 257 pounds to his advantage and wear down Toney.

Samuel Peter makes the upset winning a 12 round decision over James Toney, breaking the computers 2 fight win-prediction streak.

Not many of the experts would have predicted Peter winning via decision.

But the computer is right [considering Toney is 233] that Toney could have been undertrained for this fight, as it did give Peter the edge by 8 points.

Toney did manage to win one out of the three judges score cards, but nevertheless the "Nigerian Nightmare" proved the critics wrong, and I should have stuck to my instinctive feeling that Peter would pull the upset.

Nevertheless this computer testing will continue as planned.

Computer is now 2-1 in it's predictions.

IrishRufusMurphy wrote:Samuel Peter makes the upset winning a 12 round decision over James Toney, breaking the computers 2 fight win-prediction streak.

Not many of the experts would have predicted Peter winning via decision.

But the computer is right [considering Toney is 233] that Toney could have been undertrained for this fight, as it did give Peter the edge by 8 points.

Toney did manage to win one out of the three judges score cards, but nevertheless the "Nigerian Nightmare" proved the critics wrong, and I should have stuck to my instinctive feeling that Peter would pull the upset.

Nevertheless this computer testing will continue as planned.

Computer is now 2-1 in it's predictions.

Rufus,

Regarding the Holyfield - Bates gimme.

Did you or did you not put the computer's pick up AFTER the bout had taken place?

Did you even ask anyone if it should be included?

Using the agreed formula, the computer is 1 from 2 so far. That's 50% in case you were wondering...

You don't want upset the few people who are entering into the spirit of things by fecking it up, mate.

IrishRufusMurphy wrote:Samuel Peter makes the upset winning a 12 round decision over James Toney, breaking the computers 2 fight win-prediction streak.

Not many of the experts would have predicted Peter winning via decision.

But the computer is right [considering Toney is 233] that Toney could have been undertrained for this fight, as it did give Peter the edge by 8 points.

Toney did manage to win one out of the three judges score cards, but nevertheless the "Nigerian Nightmare" proved the critics wrong, and I should have stuck to my instinctive feeling that Peter would pull the upset.

Nevertheless this computer testing will continue as planned.

Computer is now 2-1 in it's predictions.

Well actually the computer picked Peter to win since Toney was coming in undertrained. What's the point of having the feature of "undertrained" if we're not gonna use it. I'd say it got the Peter-Toney fight correct...and the percentages were very close, 10% for the draw was it? Very reasonable.

Also Collins, I ran the test on Holyfield-Bates last week (with software from 2005) and it predicted a 4th round KO for Holyfield.

A good case could be made that the sim is 3-0, That is 100% incase you were wondering