Everyone has an opinion. Even I have mine, but now I am not going to talk about it. But what if we tossed opinions aside and just looked at the results. Take every top-boxers of the decade, and see who won who... and based on that estimate the rest of the matches - matches that never happened. As a statistician I found this exercise to be very interesting. And I really do not mean that this is "the absolute truth" - only one way more to take a look at things. So in a nutshell, I collected h2h results, and used statistical imputation methods to fill in the blanks (head-ups that never happened). And I always looked just the results in that particular decade, it didn't matter what happened before or after that. The higher ranked opponent the boxer won, the more points he received. So getting many victories over nobodys takes you nowhere - winning the best is the way to reach the top! So here we go. Every boxer has an index rating based on the average points of decades top 15 fighters. I would love to hear some comments, criticism, questions... all feedback is appreciated. I had fun doing this, hopefully someone gets something from reading the results. 1890's 1. James J. Jeffries, 134,2 pts 2. Tom Sharkey, 123,7 pts 3. Charles "Kid" McCoy, 120,7 pts 4. Peter Maher, 118,5 pts 5. Joe Choynski, 107,3 pts 6. Gus Ruhlin, 100,1 pts 7. Frank Childs, 99,5 pts 8. Bob Fitzsimmons, 98,4 pts 9. Peter Jackson, 97,8 pts 10. Walter Johnson, 87,4 pts (James J. Corbett, 65,1 pts) 1900's 1. Jack Johnson, 120,0 pts 2. James J. Jeffries, 116,7 pts 3. Jack O'Brien, 111,6 pts 4. Tommy Burns, 106,1 pts 5. Marvin Hart, 99,0 pts 6. Sam Langford, 98,9 pts 7. Al Kaufman, 98,7 pts 8. Bill Lang, 97,8 pts 9. Joe Jeannette, 97,5 pts 10. Sam McVea, 95,6 pts 1910's 1. Jack Dempsey, 117,4 pts 2. Fred Fulton, 113,4 pts 3. Harry Greb, 108,1 pts 4. Joe Jeannette, 106,1 pts 5. Sam McVea, 105,4 pts 6. Kid Norfolk, 105,0 pts 7. Harry Wills, 104,6 pts 8. Jack Johnson, 98,3 pts 9. Jess Willard, 97,6 pts 10. Sam Langford, 97,5 pts 1920's 1. Gene Tunney, 121,7 pts 2. Jack Dempsey, 112,7 pts 3. Jack Sharkey, 107,1 pts 4. Jim Maloney, 103,6 pts 5. Harry Greb, 103,5 pts 6. Billy Miske, 101,9 pts 7. Harry Wills, 101,6 pts 8. Young Stribling, 100,5 pts 9. George Godfrey, 95,5 pts 10. Jack Delaney, 94,8 pts 1930's 1. Joe Louis, 132,9 pts 2. Ernie Schaaf, 110,5 pts 3. John Henry Lewis, 105,6 pts 4. Primo Carnera, 104,1 pts 5. Max Schmeling, 103,2 pts 6. Nathan Mann, 102,0 pts 7. Charley Retzlaff, 95,3 pts 8. Maxie Rosenbloom, 95,1 pts 9. Mickey Walker, 94,8 pts 10. Lou Nova, 93,9 pts (Max Baer, 86,9 pts) (James J. Braddock, 86,6 pts) 1940's 1. Joe Louis, 125,9 pts 2. Ezzard Charles, 119,5 pts 3. Bob Pastor, 102,8 pts 4. Archie Moore, 102,6 pts 5. Jersey Joe Walcott, 100,0 pts 6. Melio Bettina, 97,9 pts 7. Roscoe Toles, 96,8 pts 8. Jimmy Bivins, 96,7 pts 9. Lee Q Murray, 96,1 pts 10. Billy Conn, 95,0 pts 1950's 1. Rocky Marciano, 125,0 pts 2. Ingemar Johansson, 124,4 pts 3. Eddie Machen, 110,0 pts 4. Floyd Patterson, 103,7 pts 5. Archie Moore, 103,2 pts 6. Harold Johnson, 101,1 pts 7. Tommy Jackson, 97,9 pts 8. Willie Pastrano, 93,9 pts 9. Don Cockell, 93,4 pts 10. Nino Waldes, 92,5 pts (Sonny Liston, 92,0 pts) (Joe Louis, 88,5 pts) 1960's 1. Muhammad Ali, 137,0 pts 2. Joe Frazier, 130,6 pts 3. Jimmy Ellis, 122,3 pts 4. Sonny Liston, 109,0 pts 5. Leotis Martin, 101,2 pts 6. Oscar Bonavena, 101,2 pts 7. Ernie Terrell, 98,2 pts 8. Floyd Patterson, 95,2 pts 9. Zora Folley, 90,2 pts 10. Bob Cleroux, 89,2 pts 1970's 1. Muhammad Ali, 142,2 pts 2. Larry Holmes, 128,3 pts 3. Joe Frazier, 116,3 pts 4. George Foreman, 114,3 pts 5. John Tate, 106,1 pts 6. Ken Norton, 105,4 pts 7. Gerrie Coetzee, 103,2 pts 8. Jerry Quarry, 95,2 pts 9. Jimmy Young, 93,9 pts 10. Ron Lyle, 88,7 pts 1980's 1. Mike Tyson, 143,1 pts 2. Larry Holmes, 122,0 pts 3. Evander Holyfield, 119,5 pts 4. Tim Witherspoon, 107,1 pts 5. Michael Spinks, 103,9 pts 6. James Douglas, 99,2 pts 7. Michael Dokes, 98,3 pts 8. Tyrell Biggs, 97,6 pts 9. Frank Bruno, 92,4 pts 10. Pinklon Thomas, 91,7 pts 1990's 1. Lennox Lewis, 134,5 pts 2. Riddick Bowe, 128,2 pts 3. Evander Holyfield, 120,9 pts 4. Mike Tyson, 107,0 pts 5. Michael Moorer, 106,3 pts 6. Oliver McCall, 102,7 pts 7. Frank Bruno, 100,6 pts 8. Brian Nielsen, 99,1 pts 9. Tommy Morrison, 94,2 pts 10. Ray Mercer, 93,0 pts (George Foreman, 88,1 pts) 2000's (so far) 1. Lennox Lewis, 125,0 pts 2. Wladimir Klitschko, 124,5 pts 3. Vitali Klitschko, 123,3 pts 4. Nikolay Valuev, 119,4 pts 5. Chris Byrd, 118,7 pts 6. Ruslan Chagaev, 109,3 pts 7. James Toney, 106,5 pts 8. John Ruiz, 102,9 pts 9. Samuel Peter, 102,3 pts 10. Joe Mesi, 98,7 pts (Evander Holyfield, 76,9 pts)
Welcome to the board. I can tell that you are going to make a bit of a splash. My main criticism of your method here is that it is biased in favour of fighters whose prime fell within a calendar decade, and against those whose primes crossed between two calendar decades. I do think that there is a lot of potential for such statistical aproaches.
That is true, and know that too... At one stage I thought that I would do this in a way where I would use overlapping time periods (for example 1950-1959 would be followed by 1955-1964), or some moving average method, but... well, I run out of gas for now!
1. I don't know where the ratings come from so it is difficult for me to see how these numbers are derived. Questions: 1. The black dynamite fighters seem to be rated very low. Wills behind Fulton? Langford behind everybody. You are the first I know of who considers Bill Lang ahead of Jeannette and McVea. 2. 1890's--Why Maher (and Sharkey) ahead of Fitz? Especially Maher. 3. 1900's--Why Burns and Hart ahead of Langford, Jeannette, McVea? Your system does put Kaufman fairly high--I do think he has been somewhat historically underrated. 4. 1910's--Fulton way ahead of Wills? 5. 1920's--Maloney ahead of Godfrey? 6. 1930's--Ernie Schaaf ahead of everyone but Louis? 7. 1940's--How does Pastor get in front of Conn? It is very hard for me to see how Elmer Ray with his victories over Charles and Walcott, 50 plus bout winning streak, 72 out 74 run, etc, does not make this decade's list. 8. 1950's--Why not Rex Layne instead of Cockell? Layne defeated Walcott and Charles. I will let the young whippersnappers take over now. Welcome aboard.
Tom Sharkey so ahead of Fitzsimmons? Well, if you're crunching the official record, it makes a certain sense, since officially Sharkey won by foul in their first meeting. I'd be curious how the 1890's ratings would change if the first Fitzsimmons/Sharkey meeting was figured as a KO win for Fitzsimmons.
Answers: 1. Well, firstly - I these are not my opinions, just results of the analysis. The reason, why these black fighters are rated low is surely related to that, that they had to fight against each other - and had no chance to prove themselves against white guys. Although this analysis should take that into an account at least in some degree, I believe that this is the main reason to this matter. 2. Looking at the data Sharkeys rating is affected for example by win over Kid McCoy... and he did beat Fitz too. On the other hand Maher lost to Fitz twice, but had some great wins too. Fitz on the other hand was quite inactive during this decade against top heavyweights, and that reflects to his rating. He did win Maher twice and of course Corbett - but that was about it (in terms of this analysis). 3. Again - the black boxers had to fight mainly against each other, and in this analysis they all suffered of it. Burns actually has a good record during the decade, and victories over Hart, Lang and O'Brien (although he also lost against him) helped his ranking. The Hart-case... well, he might be overestimated (I don't know) because of one victory - the victory over Jack Johnson. But that's how this went, beating the best gives you a lot of points. 4. Fulton really surprised me, too! Losing to Jim Johnson was one case that dropped Wills ranking. 5. Godfreys record during the decade (63-15-1) actually wasn't that great, but he also fought the best. Therefore, although he suffered many losses he is ranked this high. 6. Maybe because he wasn't beaten by Louis Seriously, it was quite a close call behind Louis who was dominating. 7. Elmer Ray did win Charles and Walcott, but he did lose to them too which affected his points. In addition to that he did have three wins against Colion Chaney, but then again he lost for example to Turkey Thompson, and this all made his ranking quite mediocre. 8. Rex Layne did win Walcott and Charles, but winning them didn't credit so much points during this decade. And he did lose to Charles twice. And Layne had his share of losses - three times against Baker, twice against Walls, twice against Jackson... ----- So not only victories, but losses affect too in rating. And because Langford, Jeannette, Wills and McVea (among others) fought a lot against each others, they made each others life difficult - even in this rating system. And this system is far from perfect, I am the first to admit that
That is something I thought too. But I decided to leave my opinions aside and go with the official records. There are cases, where were DQ matches, poor refereeing or even suspects of agreed results... but I still went with official records. If Fitz was credited a victory in that match, he would be ranked fourth during the decade, and Sharkey would drop down to fifth place.
Your ratings are interesting and especially because they show you know your HW's across the total timespan 1890-2008. Would really like to know the methodology so we can see why certain results occur. For example does it take into account the form of an opponent up to the time of a fight or his career form in total? The late Mike Paul had a similar approach with his computerised ratings, but in greater detail and his results really hold up IMO.
Nice first post! I feel the names in green should be rated 1-4 spots higher than where you placed them, and the names in red shoulds be rated lower 1-4 spots lower than where you placed them..
