Rating My Own Fighters: Amateur Questions

[Chadness]

I've begun the project of rerating fighters and have done a good job but now want to ask some things to be able to go back and be as accurate as possible.

- Proficiency, what is this?

I think I have an idea of it but there could be a different meaning. Is this like an experience statistic? 8 proficiency would mean a guy at the 8 round level so his endurance rating would be pro-rated to reflect that? Or would it mean the guy gives 8 good rounds and tends to take his foot off the effort pedal from round 9 on?

- What are some tips for Punching/Missed/Clinching?

I use 34 as average for landed (22+46 = 68/2 = 34) and the same formula to get 49 as average for missed. Using this, numbers are generally lower than what are given and to me has produced more realistic fights.

I'm only doing Welterweights right now and of course the fighter with the best numbers here is Floyd Mayweather with 38 landed and 40 missed. On average he produces Mayweather numbers with 48%+ landed and doesn't give up much to counter. Manny Pacquiao is, I believe, 43 landed and 58 missed. He throws more punches, puts up a respectable 38%+, and gives up a realistic connect average in the mid 20's. Most everybody following along the same system feels right but a few are giving me trouble.

For instance, Zab Judah is a punching machine. I think he is 36 landed and 60 missed. The idea being he can land better than average on speed alone but he is also somewhat inaccurate for a guy as fast as he is. If not for the high Punches Missed, he'd be similar to Floyd Mayweather. With the high Punches Missed, he has a pretty high average of punches per round. What do I do to make a fighter more inaccurate but not because he throws in high volume?

And in general what I want to understand is Clinching/Running numbers. I use the same 49 as average but feel I definitely am missing out on intricate details that may better determine if a guy is more or less active in a round.

- What's up with Carlos Baldomir!?

This is a guy that wasn't a terrible fighter. Slow, accurate, and only knew how to win with volume. I have his style replicated as well as possible but he always gets pounded on and stopped waaaay more than he should be. I believe I have 32 landed and 60 missed. He throws a lot of punches, good. He lands an accurate percentage, good. He also tends to gas out REALLY badly before he really can get going. Here are some other stats...

Counter: 40, to replicate his timing. He had a weird way of actually counterpunching with timing rather than leading. This is probably why his connect percentage is very good.

Fast Starter: 4, he was a slow starter. I think I may need to go to 3?

CF: 6/6, my rationale being that he really had to fight at the stylistic disadvantage of his opponent. He was too slow to box and too featherfisted to slug. I use 8 as average value.

HP: 3, feels right but may be low in practical use.

3pt Punches: I didn't give him much and wonder if this is a problem. He really didn't throw much beyond the right cross that landed hard and flush. Is he just getting walked through too easily by having a 3 on cross and no other 3pt punches?

Obviously he SHOULD be mostly losing as he's fighting a lot of elite fighters but some of the guys rated are accurately rated fringe condensers that also demolish him in the same way. Freddy Hernandez and Luis Abregu, mainly, the two guys that shouldn't so easily beat up Carlos Baldomir.

[sepia warrior]

I find discussions about rating fighters fascinating and always get something out of them. I welcome this thread. It's hard to understand the effect your rating changes have on these fighters without seeing all of the ratings for the fighter as each rating category influences the fighters performance with the exception of proficiency. For what they are worth, I share the following comments.

Proficiency seems to play no role in TB2.5. It's there because it will be (was supposed to be) used in TB3 to help sort fighters in the Auto-Scheduler. It indicates the number of rounds the fighter usually fights. 4=beginner, novice; 6=beginner or pre-prime fighter gaining experience; 8=fighter nearing prime; and 10= experienced fighter, main eventer. Most of the fighters rated in the game are a 10 proficiency because they are experienced at that distance. I assign lower proficiency ratings to fighters who only have had few fights. I use their records on BoxRec to help in that determination. For example, if a fighter has fought only 4 and 6 round fights, I'll give him a six. If there's a fighter who has fought mostly 6-8 round bouts and a single 10 rounder, I would still give him an 8.

