BoxRec

IRM

I'm on your case a bit here but I think there are a lot of things that need pointing out. Your idea of a compuer tournament is a good idea and fun, of nothing else, but there are some serious pitfalls.

Your last post perfectly illustrates my point.

Firstly how do you know how Title Bout arrive at their ratings for Tunney?

Secondly, you believe Tunney is overrated in the game. I believe Marciano is overrated in the game. So who is right? the game, you, me, none of us??? If you accept their ratings when they suit you and reject their ratings when they don't then we are merely playing the ratings game that we all enjoy so much on this forum. there's nothing wrong with that but we can't claim that the game has done anything other than reflect our beliefs on who is better.

Firstly, I don't under-rate Tunney. What I have ever pointed out is that Tunney's bulk of his career was at LHW and he beat a shot Jack Dempsey, after beating a green Johnny Risko and Tommy Gibbons, who was also a LHW who fought as a HW on occassion.

I don't particuarly look at the rankings on TITLE BOUT on an ATG level, but in my opinion a fighter should be judged strictly at the weight class and era that they fought in. In Tunney's case I place him in the top 15 LHW's of all time, but as a HW I place him even lower, as his career as a HW was far too brief to really prove anything, despite winning the title and retiring as champion.

ENough on Tunney.

As far as if I believe TITLE BOUT rankings are greater than other rankings, I will not say that. But strictly on a fantasy fight basis, in my tournament, I matched up 24 of the best HW's and the TITLE BOUT simulator used virtually the same amount of variables Murry Woroner used, looked at the fighters records, judged by the eras they fought, and determined who would beat whom.

I think many would agree with me that having Ali and Louis in the top three [as my tournament said] is far more accurate than BOXREC's that has Ali as number one and Joe Louis as number five, with Gene Tunney, whose career was bulked as a LHW as number two?

That would be like saying if James Toney won a HW title, that due to all his past accomplishments at lower weights, belts and all, would make him the best HW of all time, because of what he did at lower weights.

No, to judge a fighter correctly you must judge by the weight class, the era and who they fought. Take Toney as an example, ok? As a HW Toney beat a 40 year old Holyfield, drew Hasim Rahman, tested positive for steroids against John Ruiz, beat mediocre Davaryl Williamson, lost to slow, ponderous Samuel Peter. A pretty decent record as a HW, but just because he beat Holyfield does that make him a great fighter? No.

The simulator judges not on rankings so much, but rather on the 129 variables, as well as who they fought and how well they did in their era. Take for example Marciano, he never lost, but his era wasn't exactly the 1970's either. It takes away from him, even though he never lost. That's how the computer judges. With facts, and no human elements to detour it.

Do I personally rate Marciano higher than most? Yes and I will debate and debate my reasons why. But I ain't the one determining the outcomes of these fights, and Marciano didn't need me to catapault him into the finals of this tournament, it was his accomplishments that the computer reviewed and it was the era and the 129 variables it took into consideration that got him there.

Yes, I take this whole thing in fun, and I know that despite how accurate this simulator ever proves to be, and despite who is the winner in the end, I know the debates of who is the greatest, who is the best will never stop. I was offered a challenge and I took it, and this computer has proved to be remarkably accurate.

Again, I will say this, I believe the majority would agree to my findings in my tournament with the top three being Ali, Marciano and Louis. Despite the computer saying Marciano would beat Louis [which was the catalyst for the whole testing to begin with as many said it was a bullshit decision] the top three remain to be Louis, Marciano and Ali.

BoxRec's rankings? Ali, Tunney, Marciano, Johnson and Louis. And I bet at least a solid 95% of the people here would say Tunney has NO business being in the top five. Sure people could argue that Marciano doesn't either, but you could also make the same debates on any of the fighters on that list.

************************************************************

Sorry for debating/arguing/bitching/ranting/raving...

I would like to see the computer determine the final five fights to make this experiment complete. Speaking of which, here are a few to list off, to take into consideration:

Shannon Briggs vs Sergei Liakhovich-WBO HW title

Floyd Mayweather vs Carlos Baldomir- IBF WW title

Jermain Taylor vs Kassim Ouma- MW title

Markus Beyer vs Mikkel Kessler- WBC/WBA SMW title(s) October 14th

Wladimir Klitschko vs Calvin Brock- IBF HW title

IrishRufusMurphy wrote:Firstly, I don't under-rate Tunney. What I have ever pointed out is that Tunney's bulk of his career was at LHW and he beat a shot Jack Dempsey, after beating a green Johnny Risko and Tommy Gibbons, who was also a LHW who fought as a HW on occassion.

