BoxRec

I think that whole rumor about Marciano having a 50th fight was partly due to confusion, as when he was an amatuer and fought a pro fight as Rocky Mack and then went back as an amatuer---it's also a rumor that branched from this that Marciano actually had a loss as a pro---which is not true, although after his fight as 'Rocky Mack' he lost twice in the amatuers. Plus there were alot of ignorant people who saw the computer fight with Ali and just saw two men in a ring and figured, computer or no computer, that Marciano had his 50th fight. That or it could be that Marciano had an exhibition somewhere and it was thought as a professional bout; things like that have happened, for instance Joe Louis fought a man who was 3-3-0 and it was just an exhibition, but the state of New York put up the title in that bout anyways...so for many years there was always talk of Joe Louis having a 26th title defense.

And THANK YOUR BROCKTON for voting!

3-1 Ali (Marciano vs Ali)

1-1 (Louis vs Marciano; I voted before for Marciano the last go-around)

Note- Still need votes for the Louis vs Ali fight

My opinion on that particular vote is that Louis could win a very close decision over Ali.

Louis TKO6 over Marciano
Ali TKO9 over Marciano

If I don't like the results you give me, I declare the matchup "close" and make the result up mysefl - ahem, ask my "computer"....

Yous seriously ask from us that we should accept this?

P

The Great John L wrote:

IrishRufusMurphy wrote:I find it surprising that nobody has complained of my methods up until now.

You're right. I think up to this point you've had several others lose that were actually ahead in the balloting and nobody complained about those.

I'm sorry I didn't follow this thread closely and didn't realize how silly your methods were. I didn't realize that if the vote was close that you could throw out the result and just say that the computer favors the guy that you liked better. It wasn't fair either that Liston was out when he more votes than Tyson either. however, it's your rules, so it's your tournament. I just feel that it's silly.

Ambling Alp wrote:

The Great John L wrote:

IrishRufusMurphy wrote:I find it surprising that nobody has complained of my methods up until now.

You're right. I think up to this point you've had several others lose that were actually ahead in the balloting and nobody complained about those.

I'm sorry I didn't follow this thread closely and didn't realize how silly your methods were. I didn't realize that if the vote was close that you could throw out the result and just say that the computer favors the guy that you liked better. It wasn't fair either that Liston was out when he more votes than Tyson either. however, it's your rules, so it's your tournament. I just feel that it's silly.

Exactly. I was agreeing with Irish that I was surprised that no one had complained to this point, because there were a few times when I wondered why he ignored the voting and conducted a game sim that usually went against the vote. I’d have to check, but I thin there may have been one that wasn’t really even close and he ran a game sim.

At the time I just thought the same thing you’re saying here -- that it’s his thread and he can do what he wants. Might be interesting to revisit some of the times he ran sims when he said the voting was close…

The Great John L wrote:

Ambling Alp wrote:

The Great John L wrote: You're right. I think up to this point you've had several others lose that were actually ahead in the balloting and nobody complained about those.

I'm sorry I didn't follow this thread closely and didn't realize how silly your methods were. I didn't realize that if the vote was close that you could throw out the result and just say that the computer favors the guy that you liked better. It wasn't fair either that Liston was out when he more votes than Tyson either. however, it's your rules, so it's your tournament. I just feel that it's silly.

Exactly. I was agreeing with Irish that I was surprised that no one had complained to this point, because there were a few times when I wondered why he ignored the voting and conducted a game sim that usually went against the vote. I’d have to check, but I thin there may have been one that wasn’t really even close and he ran a game sim.

At the time I just thought the same thing you’re saying here -- that it’s his thread and he can do what he wants. Might be interesting to revisit some of the times he ran sims when he said the voting was close…

With this ingenious method, it would ahve been better he had announced his winner from the start - it may have saved some bytes.

When the votes were close, I had the computer do a best two out of three...whoever won two out of the three, that number was added on to the votes that everybody had given them---so it was fair, it wasn't done solely by the computer, just if someone won two out of three, the two 'wins' would be considered as good as votes and be added to their votes.

If you still don't understand what I am saying here is a diagram:

Let's say Sonny Liston fought Tyson and the votes were 3-2 in favor of Tyson, and since that is a close vote (it wouldn't be if it was 3-1 or 3-0), I let the computer do a best two out of three---and let's say that the computer picked Tyson two out of the three times---then I add those wins to his already 2 votes, and that would make it 5-2 in favor of Tyson, thus making him move forward in the tourney.

Understand now that I explained it better?

Don't accuse me of cheating, cus I haven't.