I'm not sure what you are doing with the Punches Missed category. The manual states that the normal range is from 57 to 64 although the game allows a range of 21 to 78. (Interestingly, the game only allows a range of 24 to 46 for Punches Landed and CounterPunching. Go figure) Anyway, the game numbers range up to 78 total sequenced in the following order - Punches Landed (PL); Punches Missed (PM); Clinches (CL) and Ring Movement (RM). Counterpunching (CP) only comes into play when the opponent has missed and a counterpunching opportunity has arisen. A typical rating for an average generic fighter would be 1-34 (PL); 35-60 (PM); 61-69 (CL) and 70-78 (RM) Without any adjustments for the defensive rating of his opponent, he would land punches approximately 44% of the time, miss 33% of time; clinch 11.5% of the time and move 11.5% of the time when he is in control. The number of punches attempted is a function of the Control Factor - the more a fighter is in control, the more opportunities he will have to punch. The worse the opponent's defense factor, the more accurate the fighter's punches will be. Thus, if the above generic fighter fought a poor defensive fighter with a 4 DEF rating, his punches landed would be 38 (48.7%) in that fight as long as both fighters had endurance points remaining. The converse would be true if he was fighting a skilled defensive fighter. These numbers change further if either or both of the fighters exhaust their endurance points.

Everything starts with the Control Factor (CF). The fighter who is in control most of the time gets to dictate what happens. Obviously control shifts back and forth but the fighter with the better control factor has an advantage. Nonetheless, there are things which can negate that such as defense, the opponent's punching power, the fighter's susceptability to be hurt or knocked down by a big punch, lack of endurance points, cuts etc.

Carlos Baldomir is a tough guy to rate because of his style or lack thereof. His endurance and durability ratings should be very high. 9 for Endurance and a 1 or 2 TKO rating. The official rating gives him a high CF because he has to be in control in order to avoid being pounded. He doesn't have much power but the high CF, END and PL numbers mean that he will wear down most of his opponents. His stoppages will come later in fights and will be of the TKO variety rather than clean knockouts. His relatively high aggressiveness rating means that he will start in control after a close or even round and get points if the round is close. The official rating gives him an 8 killer instinct rating to enhance his low hitting power rating and absence of 3 point punches. The killer rating means the number of opportunities a fighter gets to punch without having the game check to see who is in control. Even with unorthodox or unconventional fighters, I try to avoid extreme ratings in minor categories because even they can distort performance.

Finally, a fighter rating doesn't exist in isolation. It has to be measured against other fighters and I find it useful whenever possible to test fighters I rate against their actual opponents who are officially rated in the game or rated by some of the real veterans who've contributed to this forum.

[sepia warrior]

An amendment to my comments. The Baldomir rating I looked at was the "old" official rating rather than the most recent one. In the current rating, Baldomir's Killer Instinct was downgraded to 5 and he was given a few 3 points for 3 point punches. He's not as effective when he hurts an opponent but he has a marginally better chance to get someone in trouble. Are you there Zab Judah? The other things that jumped out at me were his conditioning and strategy ratings. His conditioning rating of 1 means he'll be in top condition 90% of the time and his coverup rating of 4 will help him to avoid stoppages.

[Chadness]

What I don't understand is how it is a "range" of numbers when I can only give one value. In your example fighter would I have...

Punches Landed: 34
Punched Missed: 60
Clinching: 69

as the values? I mean I sorta understand that too but not so much the ability to set nonsensical ranges. What if I set 40 Landed and 34 Missed? What the heck is happening there? Am I technically making the guy less accurate as 1-34 is Missed and 35-40 is Landed? Perhaps an interesting way to rate Mayorga if that's what I'm doing - making a guy miss more than he lands but not by volume.

Time to play with what you've said.

[sepia warrior]

When you put in 34 for Punches Landed, as it is the first value, the game understands it as 1-34. If your Punches Missed is 60, the game identifies the range of numbers as 35-60. The same applies to Clinches and Ring Movement up to 78 which is the end value. If you want Mayorga to miss more often, you can increase his Punches Missed value beyond the suggested high rating of 64. You could make it 66 or any number up to 70. In doing so, you will be decreasing the numbers available for clinching and ring movement. Or you can decrease his Punches Landed number so that the Punches Missed range is wider. (To show the game's flexibility you can even compensate somewhat for the his lower connect rate if you reduce his punches landed by increasing the % of his 3 point punches.)

The way the computer game works now in its latest iteration is pretty much the same as its parent, the board game Title Bout. While you can do many things with the ratings for a particular fighter, that fighter will still have to interact with other fighters who have been rated using the standards developed by the Trunzos and refined by them and others as different computer versions have been issued. Nonetheless, the rating categories provide different paths to accurately rate a fighter. Testing your ratings will tell you whether you've modeled the fighter the way you want.

