FORUM HOME > TNM > Discussion
Question to Oliver about Ratings

AnubisPosted on 09/30/04 at 09:18:28

How does TNM decide exactly what rating a match gets?  I know all the info is in the LOGS.TXT file, but I can't decipher this.

So I don't suppose you'd be willing to decipher the weird stuff?  Here is the log from my most recent main event match:

Match no. 0
Participants:          Jimmy Dragon Dawson Polaris Jerry Dunvegan Brandon Harwell
Total number of moves:  193
Total damage:           7905
Moves/time:             .128154050464807
Damage/time:            5.24900398406375
Damage/moves:           40.9585492227979
Average heat: 10 -> 2.5
Damage/move bonus: .5
Finish class: 2             bonus: 0
Workrate bonus: 5
Cumulative rating (so far):*** 3/4
Near falls:             32 and 13 - 32 and 13 per fall
Adjusted to:  4.25 - would have been 5
Final rating:      **** 1/4
Jimmy Dragon: 9 -> 9
Dawson Polaris: 9 -> 9
Jerry Dunvegan: 9 -> 9
Brandon Harwell: 9 -> 9

I don't suppose you'd be willing to explain each line of this?  I'm just profoudly curious is all.  I would especially like to know about the line "Adjusted to:  4.25 - would have been 5".  All of them would be nice, though.  Thanks!
Oliver CoppPosted on 09/30/04 at 11:08:03

On 09/30/04 at 09:18:28, Anubis wrote:Average heat: 10 -> 2.5
Takes the average of all the wrestlers' heat value and turns that into a basic rating. It's like in real life: matches where the fans are on the edge of their seats will seem deceptively good (see: Rock vs Hogan I) while even great matches without crowd reactions appear dull.


Damage/move bonus: .5
The more hard-hitting the moves in the match are - on average - the better it will be received. Think "restholds" vs "highspots".


Finish class: 2 bonus: 0
Finish classes:

1: Outside Interference
2: Finisher
3: Small Finish (e.g. small package)
4: Big Finish (e.g. stardust press)
5: Counter Finish
6: Submission Finish
7: DQ Finish
8: Countout Finish
9: Draw
10: Gimmick Finish
11: Unspecified Pinfall Finish / Foreign Object Finish

I won't go into the ranges of bonus / malus values for the finish types, though.


Workrate bonus: 5
The better the wrestlers' average workrate is, the higher this will be. Can also be negative =)


Adjusted to: 4.25 - would have been 5
Everything higher than **** 1/4 has to meet the criteria in the following areas:
- work rate
- match length
- good to great finish
- great heat
- many near falls


Jimmy Dragon: 9 -> 9
Dawson Polaris: 9 -> 9
Jerry Dunvegan: 9 -> 9
Brandon Harwell: 9 -> 9
After the match, everybody's base heat (the one they start out with) may be increased or decreased based on their performance in the match.
AnubisPosted on 09/30/04 at 11:17:36

How sly.  You answered the questions without "answering" the questions.  What I'm really interested in are the equations.  Questions like, what are the parameters for the ***** criteria you mentioned?  How does the work rate translate into the bonus?  How do all those massive numbers "translate" to the specific ratings?

Based on the numbers, I'm assuming the math behind it is actually simpler than it appears at first glance.  It's a hunch.  Something that each * requires so many points or something.  The near falls thing is the biggest mystery of all, I can't figure out what it's trying to say, honestly.

So did you just misunderstand the question, or are you being sly to keep the inner workings a secret so we can't figure out the "formula" for a ***** match?  Just curious there.  It's very interesting, and all equations always interest me.