I don't particuarly look at the rankings on TITLE BOUT on an ATG level, but in my opinion a fighter should be judged strictly at the weight class and era that they fought in. In Tunney's case I place him in the top 15 LHW's of all time, but as a HW I place him even lower, as his career as a HW was far too brief to really prove anything, despite winning the title and retiring as champion.

ENough on Tunney.

As far as if I believe TITLE BOUT rankings are greater than other rankings, I will not say that. But strictly on a fantasy fight basis, in my tournament, I matched up 24 of the best HW's and the TITLE BOUT simulator used virtually the same amount of variables Murry Woroner used, looked at the fighters records, judged by the eras they fought, and determined who would beat whom.

I think many would agree with me that having Ali and Louis in the top three [as my tournament said] is far more accurate than BOXREC's that has Ali as number one and Joe Louis as number five, with Gene Tunney, whose career was bulked as a LHW as number two?

That would be like saying if James Toney won a HW title, that due to all his past accomplishments at lower weights, belts and all, would make him the best HW of all time, because of what he did at lower weights.

No, to judge a fighter correctly you must judge by the weight class, the era and who they fought. Take Toney as an example, ok? As a HW Toney beat a 40 year old Holyfield, drew Hasim Rahman, tested positive for steroids against John Ruiz, beat mediocre Davaryl Williamson, lost to slow, ponderous Samuel Peter. A pretty decent record as a HW, but just because he beat Holyfield does that make him a great fighter? No.

The simulator judges not on rankings so much, but rather on the 129 variables, as well as who they fought and how well they did in their era. Take for example Marciano, he never lost, but his era wasn't exactly the 1970's either. It takes away from him, even though he never lost. That's how the computer judges. With facts, and no human elements to detour it.

Do I personally rate Marciano higher than most? Yes and I will debate and debate my reasons why. But I ain't the one determining the outcomes of these fights, and Marciano didn't need me to catapault him into the finals of this tournament, it was his accomplishments that the computer reviewed and it was the era and the 129 variables it took into consideration that got him there.

Yes, I take this whole thing in fun, and I know that despite how accurate this simulator ever proves to be, and despite who is the winner in the end, I know the debates of who is the greatest, who is the best will never stop. I was offered a challenge and I took it, and this computer has proved to be remarkably accurate.

Again, I will say this, I believe the majority would agree to my findings in my tournament with the top three being Ali, Marciano and Louis. Despite the computer saying Marciano would beat Louis [which was the catalyst for the whole testing to begin with as many said it was a bullshit decision] the top three remain to be Louis, Marciano and Ali.

BoxRec's rankings? Ali, Tunney, Marciano, Johnson and Louis. And I bet at least a solid 95% of the people here would say Tunney has NO business being in the top five. Sure people could argue that Marciano doesn't either, but you could also make the same debates on any of the fighters on that list.

************************************************************

Sorry for debating/arguing/bitching/ranting/raving...

I would like to see the computer determine the final five fights to make this experiment complete. Speaking of which, here are a few to list off, to take into consideration:

Shannon Briggs vs Sergei Liakhovich-WBO HW title

Floyd Mayweather vs Carlos Baldomir- IBF WW title

Jermain Taylor vs Kassim Ouma- MW title

Markus Beyer vs Mikkel Kessler- WBC/WBA SMW title(s) October 14th

Wladimir Klitschko vs Calvin Brock- IBF HW title

I think Title Bout ranks fioghters on how they performed at that weight class. Mike McCallum at MW is not as good as McCallum at JMW, same for Hearns, Leonard, etc... You say that you don't underrated Tunney but how can you know that? I'm just using him as an example.

In you re-run Louis-Marciano over 1000 fights you will see that Louis comes out on top.You say thast there are no human elements to detour it and oncer the fight sim is under way you are right, BUT a human, or group of them, decided to rate Foreman as 13/13 for punch power (and likelihood to throw KO punches) and Louis 11/13, Tunney 5/13, Holyfield 8/13. This is the human factor.

You are doing a great job but I doubt whether the majority here would have Ali, Louis and Marciano as their top 3.

Irish,

I know I'm a sucker for asking this as you have stated you know feck all about how the simulator works, but what are the '129 variables' you keep mentioning?

I saw that interview with Tracy Callis on the Marciano - Ali dvd and he said he used something like 20 variables for his simulation.

I can imagine 20 but 129???????????????

What are they?

Please don't post 15 rambling paragraphs without answering my question. I'm only interested in hearing about the '129 variables' you keep mentioning.

Tracy Callis said he used 20 variables for his computer simulator. Murry Woroner used 129 variables [Hank Kaplan wrote down the various variables] that revolved around boxing and what boxers do.

For instance, let's take into consideration Muhammad Ali. His hand and foot speed was his greatest strengths, as well as lateral foot movement, ring generalship, and he proved to have a tremendous resistance to pain.

His weaknesses was that he carried his hands too low, didn't pack a powerful punch, had less than good defense against swarmers [his infighting ability was his weakest point].

Now let's say Ali was faced against Joe Louis, who was a better tactician, had more power, could box better than Ali. But you could say that Ali could beat Louis, because if Louis had problems with Billy Conn, then Ali is Billy Conn with more speed, more power and forty more pounds bigger.


As far as the variables are concerned, I don't know every single one off the top of my head, but here is a few articles from Hank Kaplan who drew up the 129 variables and gave them over to the NCR company to transform those variables into mathematical formulas, and then the computer would create one gigantic mathematical equation that would turn out the 'script' for the computer fights.

http://www.miaminewtimes.com/Issues/199 ... ure_2.html


I will say that the computer uses virtually every aspect of the game as a variable, from punching power to cut resistance to speed and stamina to desire and determination, as well as the various styles and movements that a boxer does [jabs, hooks, bobbing and weaving etc].

For example, Jack Dempsey bobbed and weaved [2 variables there] and was a swarmer [another variable]. He had great punching power [another variable] had a solid chin [another variable] and rarely cut [another variable]. He was very determined [another variable] but seemed to slow down in the middle rounds [another variable] and he could get up off the floor to win [another variable].

Just that short a sentence on Dempsey yeilded nine variables. Also, film of these peoples fights were fed into the NCR computer and it judged how the fighters reacted when their opponent tried to retaliate---let's say when someone try a jab [variable] at Dempsey he would bob down [variable] OR if someone threw a hook [variable] at Dempsey he would weave [variable] out of the way.

Starting to understand? If not here is the TITLE BOUT boxing website and it goes into variables that the computer takes into consideration, though I don't think it lists every single variable.

http://www.ootpdevelopments.com/tb/index.php

IrishRufusMurphy wrote:Tracy Callis said he used 20 variables for his computer simulator. Murry Woroner used 129 variables [Hank Kaplan wrote down the various variables] that revolved around boxing and what boxers do.

For instance, let's take into consideration Muhammad Ali. His hand and foot speed was his greatest strengths, as well as lateral foot movement, ring generalship, and he proved to have a tremendous resistance to pain.

His weaknesses was that he carried his hands too low, didn't pack a powerful punch, had less than good defense against swarmers [his infighting ability was his weakest point].

Now let's say Ali was faced against Joe Louis, who was a better tactician, had more power, could box better than Ali. But you could say that Ali could beat Louis, because if Louis had problems with Billy Conn, then Ali is Billy Conn with more speed, more power and forty more pounds bigger.


As far as the variables are concerned, I don't know every single one off the top of my head, but here is a few articles from Hank Kaplan who drew up the 129 variables and gave them over to the NCR company to transform those variables into mathematical formulas, and then the computer would create one gigantic mathematical equation that would turn out the 'script' for the computer fights.

http://www.miaminewtimes.com/Issues/199 ... ure_2.html


I will say that the computer uses virtually every aspect of the game as a variable, from punching power to cut resistance to speed and stamina to desire and determination, as well as the various styles and movements that a boxer does [jabs, hooks, bobbing and weaving etc].

For example, Jack Dempsey bobbed and weaved [2 variables there] and was a swarmer [another variable]. He had great punching power [another variable] had a solid chin [another variable] and rarely cut [another variable]. He was very determined [another variable] but seemed to slow down in the middle rounds [another variable] and he could get up off the floor to win [another variable].

Just that short a sentence on Dempsey yeilded nine variables. Also, film of these peoples fights were fed into the NCR computer and it judged how the fighters reacted when their opponent tried to retaliate---let's say when someone try a jab [variable] at Dempsey he would bob down [variable] OR if someone threw a hook [variable] at Dempsey he would weave [variable] out of the way.

Starting to understand? If not here is the TITLE BOUT boxing website and it goes into variables that the computer takes into consideration, though I don't think it lists every single variable.

http://www.ootpdevelopments.com/tb/index.php

Simple questions - so try to keep the answer simple rather than rambling on about things I am not interested in at this point.

1. Does the simulator you are using here use the same 129 variables Woroner's simulator used?

2. If not, why keep mentioning them?

I already gave you the link to the simulator in the post prior to your own. You also asked me what were the variables, and I tried to explain how the variables work and told you a few examples [power, speed etc].

The explaination is all there, you just need to look for it yourself. As far as your new point blank question as to whether or not this simulator uses the same amount of variables that Woroner used, I don't think it does. But then again, irregardless the number of variables, whether it be twenty like Callis uses or 129, this simulator has proven thus far to be just as accurate if not more in some aspects than Woroner's.

Like I said in the MODERN DAY ALL-TIME HEAVYWEIGHT TOURNAMENT the Title Bout simulator can't be any worse than the old dated NCR computers. But just to humor you, I will check the Title Bout site and see where the page is where it says the variables it uses.

Yes, I do take this 'testing' serious. But then again, I guess I can't really say with 100% confidence that the computer would always be right. I guess the 1969 SUPERFIGHT between Ali and Marciano could almost be comparable to the 1994 music video 'Tear In My Beer' with Hank Williams Jr doing a digital duet with his father, with then state of the art computers and doubles.

But even if the majority wishes to view the computer testing to be nothing more than visual junk food, the computer has predicted all but one fight right so far with five more to go. That's pretty good considering this was a challenge to see how 'accurate' the computer is.

http://www.ootpdevelopments.com/tb/tbarticle.php?id=242

The above link goes into some length of how it works.

Now mind you, when this simulator first came out with even less details and variables, the prototype did a tournament. Their ATG HW final [with the prototype] was between Joe Louis and Gene Tunney. Since that time, they have made the simulator more advanced and have yet to run another ATG HW tournament.

Another note is that my tournament was set up differently than their's as mine was two tournaments with the best puncher/brawlers and the best tactician/boxers. Ali won the tactician/boxers tournament, while Marciano won the puncher/brawlers tournament. I had it in mind that styles make fights, and why not the best boxer versus the best puncher? So after testing it will be Marciano vs Ali for the ATG HW CHAMPIONSHIP.

I will also email the TITLE BOUT people and ask them to give a detailed list of how many exact variables that they use, so when they do reply, I will post those in here.

For the time being, I hope Cube can run a few more simulations, the final five, so this can be judged to either be accurate or not accurate enough to be a simulation of fights.

IrishRufusMurphy wrote:I already gave you the link to the simulator in the post prior to your own. You also asked me what were the variables, and I tried to explain how the variables work and told you a few examples [power, speed etc].

The explaination is all there, you just need to look for it yourself. As far as your new point blank question as to whether or not this simulator uses the same amount of variables that Woroner used, I don't think it does. But then again, irregardless the number of variables, whether it be twenty like Callis uses or 129, this simulator has proven thus far to be just as accurate if not more in some aspects than Woroner's.

Like I said in the MODERN DAY ALL-TIME HEAVYWEIGHT TOURNAMENT the Title Bout simulator can't be any worse than the old dated NCR computers. But just to humor you, I will check the Title Bout site and see where the page is where it says the variables it uses.

Yes, I do take this 'testing' serious. But then again, I guess I can't really say with 100% confidence that the computer would always be right. I guess the 1969 SUPERFIGHT between Ali and Marciano could almost be comparable to the 1994 music video 'Tear In My Beer' with Hank Williams Jr doing a digital duet with his father, with then state of the art computers and doubles.

But even if the majority wishes to view the computer testing to be nothing more than visual junk food, the computer has predicted all but one fight right so far with five more to go. That's pretty good considering this was a challenge to see how 'accurate' the computer is.

Just to humour YOU we won't mention the fudging of the Toney fight ............ or the fights you said were in and then you and/or your buddy didn't simulate them.

Anyway, now you have admitted you have no idea what variables it uses maybe you can stop padding your long-winded rambling posts with the '129 variables' crap.

Did you happen to notice how interest in your 'testing' went very quiet after you started fudging the results? Sort of like when you messed up your own contest and everyone lost interest wasn't it?

I misunderstood what Cube was saying when he posted the Toney-Peter fight because he did the 1,000 runs , rather than my own method which was 'best two out of three'. All I seen was the higher percentages going to Toney, but it also gave statistics if Toney was under-trained [over 230 pounds] giving him the higher percentage.

The weigh-in didn't even happen when Cube did the simulation, so I said Toney was picked as the winner, because it was my own fault for talking before seeing how much Toney weighed. As far as the Holyfield-Bates affair is concerned I have apologised repeatedly for posting the results late.

I have also said that the 'gimme' as everyone called it, would be scrapped and so would Toney-Peter. You yourself said that I would never admit to a loss if the computer was wrong---but the computer was wrong in one of the fights [check page 4] and I have admitted that loss.

If you count the Peter-Toney fight as a loss for the computer, even though it was my own mistake for pointing out Peter as the winner before the weigh-in, out of seven fights there would be two losses, making the computer running at almost 72% accuracy [the goal is 75%].

If you keep the Toney-Peter flub out the computer would be running at almost 85%.

As far as 'popularity' is concerned, I didn't see too many of the so-called experts taking bets on this from the beginning anyways. I was given a challenge and screw ups or not, I have taken it.

Considering there's a few fights this month and the upcoming two months, I figured I would throw around these fights in here, to see if Cube or Ezzard can simulate them.

Herbie Hide vs ???? [opponent yet to be announced]

Wladimir Klitschko vs Calvin Brock

Shannon Briggs vs Sergei Liakhovich

Floyd Mayweather vs Carlos Baldomir- IBF WW title

Jermain Taylor vs Kassim Ouma- MW title

I looked over this entire thread and realised I have alot of the numbers wrong, as there were other fights that were simulated, but were forgotten about. I looked them up to see who won the fights, and here are the official stats of the computer thus far [bold is computer picks that were correct]:

Oleg Maskaev vs Hasim Rahman

Marco Antonio Barrera vs Rocky Juarez

Nicolay Valuev vs Monte Barrett

Evander Holyfield vs Jeremy Bates*

Oliver McCall vs Darroll Wilson

Nelson Dieppa vs Hugo Cazares

Jhonny Gonzales vs Israel Vasquez

Samuel Peter vs James Toney**

Barrios vs Guzman [forgot their first names but check page 3-4]

The Guzman/Barrios fight was technically the only fight the computer wasn't correct with.

Holyfield/Bates was simulated by both me and Cube, but my results was posted after the fight and Cube was yet apart of this testing, though his simulation of the fight was posted on his website (?)

The Peter/Toney fight was somewhat a shambles, as the computer picked Toney to win if he was in shape, but it also picked Peter to win if Toney was under-trained. At the weigh-in [after the simulation was done] Toney weighed in at over 230 pounds, leading towards being under-trained. And of course Peter won, but it was my mistake for saying it picked Toney.

The Bates fight was considered a 'gimme', so I threw it out. The Toney/Peter fiasco was not the computers fault but my own, so I take responsibility for that, but I still count it as a win in my own mind.

The computers record is officially 8 wins and 1 loss, if I don't take the Holyfield/Bates fight out.

Irish,

You are playing to an empty house, mate.

You've made such a mess of this whole thing that I'm the only one left that will even reply. And I only do it for a laugh.

For once in your life, why don't you try and do something professionally. Come up with a proper test plan, seek the agreement of your peers, accept input from others and THEN maybe you will get some respect.

This completely haphzard, half-arsed 'test' is becoming embarressing.

Actually the simulator has some good things going for it but you have succeeded in reducing it to farce.

I do apologise to all of BoxRec if I did turn this thing into a 'farce' into everyone's eyes...but like I said, don't take it out on the computer, take it out on me.

Maybe I should do another tournament altogether and just post up what the computer says, with no other input from myself. I think the computer has shown its self to be incredibly accurate considering its just a machine.

Irregardless of my own mistakes, I am proud of this tournament, at least concerning the computer being so accurate.

Again I apologise, and if I lost any respect to the other posters, I am saddned by that.

IrishRufusMurphy wrote:I do apologise to all of BoxRec if I did turn this thing into a 'farce' into everyone's eyes...but like I said, don't take it out on the computer, take it out on me.

Maybe I should do another tournament altogether and just post up what the computer says, with no other input from myself. I think the computer has shown its self to be incredibly accurate considering its just a machine.

Irregardless of my own mistakes, I am proud of this tournament, at least concerning the computer being so accurate.

Again I apologise, and if I lost any respect to the other posters, I am saddned by that.

There's no need to apologise.

Why not just post the simulator's predictions, without comment, and without aiming for any specific % right / wrong etc? That way we can maybe communicate and comment without any agendas.

That's one of the reasons for the confusion / disagreement. For the Toney fight, I believe the simulator said "If Toney is undertrained" and CUBE then came up with the figure of 230 pounds arbitrarily. Correct me if I am wrong and if it was the simulator that came up with that figure.

Here's some bouts you have proposed in the past, so why not run simulations now for them? You can always run the simulations again later if some more relevant info comes to hand.

Wladimir Klitschko vs Calvin Brock
Shannon Briggs vs Sergei Liakhovich
Floyd Mayweather vs Carlos Baldomir
Jermain Taylor vs Kassim Ouma

In the past you have said you don't have the latest 'rankings' for the simulator. CUBE (or someone else) showed you where to download the latest data for it. So, if you REALLY are serious, there is no reason for you not to run the simulations now.

I will not run the simulations. Only because I was criticised in the ALL TIME HEAVYWEIGHT tournament I ran because Marciano beat Louis. I made that promise before, and I won't do it.

As far as getting rid of the past fight simulations, I think its not such a good idea, only because alot of the simulations were proven correct without faults by myself and others, like these ones:

Oleg Maskaev vs Hasim Rahman

Marco Antonio Barrera vs Rocky Juarez

Nicolay Valuev vs Monte Barrett

Oliver McCall vs Darroll Wilson

Nelson Dieppa vs Hugo Cazares

Jhonny Gonzales vs Israel Vasquez

Barrios vs Guzman [forgot their first names but check page 3-4]

Those seven were done without any faults by anyone, and the computer was wrong once, being 6-1 in predictions.

But if getting respect from the other posters means forgetting these simulations and doing others, then I will move on and do a new series of simulations [ten in number, like this one was originally].

But I'd need other TITLE BOUT users help, as I said before, I will not run the simulations, because of prior criticism.

As far as the 1,000 simulations to get a win/loss percentage, I somewhat agree even though that way is far more accurate. When this tournament was originally done it was a 'best two out of three' method, and with evidence of the Maskaev-Rahman fight, it showed it was accurate enough to still pick a correct winner.

Here's for Collins2000 when he asked how many variables the TITLE BOUT simulator used. I posted this question in the TITLE BOUT forums, and this was their response.

Looking at Title Bout's ratings, I count 33 different categories assigned/rated for each individual boxer, specifically power, defense, chin, style, punches landed, etc. There are also several other variables, such as cornermen, left-or right-handed, career stage and whether or not the fighter is champion or favored, that come into play as well.

While Woroner's simulations are important from a "historical" and gaming standpoint, I was never particularly impressed by the "realism" of either the HW or MW tourneys he ran, regardless of any "129 variables" (seems excessive,no?). They are an interesting read, however, and I expect you can find the results somewhere on the web. There is a detailed synopsis of both in the original "Peoples Almanac" edited by David Wallechinsky and Irving Wallace, now long out of print.

As an added note, I just went back and re-read the article and it states that Woroner used "58 factors" to rate the fighters. These were derived from responses sent to him by "over 250 boxing experts and writers" whose opinions/ratings he solicited for the project.

I think those of us long-time Title Bout addicts will stand by the Trunzos' more accurate and realistic ratings as well as those of Dr. Julian Compton's Data Boxing and Gary Brown's Legends of Boxing. Dr. Compton's game was featured in "The People's Almanac 2", with a HW tourney of his own, that in my opinion was far more realistic than Woroner's.

As far as Tracy Callis goes, I've enjoyed his writing and find his simulations to be quite realistic. Matter of fact, I kinda think he may have used Title Bout to simulate his "dream matchups". Further, I think he may even be here on the TB boards now and then under an assumed name. Hopefully, someone else here with more knowledge than I can fill in some of the blanks regarding Tracy's contributions.


I hope this answers your questions, and I only want to clear up that the '129' variables that I kept bringing up was to go 'back' in history and remind people that Woroner used that many in the original tournament done in 1967, as well as the SUPERFIGHT in 1969.

And, considering Tracy Callis of CyberBoxingZone, according to this guy, gave his stamp of approval to the computer, then I reckon that since Mr. Callis works on computers and has ran tournaments before, that it's a good thing.

