Division depth according to boxrec

I started to get first results and first ranking changes. And they are puzzling to say the least. I was hoping that boxing ranking works similarly to chess ranking (or any ranking, basing on ELO system). Does not look like this is the case. Over this weekend there was 11 fights, where I had ranking for both fighters previously recorded and Boxrec shows the ranking change. The thing is - in the majority of cases the winner gets substantially more points added than the loser subtracted. I thought that it shall be roughly the same (plus-minus a point for rounding). But it isn't!!! For instance - Liam Smith won against John Thompson. Winner's ranking changed from 163 (on oct 7th) to 273 (today). Loser's ranking dipped from 178 (on oct 7th) to 117 (today). Winner got 110 ponts, while loser lost 61 points. 49 points differential appeared out of the thin air!!! In other cases the differential is not that outrageous... But still noticeable...Ustinov beat Mo Harris and got 27 points added, while Harris lost just 15 points. There are also cases when ranking drifts without the change of record... Saul Canelo Alvarez had a ranking of 1074 last Wednesday, but today his ranking is 1075. Why???? Stop being a fuccboi and just read this. http://boxrec.com/media/index.php/BoxRec_Ratings_Description all your questions will be answered

I started to get first results and first ranking changes. And they are puzzling to say the least. I was hoping that boxing ranking works similarly to chess ranking (or any ranking, basing on ELO system). Does not look like this is the case. Over this weekend there was 11 fights, where I had ranking for both fighters previously recorded and Boxrec shows the ranking change. The thing is - in the majority of cases the winner gets substantially more points added than the loser subtracted. I thought that it shall be roughly the same (plus-minus a point for rounding). But it isn't!!! For instance - Liam Smith won against John Thompson. Winner's ranking changed from 163 (on oct 7th) to 273 (today). Loser's ranking dipped from 178 (on oct 7th) to 117 (today). Winner got 110 ponts, while loser lost 61 points. 49 points differential appeared out of the thin air!!! In other cases the differential is not that outrageous... But still noticeable...Ustinov beat Mo Harris and got 27 points added, while Harris lost just 15 points. There are also cases when ranking drifts without the change of record... Saul Canelo Alvarez had a ranking of 1074 last Wednesday, but today his ranking is 1075. Why????

Average is not calculated "against" anything ))) It is just an average. IDK what "smaller words" shall I use... In each division only top 100 was used, therefore the number of fighters below 100 did not matter. While I agree that total number of ranked fighters in the division is also a sign of depth, it is a different matter all together. A better way to frame my question: what is your information trying to tell me? The #95 guy at minimumweight, from a talent level, isn't as good as the #95 guy at welterweight. BoxRec has changed their setup to list every fighter on the lb for lb database; rather than taking 100 fighters from each of the 17 weight classes (1700 fighters, if the full list was done), why not just look at the top 1000 fighters on BoxRec's p4p list? Maybe I'm dense, but I don't see the usefulness of the initial chart.

BoxRec rankings are rubbish therefore so is this

Bullshyt.Light Heavy and Cruiser are the most depth divisions.

They have formal criteria of activity in place. Grant had last fight 2014-10-24. If he does not fight for the next couple of weeks, I guess he will be out of ranking list.

Average against what, exactly? BoxRec rates near 1800 fighters at 147, yet rates under 300 fighters at minimum weight. Every weight class bantamweight on down has fewer than 800 fighters rated by BoxRec. Beyond that, there are also "zombie fighters" rated (Michael Grant, Kevin Johnson, and Jonathan Banks are all in the heavyweight top 100, for example). Honestly not sure what you're seeking to measure. I was just about to post how this analysis doesn't and by still rating guys like Grant is madness. I agree ,welterweight probably has the most depth but there is noway heavyweight has more depth than light welter, light middle, feather/super feather and superbantam. No way in hell. That's where the analysis falls apart. Plus, you can't measure depth simply by how highly a fighter or fighters are ranked within that division, because a name like Pacquiao would skew the statistical data in the welterweight division. Take him out, what would happen? Or Wlad at heavy. Nice try by OP though, as it probably took a lot of work but I feel boxing is very different from say cricket or baseball where stats can mean a lot. In boxing, I don't think they work out so well.

We doing this stiff, slower upper division fighters beat faster, mobile lower division fighters again? Boxrec gives more worth or points to higher weight class fighters thus inflating their ranking. According to boxrec, a win over Viloria is worth less than a win over Robert Guerrero, and thats just bullsh*t. Thats why if you check their national P4P lists, you'll see someone like Estrada next to someone like Soto-Karass and Humberto Soto. lol Flyweight is the very last one. :rofl: :nonono: Lighteweight, Middleweight and Junior Middleweight, three dead wasteland divisions are near the top. The highest lower division weight class is Featherweight which is still very low. And thats ignoring the fact Flyweight is like a million times better. lol Don't people usually just make fun of boxrec rankings since you know, they are computerized. But I guess when there is an agenda... :lol1: Are you sure that Boxrec gives more ranking points to the higher division fighter if the respective rankings of two fighters before the fight are the same? This means for instance that we have fighters A and B at middleweight with say rankings 300 and 400 and fighters C and D at bantamweight with the same rankings 300 and 400. And you say that if A(300) beats B(400) and C(300) beats D(400) then A gets more points than C. You maybe right, but I want to be sure of that myself... This is the main goal why I started to collect these data - to figure out how it works...

