Updating the BOXREC Forum Heavyweight Rankings CLOSED

[Horse]

I think something needs to be done about the joint rankings.

Maybe have a vote on which fighter should be higher.

[jezzamundo]

I think something needs to be done about the joint rankings.

Maybe have a vote on which fighter should be higher.

Joint rankings only occur when there's a draw in the voting. It's relatively rare and this is the first time in nine years there have been three joint rankings. I'd rather not make a new topic every time there's a tie, so I'm considering using the mean vote to split them in the future. If the mean is a draw too, then the fighter ranked higher before the topic remains higher.

[Horse]

Joint rankings only occur when there's a draw in the voting. It's relatively rare and this is the first time in nine years there have been three joint rankings. I'd rather not make a new topic every time there's a tie, so I'm considering using the mean vote to split them in the future. If the mean is a draw too, then the fighter ranked higher before the topic remains higher.

[greg]

Joshua to #9,
Martin remains unranked

...why not top 5? Joshua surely has the potential but considering the level of opposition he's not there yet...

[Belanio]

My vote: Joshua to #6, Martin to #8.

[Heretic]

My vote.

Joshua to # 11

Beating Martin does not mean anything. Hes only meaningful win is against Glazkov and that match was ruined by the injury. I am not sure he would have won against Glazkov in fair match. The opponent was still better than any average journeyman and i think that some activity points are in order.

My opinion on the ties is that they should be resolved somehow. The list looks quite bad at the moment.

I have enjoyed this list. Keep up the good work

[jezzamundo]

The votes have been tallied, Joshua rises to #6, Martin remains unranked.
Based on popular demand, I have got rid of the split rankings. In the case of a drawn vote, the higher ranked fighter before the topic remains on top.

Champion: Tyson Fury
1. Wladimir Klitschko
2. Alexander Povetkin
3. Deontay Wilder
4. Luis Ortiz
5. Kubrat Pulev
6. Anthony Joshua
7. Bryant Jennings
8. Carlos Takam
9. Vyacheslav Glazkov
10. Lucas Browne
11. Dereck Chisora
12. Bermane Stiverne
13. Artur Szpilka
14. Malik Scott
15. Joseph Parker

[BroughtonRulesRefuge]

I'm sorry, but either you didn't read my reply to your accusation in a previous topic, or you didn't understand it for some reason. Joshua to #5 is my vote, which carries no more weight than anyone else's.

Whenever asked, I've always been very upfront with the way these rankings are calculated - I use the median vote. I clearly explained what this means for those who didn't know. Using the median is a much fairer way of counting votes because it means that the majority rules, with no exceptions. I don't know what gave you the idea that I don't count every vote, but that is absolutely 100% incorrect and frankly, insulting, as I've run these rankings in an open and honest way for the past nine years.

Median is a very basic mathematical concept that is taught to ten year olds. If you can't understand that this is a better, fairer way to count votes than the mean, then I'm surprised you have intelligence enough to operate a keyboard.

- Since you privately emailed me with more data, let's be clear here, somewhere in the past I questioned the poll methods and as I recall you stated you didn't count so called outlier votes, but perhaps you meant to couch it in other terms, but it seems that is the net effect as the example you provided me, ie:


Votes for Joshua: 4, 10, 12, 12, 13, 14, 14, 14, 14, 15, 15.

I use the median vote to determine the fighter's rank. In this case there were eleven votes, so the 6th highest vote is the median, which you can see in bold. If it were an even number of votes, I would take the mean of the two middle votes - this sometimes results in a number ending in point five, which results in shared rankings, something I'm looking at introducing a new rule to avoid in the future.

Had I used the mean vote in the Joshua vs Whyte topic, the number we get is 12.45, which would round down to 12. This would be a very unfair way to count votes, because now not every vote is equal - they guy who voted Joshua to #4 has had a far greater impact on the result than anyone else. The voters wanted Joshua ranked behind the #12 guy by a ratio of 7:4 (and behind the #13 guy by a ratio of 6:5), yet the minority get their way largely because of one outlying vote. Using the median, every vote is counted and the majority rules with no exceptions. This is demonstrably the fairest way to count votes - a true one man one vote democracy, as you said.

- The mean of the two middle votes in an even count poll is still a variation of the median. The median is just one of many different tools to look at stats, hardly the holy grail you make it out. There is virtually no change from the top highest votes, yet there is significant change for the lowest 5 votes, all of whom think Joshua is much better than the seeming herd mentality of the top 5. So in effect you have muted the voice of the lower 5 from the poll, hardly the democracy you make out. Political polls often use all sorts of stat monkey business to get the results they are being paid to produce sorta like CPAs are prone to do as so many business failures like Enron for example have shown us.