[Chadness]

I've finished Junior Welterweight using the idea of it being a range rather than individual numbers in mind and gotten good results but now your post brings up a new thing. 2pt punches have better accuracy? I've not been doing anything with that in mind. Just rating 3pt as a mix of the power and flushness to come up with a value for the amount of those punches the guy will land big. Then 2pt just to get to 100. That doesn't mean more 3's than 2's but I haven't rated based on Hitting Power like it said to do. I rate 3pt on the idea of those being the punches that are memorable and win rounds so for instance Paulie Malignaggi has some 3pt jabs as he lands solid enough to win rounds off those alone. I had wondered what would happen if a guy had a LOT of 3pt punches because all they land are haymakers. Any clue if the difference in accuracy is enough to offset the advantage of so much extra power and keep things 'fair'? Cause while thinking of Mayorga I thought he either blatantly misses or lands a big one. Little bit the same with David Haye.

[sepia warrior]

There is no difference in accuracy between 2 pt. punches or 3 pt. punches. Accuracy is determined by the Punches Landed rating modified by the opponents defense rating. The notion is that not every punch landed does damage or hurts the opponent. 3 point punches are the "big" punches which not only pile up points but could cause an opponent to be knocked down or so dazed that the killer instinct rating kicks in allowing the puncher a number of actions without the game conducting a control check. (Every time there is a 3 pt. punch, the game conducts a check to see if causes a knockdown or some other major damage) Thus, there is a correlation between Hitting Power and the number of 3 point punches allocated to a fighter. The more power a fighter has, the higher the number of 3 pt. punches.

The suggested number of 3 point punches based on HP is as follows: 1 HP = 0-3; 2 HP = 3-5; 3 HP = 6-9; 4 HP = 10-12; 5 HP = 13-15; 6 HP = 16-20; 7 HP = 21-25; 8 HP = 26-30; 9 HP = 31-36; 10 HP = 37-42; 11 HP = 42-44; 12 HP = 45-47; 13 HP = 48-50; 14 HP = 51-54; 15 HP = 55-57. While there is nothing that says you can't vary these ranges, allocating too many 3 point punches to a fighter will distort his performance.

There are a number of posts devoted to rating fighters and some articles on the website. Another way to understand the rating system is to select some boxers with whom you are familiar. Examining their ratings will give you an idea of how the rating categories are interrelated and can be used to identify the fighters strengths and weaknesses. The official ratings are really very good and are the product of refinements made over and over again. The Fighter List is also useful for seeing how the ratings have evolved. Fighterlist If something strikes you as wrong with a rating, try tweaking it (making small changes in key categories) to see if the fighter performs more realistically with those changes.

Once again I hope this helps.

[Chadness]

I've done a good amount of fighters. 135, 140, and now back at Welterweight having done Oscar, Manny, and Floyd. I've gotten very good results. There's a good mix of knowing who has an advantage and knowing an upset can happen with almost any fight. Of course, Manny and Floyd are pretty dominant. Floyd usually wins between them.

Reading their fights I've noticed subtle differences that make them interesting. Floyd is the +/- king like IRL as he usually lands close to 50% and holds opponents somewhere just under or above 20%. He wins rounds with precision. He tends not to get the stoppages he could with more aggression but almost always makes his mark.

Manny Pacquiao rounds I've noticed tend to be won by Manny after scoring large volumes of points near the ends of rounds or with just a steady stream of aggression adding up while the other guy goes defensive during his assaults and gives up the round lead. He scores a lot of stoppages. Something funny is that aside from the fact Mayweather can beat him is Juan Manuel Marquez is the only other one that really challenges him. Makes me feel I've done a good job, lol

[sepia warrior]

It certainly seems like you've done a good job. You should post some of your ratings so we all can benefit from your approach. I always learn something new from these discussions or from looking at how particular fighters are rated whether it's the official rating or by one of the posters in this forum.

[Jackyle]

great thread guys, very interesting to see different ideas toward rating fighters. Would love to try some of your rated fighters out Chadness.

[bear]

Originally Posted by Jackyle

great thread guys, very interesting to see different ideas toward rating fighters. Would love to try some of your rated fighters out Chadness.

Ditto from me.

[Chadness]

I'll put a division up then. Right now I'm watching Prescott vs McCloskey and going to watch other similar fights that'd help me get a better idea of fighters as current as possible. Prescott is rated as he was - a bum. He looked really good against Bayan Jargal but I ignored that. Now he's looking just as good against a Paul McCloskey that made Amir Khan look bad.

Other fights on tap as refreshers:

Vernon Paris vs Tim Coleman
Vernon Paris vs Emanuel Augustus
Urbano Antillion vs Miguel Acosta
Victor Cayo vs Julio Diaz
Ricardo Torres vs Mike Arnaoutis
Devon Alexander vs Juan Urango

I'd probably let you guys play around with the 140 pound fighters as there are a lot of guys there and it feels the most like anybody could win any fight.