We doing this stiff, slower upper division fighters beat faster, mobile lower division fighters again? Boxrec gives more worth or points to higher weight class fighters thus inflating their ranking. According to boxrec, a win over Viloria is worth less than a win over Robert Guerrero, and thats just bullsh*t. Thats why if you check their national P4P lists, you'll see someone like Estrada next to someone like Soto-Karass and Humberto Soto. lol Flyweight is the very last one. :rofl: :nonono: Lighteweight, Middleweight and Junior Middleweight, three dead wasteland divisions are near the top. The highest lower division weight class is Featherweight which is still very low. And thats ignoring the fact Flyweight is like a million times better. lol Don't people usually just make fun of boxrec rankings since you know, they are computerized. But I guess when there is an agenda... :lol1: People make fun of individual fighter rankings since sometimes their algorithm results in a fighter slipping through the statistical net and being placed way too high or way too low. You aren't going to find a better tool for rating entire divisions.

Cool idea. I wonder how the welterweight average will look once Manny retires.http://gifrific.com/wp-content/uploads/2012/10/Felix-Baumgartner-Free-Fall-Red-Bull.gif

Average against what, exactly? BoxRec rates near 1800 fighters at 147, yet rates under 300 fighters at minimum weight. Every weight class bantamweight on down has fewer than 800 fighters rated by BoxRec. Beyond that, there are also "zombie fighters" rated (Michael Grant, Kevin Johnson, and Jonathan Banks are all in the heavyweight top 100, for example). Honestly not sure what you're seeking to measure. It's not really boxrec's fault that very few adult males are minimumweights. Perhaps in some "speed and reflexes per pound per 1,000 population" scale minimumweights beat everyone else but in reality a minimumweight's biggest combat advantage is being small enough to even compete within such a thin roster.

I knew heavyweight would be at or near the top. The stars aligned and boxing hipsters & USA-chanters both had the same thing to complain about, but in reality it's very deep and has been for some time.

I am also a lifelong junkee of tennis stats. In the past I tried to contact ITF, ATP and WTA. They never respond. So I reckon that Boxrec will be the same and didn't try. It is quite possible that these databases have commercial values, so why would they share it. Be thankful that they post it on the web in the more or less convertible format. Could be worse... oh true that hmm...

There is something I din't pay attention to until I started to look closely to the result's page. W,L,D and NC in the result column are clear and self-explanatory. But what "P" and "S" mean? I thought that it has something to do with the result not yet approved by commission... But what are these letters stand for?

damn mick, is there a way for us to contact boxrec to request access to database? I am also a lifelong junkee of tennis stats. In the past I tried to contact ITF, ATP and WTA. They never respond. So I reckon that Boxrec will be the same and didn't try. It is quite possible that these databases have commercial values, so why would they share it. Be thankful that they post it on the web in the more or less convertible format. Could be worse...

It took me 3 hours or so to copy-paste pages. But now I have it as MS Access. Of course only top 100 in each division. But I can run whatever SQL I like on these data )) damn mick, is there a way for us to contact boxrec to request access to database?

Nice exercise, not sure I agree with these stats though. HW has some interesting up and coming, but apart from Wlad they are all pretty untested. I do not think it's amongst the "deepest" divisions. MW , I assume, is so high because of Cotto and Canelo, both coming from lower weights.Neither of them has done much at MW. While i find SuperBantam quite low. Rigo , Quigg and Frampton make it a pretty decent weightclass, i find. I do not believe that boxrec ranks Canelo as a MW at this point in time. I believe he is still considered to be a Jr.MW.

Average against what, exactly? BoxRec rates near 1800 fighters at 147, yet rates under 300 fighters at minimum weight. Every weight class bantamweight on down has fewer than 800 fighters rated by BoxRec. Beyond that, there are also "zombie fighters" rated (Michael Grant, Kevin Johnson, and Jonathan Banks are all in the heavyweight top 100, for example). Honestly not sure what you're seeking to measure. Average is not calculated "against" anything ))) It is just an average. IDK what "smaller words" shall I use... In each division only top 100 was used, therefore the number of fighters below 100 did not matter. While I agree that total number of ranked fighters in the division is also a sign of depth, it is a different matter all together.

Average against what, exactly? BoxRec rates near 1800 fighters at 147, yet rates under 300 fighters at minimum weight. Every weight class bantamweight on down has fewer than 800 fighters rated by BoxRec. Beyond that, there are also "zombie fighters" rated (Michael Grant, Kevin Johnson, and Jonathan Banks are all in the heavyweight top 100, for example). Honestly not sure what you're seeking to measure. # of fighters in a division is also a sign of depth......and please, boxrec has records of minimum weight fighters with records like 1-10 in filipinas, japan, corea...not being in boxrec probably means your boxing comission is non-existent. "zombie fighters"? kevin johnson fought in may....he is an active fighter......that's what it takes for boxrec to rate.......

You didn't miss anything by not doing it below flyweight.....hell 100 is half the minimum division :lol1:.... And as expected the bottom 4 are the lightest weights in order..... :lol1: hipsters of boxing will not even see value to these, because "eye test doe"..... :lol1: