JCS wrote:Lennox wrote:The foundation is not flawed.
Boxers are not confined to a life in one division and their merits should be taken into account forever. You need a mechanism to factor division change and to factor length of time of the win/loss.
Any mechanism won't suit ALL boxers the same. Some go up and carry the weight, some don't. It's never going to get everyone exactly into the correct spot but it will get it pretty close. The Thurman-Garcia fight after all was a pretty even fight but their ranking is still not wildly incorrect.
I'm not saying the foundation IS flawed, but what you're doing is akin to what the English teachers in elementary school taught us...
When you define a word, don't use that word in the definition, which I think is what you are essentially doing.
I do realize it is difficult to try to prove the worth of something without scientific backing... it is even more difficult to try and argue your system over BoxRec's or vice versa, but you do keep trying to do it.
Personally, I hate an "all in" linear, ladder-based system, because it doesn't always work and nothing is worse in a rating system than hard cliffs and black/white views. The exception to this comment, is a sport that adheres to a system like this... which is not happening here. In my opinion, a good rating system has to acknowledge that there is definitely a gray area that needs to be observed. If you say you have a linear system than I can't see how Garcia beating a #1 at 140 has much bearing at 147. In a BoxRec system, it makes perfect sense.
If a not very dominant #1 fighter loses by KO to a #75... in your system, does the #75 become #1 and the #1 becomes #2? Is it that simple?
JCS- It is not a total all in ladder system because of a few things.
1. Fighters continue to have lesser fights after they have beaten a decent fighter, lets say fighter A (ranked 87) beats a fighter ranked 28.... lets say Fighter A moves to 35 and the loser drops to 36. If fighter A has a fight against the number 100 ranked fighter he still collects points, he might have several fights against lesser opposition and keep ascending.
2. Fighters can collect points against fighters in other divisions, it is the points that matter not the ranking number.
Garcia has a far better resume than Thurman going into the fight but Garcia's record is mainly at 140 he has only had a few fights albiet top ones at 147. In the PBO system a fighter is penalized for moving up from 140 to 147 his achievements at 140 are not taken 100% to welter. All things considered taking a % of his points at 140 he still figured quite highly at Welterweight and was ranked above Thurman who since 2013 when he acquired the WBA interim title against Jesus Karass defended against Julio Diaz (ranked 72) Leonardo Bundu (ranked 12) Roberto Guerrero (ranked 13) Luis Collazo (ranked 21) and then it was the Shawn Porter fight (ranked 6) where he showed he was a great fighter, up until then he had not really done so much to demand a very high rating.
If everyone stayed in the same division for their whole career it would be a lot easier.
If a #75 ranked fighter beat the number 1 then what happens is there is a points calculation with two new values, one ascending and one descending. If the #75 has no tags (defeats) he could go to number 1, more likely is that the number 75 would move to 3 or 4, the former number 1 would be 4 or 5 and the previously ranked 2 3 and 4 all move up a spot. Each situation is different. Zlatican was number 1 recently and Mike Garcia beat him, it moved Linares to number 1. Garcia to 2 and Zlatican to 3.
It is primarily linear based is not all in linear simply because fighters mainly avoid other rated fighters so the best matches are not always made. Big leaps upwards are made by beating good fighters and 'the ladder'. Some credit is given for fighting journeymen.
that's another thing. Why in the world would a computer system only publish monthly ratings? That puts the m in moronic.
