For Whom the Bell Toles (Roscoe)

[mking55]

First a disclaimer on this post and any future ones I make along these lines. I am not a boxing expert, historian, connoisseur or anything of that nature to do with boxing.

I have watched various fights over the last 40 years or so. My all-time favorite is George Chuvalo, so that shows what I know.

As I have mentioned in 2 of my larger posts on here (Lem Franklin and Greg Page), I want to understand the ratings. All I can do is look at the records in boxrec.com and try to see where the record fits with the rating. If I don't understand it, I ask.

I never want anyone to think I am trying to say I know better. What I am always trying to say is, "What does this mean? And if I am not interpreting it correctly, then please show me my error".

That's it by way of preamble. Now on to this guy. (cue Rod Serling).

Presented for your consideration, one Roscoe Toles, a heavyweight boxer who had a career in the ring from approx. 1934 to 1950.

Roscoe Toles had a career record of 35-10-6. That is better than some and not as good as others. He won 69% of the times he entered the ring. He is rated a 9 in the game.

Leaving out 'Old' George Foreman, there are 20 fighters rated a 9 in the game. Toles has the 3rd worst record. There is Jack Sharkey who have I not looked at, and Jimmy Young who I will save for another post.

Of these 51 total fights, Toles fought 33 different fighters. He fought Arturo Godoy 8 times, Alberto Santiago Lovell 4 times. From November, 1941 to August, 1943 he fought only in South America. 7 times against Godoy, 4 times against Santiago Lovell, and 1 time against some other guy named Lovell. It looks to me like that is not the stuff that 9-rated fighters should have. His record during this time was 8-1-3.

Toles is the highest rated fighter of all the fighters he fought. They are either in the game with a lower rating or not rated at all. I make an assumption that if a fighter isn't rated, then he is a bottom of the barrel type guy. However, like I said, that is an assumption on my part.

I took a look at all the fighters that Toles went toe to toe with, and who they fought in common. I cannot see anything outstanding about Toles. What am I missing?

He fought Buddy Walker (not rated in the game) 4 times. He won 1, lost 2, and they drew once. He lost to light-heavyweight Maxie Rosenbloom. He lost to Bob Pastor and Abe Simon who are both rated in the game with a lower number.

He had 2 victories over fighters whose records say that was their only fight. He beat 4 fighters where the fight was either the last fight they fought, or as part of a losing streak that ended their career.

If I look at the other 9-rated fighters in the game, almost all of them at a glance look like the better fighters in the game. Toles doesn't. Compared to them, he looks weak. Compared to his real life competition he looks okay.

I would like someone to educate me on why Roscoe Toles is worthy of his numbers.

I have included a little spreadsheet I did at the bottom of the post. It's a mess but it makes sense to me. Sheet 1 is eveyone Toles fought and a brief note or 2.

Sheet 2 is a cross-reference of fights in common and their results. Reading from left to right (row) contains the correct fight result. I used short forms for the results and I hope they make sense. For example, W-KO-4/L-P-10 would mean they fought 2 times. One was a victory by KO in the 4th, the other was they were outpointed in 10.

Sheet 3 is a list of the 9-rated fighters in the game and their actual records.

The bottom line is that all this rambling is a question. Why is Roscoe Toles rated a '9'?

roscoe toles.xls

[crusadecat]

I don't have an aswer for you but enjoyed reading your informative post.

[swampdragon]

One thing I've noted is that all 9s are not created equal. Toles has 1 HP and a great chin, so most of his fights will go to decision. His other virtue besides his chin is that he will land a lot of punches. Just for fun, I fought him 100 times against Abe Simon, an 8, after I read your post. Simon won 46-37-17. I suspect if you ran a real analysis as you did with our buddy Lem, you will find that Toles actually performs more like a 7 than a 9. Take the overall ratings with a grain of salt, because the rating scoring system cannot always predict performance. A 1 HP is a major handicap.
I am also no expert, but it does appear to me that Toles was a decent Midwestern fighter who never really performed at the top level. You can certainly argue that he ought to be a 4 or 5. Maybe one of our experts on the time period can weigh in.

[djday45]

I would laso agree that the overall rating can be very misleading. As Jim's Joe Grim article shows although Grim is rated a "5" he will not win many fights.
I suspect this may be the same with Toles against competition in the 6+ skill level.
Have you ran fights agaisnt his competition in the game? this will tell you if his (or other) ratings need tweaking.
Also remeber to always take into account carrer stage in to real life results. Some boxers have very short career primes while others seem to go on forever! this can have a huge impact on real life results as can conditioning.
Thks for picking out these guys to discuss, as you say it increases all our knowledges of how the ratings work and how to rate fighters.

Rgds
Dean

[mking55]

A couple of comments on the comments so far.

I see that not all 9's (or whatever) are created equal. That is fine. Every fighter should perform uniquely against every other fighter. I do not think the overall rating is meant to be a definitive rating, but just a guideline of where the fighter falls.

To me, 2 fighters rated '9' (for example) may each fight the same 100 fighters and win 75% of their fights. I don't expect them to win all their fights against the same fighters. Fighter A and Fighter B may both win their fights against the fighters numbered 1 to 50. Fighter A wins 25 against fighters 51-75 and loses against fighters 76-100, while Fighter B loses against fighters 51-75 and defeats fighters 76-100. (I hope that makes sense).

But this way they each win approx. the same number of times, but they accomplish it in different matchups. They are not the same '9' fighters, but they both reside in the same neighborhood as far as sucess probibility lies.

To me Franklin was a case of him winning too many fights in this game (his rating did not justify his winning % in the game), and as Jim said a couple of his numbers needed changing. Toles, to me anyway, is a case of his overall rating number looking 'funny' to me in comparison to other fighters in the game, and his actual performance against these fighters in real life, and the unrated fighters who didn't even make the game database.

I suppose I could fight Toles against the masses, but he doesn't grab me the way Lem Franklin did. And I need to pace myself for the coming assault on Jimmy (9) Young. His record was 34-19-2. This is another one I don't understand but I need to calm down first so I can speak calmly about this.

Plus, any explanation from Jim I get about Toles, may help me to understand Young. And others.

And yes, it is the discussion that is so important, because I just love learning about the fighters and the fight game, and this game.

[darthpb]

Have you taken into account the era in which they fought? I sense an onslaught coming against Young but if you take him out of the 70's(the best crop of HVYs) he could have been the top boxer in a lot of eras thus the 9 rating...I guess. Some of these other guys are before my time so I won't comment

[Cap]

This discussion is great stuff, Martin. I personally think that there is too great a disparity in the OR (overall rating). Having Ali a 15 makes him seem like an unbeatable god-like warrior when compared to others in the sim. I've gone through (in TF2K1) and increased the OR on many fighters because I thought they deserved better. Maybe it doesn't really have that great a significance, but there is an unshakeable perception that anyone with an OR below (fill in the blank) is pretty much dog food.

One of the guys whose OR I increased was Sam Langford, a personal favourite and present World Champion in my universe.

Cap

[vistaman44]

Since the 1940s is my area of interest, let me chime in, too. First, I agree that a 9 is high and that the overall rankings are sometimes deceiving. Essentially, djday and dragon are right on with the technical analysis side, so here's my take on the opposition. Please take this only as context, and not as an argument for a 9, because I would make him around a 7:

First, it was not at all uncommon for middles to fight lightheavies and lightheavies to fight heavies on a regular basis back then, and the guy from the heavier weight class didn't win nearly as high a percentage of the time then as he would now.

Walker rates about a 4. He wasn't a prominent guy from the era, but he did fight quite a few of the time's contenders and the developers have included a lot of fighters from the heavyweight ranks of that time period. I posted my rating of him a while back and it is at CornerWork.

Pastor and Godoy were solid contenders. Simon and Adamick were also pretty good at the time Toles faced them.

Louis was an aging champion who was once THE BEST. Delaney, Barlund, Dorazio, Hart and Peterson were all briefly contenders (by that I mean top 15 in the world) turned trialhorses.

Belanger was at the end when Toles fought him, but was a useful trialhorse in the 1930s. He would probably rate a 2-3, though I haven't ranked him, yet.

Vorce and Schott fought at both LH and heavy, and were 1-2 type trialhorses.

Bolden was a pretty good LH who would fare well against 5-6 competition in that weight class, and I rated him a 5. Rosenbloom was a LH champion who moved and fought fairly successfully as a heavy, well enough to be ranked in the top 10 by Ring for several years.

Perroni and LQ Murray are in the game, and both were pretty good fighters and contenders. Toles beat both.

Dudas and Santiago Lovell were 2-3 level and were briefly contenders. Alf Big Boy Brown ended up with a losing record, but was ranked early in his career by Ring and fought several of the era's "names."

All of that is to say that the numbers and total results don't tell the whole story. Toles wasn't in the top quarter of his time, as a 9 might indicate on some level, but he was a force to not be taken lightly.

[swampdragon]

I do not believe that all 9s should necessarily perform at the same level. In this case, Roscoe may only win 50 fights against fighters 1-100 while another 9 is winning 75 fights. I think he would probably perform as a 7. Vistaman is an expert on the period, and he thinks Roscoe deserves about a 7. If you tweak his ratings so that he is rated as a 7, then he may perform as a 5. If there is a problem, it's with the rating formula in general. Lem was rated a 6 by the game and performed as an 8 or 9. The problem wasn't the 6 rating, it was the level of performance. Unless Roscoe actually wins the fights a 9 should be expected to win, there's no problem.

[jabbo]

Martin,

I find myself disagreeing strongly with the overall ratings of fighters in the game, but am too new to attempt to rerate them myself. For example, Alex Stewart, modern heavyweight, is rated a 7, while Jerry Quarry and Ron Lyle are rated a 6. Mike Weaver is a 5.

But I recently simmed Foreman (prime) against Lyle, and Lyle gave Foreman all he could take; Foreman pulled out a UD-10 with a knockdown in the last 40 seconds of the bout.

Maybe the problem in some cases is the formula for the overall rating needs tweaking.

Bob

[vistaman44]

Dragon;
I appreciate the vote of confidence, but expert is probably too strong a word. I have followed the fighters for more than 35 years and am very well-read and interested in the history of that era in boxing, including the not-so-common names. Beyond that, however, I agree completely with both of your posts in this string!
LOL

Keep up the good work!

[mking55]

Originally posted by swampdragon
Lem was rated a 6 by the game and performed as an 8 or 9. The problem wasn't the 6 rating, it was the level of performance. Unless Roscoe actually wins the fights a 9 should be expected to win, there's no problem.

I agree. It is not the overall number but the level of performance. However, I would like to see the overall number be reflected in the level of performance. It is with most, but not with all. And, I don't want to 'tweak' ratings myself because as I have mentioned several times, I don't feel qualified.

Lem was a 6 who fought like a 9. Now he's a 4 who fights like a 6. What does 'like a (insert number)' mean?

To me, it means that the performance of the fighters and their percentage of victories against all fighters are similar. Is that correct? I would like to know.

I don't want to see 6's who fight like 9's, and 9's who fight like 6's (on an average). Then the overall rating has no meaning. I agree that sometimes a 6 will fight like a 9, or vice versa, depending on the particular opponent.

The problem is when you give a rating as a number, it gives the appearance of an overall fixed value. Therefore 9 equals 9 equals 9. Maybe you can't create an overall rating because it is too deceiving? Maybe the overall ratings should be more open, like Excellent, Good, Average, Fair, Tomato Can.

Food for thought? Reasonable? I like these discussions as they help me understand things and learn more about the fighters and their interaction. I am also curious what Jim or Tom has to say about Mr. Toles, overall ratings, and the price of tea in China.

[Claybor]

Keep em coming Martin, your questions are very similar to many that I have, and I am sure quite a few others in the community feel the same.

[cubedrum]

The reason that Rosco (correct spelling lifted right off of Rosco's trunks) is rated so high is that his lack of hitting power mandates a high CF. The CF attribute is given too much significance in the overall rating calculation in my opinion. This has been the case with all of the controversially rated fighters. Lem Franklin, Joe Grim and now Rosco Toles all have little hitting power and a high CF.

[vistaman44]

Good point and nice pic, cubedrum!

[mking55]

Originally posted by cubedrum
The reason that Rosco (correct spelling lifted right off of Rosco's trunks) is rated so high is that his lack of hitting power mandates a high CF. The CF attribute is given too much significance in the overall rating calculation in my opinion. This has been the case with all of the controversially rated fighters. Lem Franklin, Joe Grim and now Rosco Toles all have little hitting power and a high CF.

"AHA", he exclaimed.

This is possible. However I must ask you, why do you make the claim that, "his lack of hitting power mandates a high CF"?

Why does that "mandate" it?

[Montreal]

<< I agree. It is not the overall number but the level of performance. However, I would like to see the overall number be reflected in the level of performance. It is with most, but not with all. And, I don't want to 'tweak' ratings myself because as I have mentioned several times, I don't feel qualified. >>

Hi Martin and All,

I don't own the game yet but I used to play with the ABPA computer game released around 1996 or 1997. Actually, I tend to think that what is called the Overall Rating is simply not an OR at all. It seems to be there to add or substract something to a fighter's capabilities. To understand what I mean, don't change anything but the OR of the guy you're talking about and run a few series of 500 fights against the same opponent. For instance run 500 fights with him rated at 9, then 500 rated at 6 and 500 rated at 12.

In the APBA game, this change in OR has dramatic effects. I just ran a few series of 1000 fights between Trevor Berbick (7) and Joe Bugner (7 also), except that I changed Berbick's OR to 4 and 10 in the two last series...

Berbick (7) vs Bugner (7) ... 438-455-107 (roughly 50-50)
Berbick (4) vs Bugner (7) ... 148-781-71
Berbick (10) vs Bugner (7) ... 702-228-70

A few years ago, I ran a 75-bout round robin involving the 76 HW rated from 1 to 7 in the APBA game. To my surprise, the fighter with the worst record was rated 3 and there were two guys rated 2 that were next to him at the bottom of the list. Small sample size, probably, but I decided to do some testing with the so-called Overall Rating. That's how I discovered that the OR is likely not an OR. It's possibly there to correct some effects that the different individual ratings for each fighter can't deal with. Or something.

Is the same thing happening with the current version of the game? I can't tell, obviously.

Pierre

[mking55]

Montreal,

I tried what you did and I believe that OR is not used in the outcome of a fight. I fought Berbick vs. Foley 100 times. They are both 7's.

The results were:
Berbick (7) 26-63-11
Berbick (10) 31-61-8
Berbick (3) 29-60-11

Time for you to update to the new game.


It is magnifique!

[Montreal]

Sure, Martin, will do. This is just that it's not the exact right time for me to get totally immersed in such a thing - lol

Pierre

[cubedrum]

The CF mandate is for realism's sake. Otherwise light hitting fighters would lose a lot to fighters they actually beat. I'm just contesting the overall rating equation.

Cube