In particular, although I have no clue about the actual programming, I think I've got TNM's "methods" figured out.  All except for this part, heh. :P
americamamushiPosted on 09/30/04 at 18:28:30

So did you just misunderstand the question, or are you being sly to keep the inner workings a secret so we can't figure out the "formula" for a ***** match?
He answered your question exactly how you asked it.  You asked him to explain each line and thats exactly what he did.  Not only that, but he tried to make sure that everyone knew exactly what he meant in each area.  Personally, I think he answered your question and went beyond what you asked, making sure even the casualest of fans would know what each line did.
91Posted on 09/30/04 at 18:54:05

So if a ***** has to meet criteria in every department, is that to say that pertaining to average workrate, it would be impossible for anyone below a certain workrate level to get a ***** in a singles match regardless of the quality of the opponent?
Critic of the DawnPosted on 09/30/04 at 19:17:21

I've gotten a ***** match between Edge (workrate 90) and Bob Sapp (workrate 50), so I doubt that's the case.

Eric "Critic of the Dawn"
Oliver CoppPosted on 09/30/04 at 23:16:38

On 09/30/04 at 18:54:05, 91 wrote:So if a ***** has to meet criteria in every department, is that to say that pertaining to average workrate, it would be impossible for anyone below a certain workrate level to get a ***** in a singles match regardless of the quality of the opponent?
No, that's the magic of *average*. Add up everybody's workrate and divide it by the number of wrestlers.
Oliver CoppPosted on 09/30/04 at 23:24:51

On 09/30/04 at 11:17:36, Anubis wrote:How sly. You answered the questions without "answering" the questions. What I'm really interested in are the equations. Questions like, what are the parameters for the ***** criteria you mentioned? How does the work rate translate into the bonus? How do all those massive numbers "translate" to the specific ratings?

Based on the numbers, I'm assuming the math behind it is actually simpler than it appears at first glance. It's a hunch. Something that each * requires so many points or something. The near falls thing is the biggest mystery of all, I can't figure out what it's trying to say, honestly.

So did you just misunderstand the question, or are you being sly to keep the inner workings a secret so we can't figure out the "formula" for a ***** match? Just curious there. It's very interesting, and all equations always interest me.

In particular, although I have no clue about the actual programming, I think I've got TNM's "methods" figured out. All except for this part, heh. :P
Actually, it appears simpler at first glance than it is. The reason why I didn't go into further detail is because then I'd first have to explain what a neuronal network is, how it works and how it is used in this case. Basically, a neuronal network is a program that learns from the input it gets and the output it generates.

Hence, no matter what your specific booking style it, the program will "pick up" the style you have and be educated towards it over time. That's the reason why - as Rick for one has experienced - with enough time and energy, getting over "basic" matches in a circuit *will* work if you stay with them in a consistent manner.

It *is* hard to explain, though, unless you've taken a class in advanced computer science.
91Posted on 10/01/04 at 00:28:12

On 09/30/04 at 23:16:38, Oliver Copp wrote:

No, that's the magic of *average*. Add up everybody's workrate and divide it by the number of wrestlers.
Oh yeah, I'm aware of that, but let me use an example.

For a match to get *****, let's say the workrate average has to be AT LEAST 70 (I'm well aware this probably isn't the actual figure, but bear with me). So if a wrestler with a workrate of 39 takes on even a wrestler with a 100 workrate, the average is still only 69.5, thus a workrate of 39 makes it impossible to have a *****. Granted, having a workrate of 39 means it'd be practically impossible to have a five star match in real life anyway, but in TNM terms at least, is this the case?
Oliver CoppPosted on 10/01/04 at 19:05:17

Absolutely.
91Posted on 10/01/04 at 19:41:42

Don't suppose you could reveal what that minimum figure actually is?
Oliver CoppPosted on 10/01/04 at 22:24:26

I can't because there *is no* set value. A neuronal network evolves and perpetually changes itself.
91Posted on 10/02/04 at 01:22:45

Is that then pertaining to what you mentioned earlier about the individuals style of booking - if someone books their cards in a particular way, then that value increases and decreases depending?

And then, if that's the case, would the value rise in a circuit that traditionally has a lot of strong matches and is jampacked full of ****+ matches while it would decrease for someone who generally has a lot of poor matches?