IrishRufusMurphy wrote:When the votes were close, I had the computer do a best two out of three...whoever won two out of the three, that number was added on to the votes that everybody had given them---so it was fair, it wasn't done solely by the computer, just if someone won two out of three, the two 'wins' would be considered as good as votes and be added to their votes.

If you still don't understand what I am saying here is a diagram:

Let's say Sonny Liston fought Tyson and the votes were 3-2 in favor of Tyson, and since that is a close vote (it wouldn't be if it was 3-1 or 3-0), I let the computer do a best two out of three---and let's say that the computer picked Tyson two out of the three times---then I add those wins to his already 2 votes, and that would make it 5-2 in favor of Tyson, thus making him move forward in the tourney.

Understand now that I explained it better?

Don't accuse me of cheating, cus I haven't.

You must have some random generator in your comouter to come up with results like "two out of three".

Would you bother explaining how you did this? Not the technical details, just the principles.

Cheers,
P

No what I did was, I ran the fight through the computer, and it would go by a round by round basis, telling who eventually won. I would take that as a win. I would run it again two other times. For example, when I did the Tyson-Frazier fight, I got this result:

Tyson won the first fight. Frazier won the second. Tyson won the third. Every time I would run it through the computer there would be different result, so that's what I meant by "two out of three".

Same thing happened when I used the highly praised and typically accurate "eenie meenie miney moe" method.

.....well I apologise to everyone then, the TITLE BOUT game isn't really the best choice of simulators to use, but it's all I could think of atm. I sent my whole tournament, as well as the top three fighter(s) and possible matches between the three to CBZ and I hope that they will eventually write back and do better than I could.

Again I apologise if I upset anyone.

IrishRM

I checked out Title Bout and the computer game is quite a bit different in the way it ranks fighters to the old card game. All fighters have a number of associated fields which are ranked and then an overrall figure is given to the fighter. It used to be that the overall fgure was out of 12.

From memory the HWs I think Johnson, Dempsey, Louis, Marciano, Liston and Ali were all 12/12 and Jeffries, Tunney and Frazier were 11/12...

In the new version they have an overall rating of 15. Ali was 15, Louis 14, Dempsey, Tunney, Marciano all 13, Johnson, Tyson, Holyfield, Frazier and Foreman were 12...

I'll be honest. I ended up sitting up into the wee hours organising tournaments, moaning about some of the ratings and gambling away trillions of virtual dollars on the outcomes of my favourite mythical matchups.

I don't think you need to apologise. As much fun as these simulatins and tournaments are you could run a million of them and nobody could ever say for sure just how accurate they were.

IrishRufusMurphy wrote:No what I did was, I ran the fight through the computer, and it would go by a round by round basis, telling who eventually won. I would take that as a win. I would run it again two other times. For example, when I did the Tyson-Frazier fight, I got this result:

Tyson won the first fight. Frazier won the second. Tyson won the third. Every time I would run it through the computer there would be different result, so that's what I meant by "two out of three".

I don't understand. WHY did Tyson win the first, and Frazier the second? What did your computer do differently between the first and the second time, such that he would get a different result?

IrishRufusMurphy said

"Ali had hell with Frazier, who was like Marciano. Louis never was any good with pressure fighters. We could argue this over and over, but still in the end film and records don't lie---Ali's toughest opponent was Frazier, who is ranked less than Marciano over all by most real boxing authorities.

Louis by his own admission said he never felt he could beat Marciano, and less gifted fighters with a pressure/brawling style gave Louis problems, plus Joe actually fought the man, though not in his prime the man should know best than any other phuckin source whether or not he could have beaten Marciano in his prime---his answer was NO"...


I sincerely agree Irish, people mock me when I'm constantly proclaiming that Rocky Marciano was THE GREATEST!!!

They're very quick to judge him, because he "only fought old fighters" + baiscally count him out right away...

BUT the 2 recognisably GREATEST fighters in this forum our ALI & Louis, and Marciano already beat Louis... While he gave Ali a torrid time during the filming of the Superfight, according to the Wig Incident.

I believe that there's soignificant evidence wwhich would suggest that Marciano what it takes and would defeat both of these boxers...

As for my own tournament, it's reached the final and Marciano is abest from it.

MY FINAL HAS, MUHAMMAD ALI VERSUS JOE LOUIS (which could've been predicted, and I'm not complaining about that)

HOWEVER in mine, Marciano was eliminated in the Quarter Finals by Jack Johnson, who I just couldn't see defeat Marciano or gind him down.....


BUT, to conclude, I do agree with your opinions.

pundit wrote:

IrishRufusMurphy wrote:No what I did was, I ran the fight through the computer, and it would go by a round by round basis, telling who eventually won. I would take that as a win. I would run it again two other times. For example, when I did the Tyson-Frazier fight, I got this result:

Tyson won the first fight. Frazier won the second. Tyson won the third. Every time I would run it through the computer there would be different result, so that's what I meant by "two out of three".

I don't understand. WHY did Tyson win the first, and Frazier the second? What did your computer do differently between the first and the second time, such that he would get a different result?

Pundit the simulation takes into consideration many different factors. One of which is how well conditioned fightersd were mentally and physically. It's a sort of consistency rating. So it attempts to factor in fighters like Bowe who didn't always turn up to fights in shape.

Even without this the same 2 fighters do not always produce the same result. If Wlad fought Ali 10,000 times I'm sure he'd win at least once.

Ezzard wrote:

pundit wrote:

IrishRufusMurphy wrote:No what I did was, I ran the fight through the computer, and it would go by a round by round basis, telling who eventually won. I would take that as a win. I would run it again two other times. For example, when I did the Tyson-Frazier fight, I got this result:

Tyson won the first fight. Frazier won the second. Tyson won the third. Every time I would run it through the computer there would be different result, so that's what I meant by "two out of three".

I don't understand. WHY did Tyson win the first, and Frazier the second? What did your computer do differently between the first and the second time, such that he would get a different result?

Pundit the simulation takes into consideration many different factors. One of which is how well conditioned fightersd were mentally and physically. It's a sort of consistency rating. So it attempts to factor in fighters like Bowe who didn't always turn up to fights in shape.

Even without this the same 2 fighters do not always produce the same result. If Wlad fought Ali 10,000 times I'm sure he'd win at least once.

Of course, but how do you "simulate" this in a computer?

Here's the link to the Title Bout Championship Boxing game, so you can check to see for yourself. As you can see Title Bout boxing takes into consideration many factors, from different styles, to how durable the fighters were, to how they were resistant to cuts and their power etc.

I forget how many variables they use to determine outcomes on this simulator, but when you put in the matches between the two men, it looks at their records (courtesy of BoxRec.com) as well as every bit of information the game has on the two men---and you punch it in and it will give a read out of a "round by round" basis of what happens, and it predicts a winner.

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

You can download a demo of it, or a free trial, or buy it. I think in alot of ways it is somewhat accurate---I put in alot of fights that have actually happened and pretty much got the same results back as the real results.

http://www.ootpdevelopments.com/tb/imag ... screen.jpg


This is a screen of how it looks---the example above is Lennox Lewis vs Vitali Klitschko in a computer match---as each punch is landed it tells how "hurt" a fighter is, how "hard" the punches are thrown, etc. Much like Murry Woroner's NCR-315 computer was in determining who punched where and when.

What about the question of how it comes up with different results for the same bout?

Okies...I am going to use a series of fights as an example, as to how the simulator would decide an out-come. Take into considerattion the Charles-Walcott bouts---Charles won 2 and Walcott won 2. The first two bouts went the 15 round distance---the third went only to the 7th---the fourth fight also went the distance of 15 rounds.

If put into the computer of a Charles-Walcott 5th fight, the possibility of a stoppage in seven rounds or later is possible, with the most highly outcome being a 15 round decision, as three of the four went the distance.

Now in a fight that never happened, the computer takes into consideration of different styles, and how the two men matched up against people of various styles and their strengths and weaknesses in various elements; how great their opposition were and other details. There could be many outcomes based on all these "what if" factors---that's why I did a best two out of three, having the men fight three times, seeing as who would win the most times.

I have done this before with "up-coming" matches and have had pretty decent results, one such match I placed in was Klitschko vs Johnson (Kirk & Vitali) and ironically enough I had Klitschko winning three out of three, all by KO's in 4 rounds or less---and what happened? Klitschko knocked out Johnson by KO in the 2nd.

It's rather interesting.

Why don't you run it ONCE for the Rahman vs Maskaev fight and post the full results in a thread in here now.

That will give us a chance to view the predicted outcome and, next weekend, compare it against what actually happened.

Okies, how about this. Ezzard says he been running fights off this as well, and I think some people are still accusing me of trickery---so let's have Ezzard run the MASKAEV vs RAHMAN bout, just to see what will happen.

If that is ok with Ezzard?

Now I have a good belief in the computer that it could very well come up with a decent simulation of what could happen---but I only say "iono" because of Murry Woroner's mistake in using the NCR-315 to determine before-hand the out come of a real match.

Myself I don't think it was the computer's fault, it was clearly a mistake of the men who programmed the machine and gave the computer wrong bits of information---because Woroner tried to determine the out-come of FRAZIER-FOSTER, and judging solely by records, Foster had the better wins in his weight class---and the computer picked Foster to win by KO in the 6th.

What Woroner and others failed to do was tell the computer that it was Foster's very first move to the HW division, facing Joe Phuckin' Frazier, who was at the time the UNDEFEATED HW CHAMPION OF THE WORLD!

Foster's power did not carry over into the HW division, he did have skills and speed, but his power was lacking and he found out soon enough that he couldn't take a HW punch, let alone a left hook from Frazier, as Frazier knocked him the fornicate out in 2 rounds!

This made Murry Woroner look like a joke. But I don't hold the computer wrong, I hold the programmers wrong and responsible, for not putting in that small fact, that Foster was moving up to HW for the first time, against the HW champion of the world, with one of the highest KO percentages of all time, who was undefeated.

So, when you're going to make matches, make sure that it is with fighters from the same weight classes, or at least with fighters who have had success at higher weights, because if you don't it is sure to be not exactly accurate.

Computer simulations are only as good as the data put into them and the number of variables that the game engine takes into consideration.

I did drag this out and look at it. I spent hours of my youth playing the old card game and bought the computer game a few years back as a bit of a nostalgia trip. I fire it up occasionally when I get writer's block.

Like everything there is a random factor involved. I think the version IRM has is a bit more updated than mine but it looks almost exactly the same.

The fighters are rated on various fields like

Control of a fight versus a boxer (out of 14)
Control of a fight versus a slugger (also 14)
Power (13)
How easy the fighter is knocked down (0 being the best rating)
How easy the fighter is knocked out (0 being the best rating)
Recuperative powers (0 being the best rating)
etc...
There are about 30 or so categories in all.

Fighters have a punches landed and a counter punches landed rating too.

The old game used to work with a random number from 1-80... The punches landed score for an average fighter might be 39 (This is then adjusted in accordance with the defence rating of the opponent). If the random number was between 1-39 then the fighter landed a punch (the power of the punch and the opponents reaction to it would then be ascertained using other processes).

So obviously the random numbers have an impact on what happens. A fighterX with maximum punches landed score (46), adjusted against a guy with no defence would have a final rating of 52. So he has 52/80 chance of landing a punch at a given moment. If his opponent (fighterY) starts with a cr*p rating of 36 and then has it adjusted for fighting against a defensive whizz then his final score might be 30/80...

so fighterX has a much better chance of landing a punch with his 52/80 compared to 30/80. over the course of 1,000 fights he's going to land considerable more punches but there will be times when he actually out punches his opponent.

I guess it's all just probability.

Hope this helps understand how different results are produced.

Computer simulations are only as good as the data put into them and the number of variables that the game engine takes into consideration.

I did drag this out and look at it. I spent hours of my youth playing the old card game and bought the computer game a few years back as a bit of a nostalgia trip. I fire it up occasionally when I get writer's block.

Like everything there is a random factor involved. I think the version IRM has is a bit more updated than mine but it looks almost exactly the same.

The fighters are rated on various fields like

Control of a fight versus a boxer (out of 14)
Control of a fight versus a slugger (also 14)
Power (13)
How easy the fighter is knocked down (0 being the best rating)
How easy the fighter is knocked out (0 being the best rating)
Recuperative powers (0 being the best rating)
etc...
There are about 30 or so categories in all.

Fighters have a punches landed and a counter punches landed rating too.

The old game used to work with a random number from 1-80... The punches landed score for an average fighter might be 39 (This is then adjusted in accordance with the defence rating of the opponent). If the random number was between 1-39 then the fighter landed a punch (the power of the punch and the opponents reaction to it would then be ascertained using other processes).

So obviously the random numbers have an impact on what happens. A fighterX with maximum punches landed score (46), adjusted against a guy with no defence would have a final rating of 52. So he has 52/80 chance of landing a punch at a given moment. If his opponent (fighterY) starts with a cr*p rating of 36 and then has it adjusted for fighting against a defensive whizz then his final score might be 30/80...

so fighterX has a much better chance of landing a punch with his 52/80 compared to 30/80. over the course of 1,000 fights he's going to land considerable more punches but there will be times when he actually out punches his opponent.

I guess it's all just probability.

Hope this helps understand how different results are produced.