So you are still not interested in input from others, Irish.

OK, don't run the simulations. But, please stop clogging up the forum with this now dead thread.

No one other than me was interested. And now I'm not interested. So basically no one is interested now.

You killed your own 'test' just like you killed your tournament.

Collins, I already said that I was willing to do whatever to get pplz approval. Don't see where I said I wasn't interested in others opinions, all I said was I wasn't doing the simulations because of prior criticism and that I thought it would be unfair to dismiss the simulations that had no human influences---BUT---I said if that was best, I would just redo the whole thing.

Collins never had much interest in the test in the first place except constantly criticizing.

If you do redo the whole thing, it would be better to post fights that are genuinly 50/50 (we decide,or use BoxRec rankings) and ofcourse post all the sims beforehand. I'd like to test myself against the computer

lvlarc_uk wrote:Collins never had much interest in the test in the first place except constantly criticizing.

If you do redo the whole thing, it would be better to post fights that are genuinly 50/50 (we decide,or use BoxRec rankings) and ofcourse post all the sims beforehand. I'd like to test myself against the computer

luvvy darling,

My criticisms were meant to help poor old irish. He knows that.

hahahahaha, did you put a pick in when Irish was asking for picks to test out the computer? I did, mate.

Also, how could you decide what is a 50/50 match? You don't know anything about boxing do you?

Well from reading some of your posts I got the impression you were just trying to take the piss, I didn't mean to offend you.

How do we decide what is a 50/50 match? Well we can decide on that when the time comes, or use the rankings. Witter V Hopkins for example would be a good fight to sim.

BoxRec have them 6th/7th. IBO 5th/7th. I'm guessing if we did a poll there wouldn't be much in it, as well as the bookies odds. That's how I would find a 50/50 fight.


I do know about boxing as proved by leading PTBF by a wide margin. How about you enroll and let's see how much you know, mate

I'll tell you all what. I will abandon this 'testing', but I will give reference to the prior simulations that were correct, as well as the loss.

What I think would work best is to do each simulation in an individual thread, post the simulation up, then that will give everyone enough time before fight night to cast their predictions so we can see how accurate the human predictions are to the computers.

The computer [minus the Toney-Peter flub and the Holyfield-Bates 'gimme'] was 8-1 in its record.

What I propose to do is have 8 or 10 simulations. Then after those simulations are done, then we can review how accurate the simulator is and then do the last fight from the HW TOURNAMENT that I done before, and crown the ATG HW champion.

Sound good?

lvlarc_uk wrote:Well from reading some of your posts I got the impression you were just trying to take the piss, I didn't mean to offend you.

How do we decide what is a 50/50 match? Well we can decide on that when the time comes, or use the rankings. Witter V Hopkins for example would be a good fight to sim.

BoxRec have them 6th/7th. IBO 5th/7th. I'm guessing if we did a poll there wouldn't be much in it, as well as the bookies odds. That's how I would find a 50/50 fight.


I do know about boxing as proved by leading PTBF by a wide margin. How about you enroll and let's see how much you know, mate

That's the contest in The Current Scene? Getting more right than clowns like Pringle and his ilk doesn't impress me much.

Did you put in your picks here when Irish was asking for help? I did.

IrishRufusMurphy wrote:I'll tell you all what. I will abandon this 'testing', but I will give reference to the prior simulations that were correct, as well as the loss.

What I think would work best is to do each simulation in an individual thread, post the simulation up, then that will give everyone enough time before fight night to cast their predictions so we can see how accurate the human predictions are to the computers.

The computer [minus the Toney-Peter flub and the Holyfield-Bates 'gimme'] was 8-1 in its record.

What I propose to do is have 8 or 10 simulations. Then after those simulations are done, then we can review how accurate the simulator is and then do the last fight from the HW TOURNAMENT that I done before, and crown the ATG HW champion.

Sound good?

Sure, why not? Put the first sim up now, Irish.

I'm trying to compile a list of fights, at least eight in number, only have six listed up so far:

Shannon Briggs vs. Sergei Liakhovich

Floyd Mayweather vs. Carlos Baldomir

Wladimir Klitschko vs. Calvin Brock

John Ruiz vs. Ruslan Chagaev

Jermain Taylor vs. Kassim Ouma

Oleg Maskaev vs. Peter Okhello

Anyone else know of any fights coming up soon? Preferably big fights?