I'm not going to get all out of sorts over some silly boxrec poll that was still rating the guy Martin beat, Glazkov in the middle of the pack while Martin remained unranked. I only weighed in because Joshua was being treated unfairly, and he is as I have now shown. We are just going to have to disagree on this point which is fine by me. You have turned around for this poll to suddenly rank Joshua much more favorably than you showed last time, all for beating a guy unranked in your median poll, just a little slower to see the forest for all the trees than half of us who rank him quite higher.

[jezzamundo]

I'm sorry, but either you didn't read my reply to your accusation in a previous topic, or you didn't understand it for some reason. Joshua to #5 is my vote, which carries no more weight than anyone else's.

Whenever asked, I've always been very upfront with the way these rankings are calculated - I use the median vote. I clearly explained what this means for those who didn't know. Using the median is a much fairer way of counting votes because it means that the majority rules, with no exceptions. I don't know what gave you the idea that I don't count every vote, but that is absolutely 100% incorrect and frankly, insulting, as I've run these rankings in an open and honest way for the past nine years.

Median is a very basic mathematical concept that is taught to ten year olds. If you can't understand that this is a better, fairer way to count votes than the mean, then I'm surprised you have intelligence enough to operate a keyboard.

- Since you privately emailed me with more data, let's be clear here, somewhere in the past I questioned the poll methods and as I recall you stated you didn't count so called outlier votes, but perhaps you meant to couch it in other terms, but it seems that is the net effect as the example you provided me, ie:


Votes for Joshua: 4, 10, 12, 12, 13, 14, 14, 14, 14, 15, 15.

I use the median vote to determine the fighter's rank. In this case there were eleven votes, so the 6th highest vote is the median, which you can see in bold. If it were an even number of votes, I would take the mean of the two middle votes - this sometimes results in a number ending in point five, which results in shared rankings, something I'm looking at introducing a new rule to avoid in the future.

Had I used the mean vote in the Joshua vs Whyte topic, the number we get is 12.45, which would round down to 12. This would be a very unfair way to count votes, because now not every vote is equal - they guy who voted Joshua to #4 has had a far greater impact on the result than anyone else. The voters wanted Joshua ranked behind the #12 guy by a ratio of 7:4 (and behind the #13 guy by a ratio of 6:5), yet the minority get their way largely because of one outlying vote. Using the median, every vote is counted and the majority rules with no exceptions. This is demonstrably the fairest way to count votes - a true one man one vote democracy, as you said.

- The mean of the two middle votes in an even count poll is still a variation of the median. The median is just one of many different tools to look at stats, hardly the holy grail you make it out. There is virtually no change from the top highest votes, yet there is significant change for the lowest 5 votes, all of whom think Joshua is much better than the seeming herd mentality of the top 5. So in effect you have muted the voice of the lower 5 from the poll, hardly the democracy you make out.

Political polls often use all sorts of stat monkey business to get the results they are being paid to produce sorta like CPAs are prone to do as so many business failures like Enron for example have shown us.

I'm not going to get all out of sorts over some silly boxrec poll that was still rating the guy Martin beat, Glazkov in the middle of the pack while Martin remained unranked. I only weighed in because Joshua was being treated unfairly, and he is as I have now shown. We are just going to have to disagree on this point which is fine by me. You have turned around for this poll to suddenly rank Joshua much more favorably than you showed last time, all for beating a guy unranked in your median poll, just a little slower to see the forest for all the trees than half of us who rank him quite higher.

Thanks for replying - I was wondering whether you'd got my message.

I'm aware that this is also median, although admittedly I didn't explain myself as well as I could have.

I have no idea what you're talking about when you say Joshua was being treated unfairly - he was treated exactly the same as any other fighter in any topic in these rankings. No system is perfect, but to me, median is a much fairer and more representative way to count votes than mean. Using mean also allows for people to skew the results by making outlandish votes, which median does not. The voice of the five in the lower end of the poll weren't muted by me, but by the other six voters. Six is more than five - just because the five want something more doesn't make their votes more important.

For the record, I voted for Martin to be ranked ahead of Glazkov, but was overruled by the majority.

[BroughtonRulesRefuge]

Thanks for replying - I was wondering whether you'd got my message.

I'm aware that this is also median, although admittedly I didn't explain myself as well as I could have.

I have no idea what you're talking about when you say Joshua was being treated unfairly - he was treated exactly the same as any other fighter in any topic in these rankings. No system is perfect, but to me, median is a much fairer and more representative way to count votes than mean. Using mean also allows for people to skew the results by making outlandish votes, which median does not. The voice of the five in the lower end of the poll weren't muted by me, but by the other six voters. Six is more than five - just because the five want something more doesn't make their votes more important.

For the record, I voted for Martin to be ranked ahead of Glazkov, but was overruled by the majority.

- Joshua is being treated unfairly because he has shown more skills and class in the ring than the accuracy of this poll reflects.

Even your language demonstrates the bias of the poll by dismissing out of hand the five lowest poll rankings. If we average the highest six numbers of your example, 4, 10, 12, 12, 13, [14,] 14, 14, 14, 15, 15, we get a ranking of 14.33. If we average the lowest six numbers, we get a ranking of 10.83, a significant difference, yet we've incorporated the median into both counts. If the bottom five of the six ranked Josh #1, that tallies up to a 3.17 ranking, yet your system still pegs Josh at 14, clearly out of whack.

This poll simply put is a vote on rankings 1-15. If everyone agreed to a #1 vote for Josh or whomever, that's 15 points for the best ranking possible. Add up all the rankings using your above example, that adds up to 137 points without any decimals needing rounding. If the results of another fighter come up one point over or under, then he'll be ranked over or under Josh and so on and so forth. Knowing some will not vote for a fighter in view of voter peculiarities, give those nonvotes a 20 ranking to reflect that nonvote in the total if they are otherwise voting on other fighters in the poll. Those totals should determine place. You'd get a lot less ties that plenty here gripe about.

You could easily upgrade to this system in your poll by initiating a new vote using the existing fighters currently ranked to reset every fighters's point totals instead of the deeply flawed median tool never meant to be applied in such a limited pool of voters attempting to assign rankings.

Why not conduct a yea or nay poll and let the voters mull over any changes? You've done well bearing the standard and generating the interest, and it's still you running the poll even if it's supposed to be the boxrec forum poll, so there it is.

[jezzamundo]

Thanks for replying - I was wondering whether you'd got my message.

I'm aware that this is also median, although admittedly I didn't explain myself as well as I could have.

I have no idea what you're talking about when you say Joshua was being treated unfairly - he was treated exactly the same as any other fighter in any topic in these rankings. No system is perfect, but to me, median is a much fairer and more representative way to count votes than mean. Using mean also allows for people to skew the results by making outlandish votes, which median does not. The voice of the five in the lower end of the poll weren't muted by me, but by the other six voters. Six is more than five - just because the five want something more doesn't make their votes more important.

For the record, I voted for Martin to be ranked ahead of Glazkov, but was overruled by the majority.

- Joshua is being treated unfairly because he has shown more skills and class in the ring than the accuracy of this poll reflects.

Even your language demonstrates the bias of the poll by dismissing out of hand the five lowest poll rankings. If we average the highest six numbers of your example, 4, 10, 12, 12, 13, [14,] 14, 14, 14, 15, 15, we get a ranking of 14.33. If we average the lowest six numbers, we get a ranking of 10.83, a significant difference, yet we've incorporated the median into both counts. If the bottom five of the six ranked Josh #1, that tallies up to a 3.17 ranking, yet your system still pegs Josh at 14, clearly out of whack.

This poll simply put is a vote on rankings 1-15. If everyone agreed to a #1 vote for Josh or whomever, that's 15 points for the best ranking possible. Add up all the rankings using your above example, that adds up to 137 points without any decimals needing rounding. If the results of another fighter come up one point over or under, then he'll be ranked over or under Josh and so on and so forth. Knowing some will not vote for a fighter in view of voter peculiarities, give those nonvotes a 20 ranking to reflect that nonvote in the total if they are otherwise voting on other fighters in the poll. Those totals should determine place. You'd get a lot less ties that plenty here gripe about.

You could easily upgrade to this system in your poll by initiating a new vote using the existing fighters currently ranked to reset every fighters's point totals instead of the deeply flawed median tool never meant to be applied in such a limited pool of voters attempting to assign rankings.

Why not conduct a yea or nay poll and let the voters mull over any changes? You've done well bearing the standard and generating the interest, and it's still you running the poll even if it's supposed to be the boxrec forum poll, so there it is.

We'll have to agree to disagree but I'm glad we understand each other. I understand your system, but believe it is inferior to mine - less democratic and more open to manipulation. I've taken comments about the number of tied votes on board and have now made the rule that in the event of a tie, the higher ranked fighter before the topic is placed higher.

[Tarkus]

I've taken comments about the number of tied votes on board and have now made the rule that in the event of a tie, the higher ranked fighter before the topic is placed higher.

Also I agree with the use of median. It is the optimal and simple solution. It does result in a bit illogical outcome sometimes though. Especially when prospects end up high in the rankings. Like Joshua right now ends up on 6th place after beating an unranked and unskilled football players. But rankings are as good as the people who compile them.

[greg]

..I've no problem with the rankings as well as the method used...fully aware of the fact though that as long as the top guys haven't fought each other there's no objective way to rate them to everyone's satisfaction...

[Lennox]

No rankings will be perfect to everyone's eyes if you take the subjective into the equation. If you go by results you can get nearer, things may look skewed but at least there is a reason.