I know it's a lot of questions, but I like to pry. :)
Oliver CoppPosted on 10/02/04 at 10:15:32

Yes and yes :)
91Posted on 10/02/04 at 12:47:27

So that explains how I got that ***** Naked Mideon match...
Rick GarrardPosted on 10/02/04 at 15:39:27

If you ever get a ***** Jim Duggan vs Dave Sullivan match, please let me know.  ;)
91Posted on 10/02/04 at 16:13:31

My circuit isn't THAT bad...  :P
meetzorakPosted on 10/02/04 at 16:49:11

One time i had a ***** Jim Duggan vs. Mankind and a ***** Bret Hart vs. Vincent match which i watched and actually thought it wiuld have been in the * to *1/2 range.
AnubisPosted on 10/03/04 at 08:58:48

Thanks, Oliver.  I was afraid you'd say something like that, heh.  Ah well.  Very interesting stuff, though.  You have kinda answered my question now, as best as you can at least.

That explains why my people with 82-86 work rate are regularly churning out ***** matches for sure.  I have like two or three from the same two people.  It's nice because I have a rule I follow, that anyone involved in a ***** automatically gained 1 work rate, stamina, and charisma to show that they have "learned" or "improved".  Likewise, people who take injuries lose 1 work rate, and wrestlers lose 1 work rate per year after they reach age 40.

Nonetheless, I do have one more question.  Do the criteria for a ***** match change as well, or is that part set?
Oliver CoppPosted on 10/03/04 at 09:37:58

That is also adapted as you go because I think people should be rewarded for solid efforts, no matter what route they take. If they stay consistent to their own booking style, TNM will reward them.
91Posted on 10/07/04 at 22:59:33

You know, here's a thought. While we're talking about ratings at the high end of the scale, there's one thing that's always irritated me about those at the other end. Say you had the following...

Sir Mo pinned Sycho Sid after a 450 splash in 0:00:42.
Rating: *

It seems that you can have a match that goes under a minute and get as high as *, doubtless you could in theory go higher, but realistically, nobody could put out a match that good in just a few seconds. Ideally anything less than roughly a minute and a half would almost always get a DUD.
Snabbit888Posted on 10/08/04 at 00:49:23

I have another question regarding this.  When a match ends and you continue it using the 'c' button, what does that get counted as when considering match ratings?  A nearfall?
AnubisPosted on 10/08/04 at 23:16:13

Actually, I can understand perfectly why the shortest matches get *.  I think most people misinterpret just how bad DUD and negativem matches are.  DUD is a match that just makes you say "WTF was that all about?", and negative matches are matches that are actually bad.

Matches lasting under a minute, first off, don't have enough time to actually be bad.  Bad matches take time to happen, plain and simple, and usually involve putting grossly untalented workers like Big Show, A-Train, or Mark Henry in the ring for extended periods of time.  (If you think they're good, get a CAT scan and/or just substitute your favorite suckasses on that list.)

A short match, however, is usually good at getting an insta-pop.  Think Goldberg.  I don't think any of his squash matches wwould rank as a DUD.  They weren't good matches by any stretch of the imagination, but they weren't bad either.  DUD isn't the state of being neither good nor bad, DUD is actually, I believe, the first negative, or a rating of 0.  As such, since squashes can't get that bad that quickly, they always rank in at the "minimum positive", or *.  (You might think 1/4*, 1/2*, and 3/4* would be considered positive, but from what I've seen in the files, they're not.  It APPEARS that the "base score" is 1 most of the time, so a match that has nothing good nor bad would come in at *.

To back this up, I can say that I've enjoyed squash matches in TNM far more than the matches that ended up at fractional ratings.
91Posted on 10/08/04 at 23:48:27

Actually the general consensus of those who actually use ratings for real shows is that DUD is the starting point and it goes up or down from there. Obviously a ten minute DUD is a lot worse than a ten second DUD, but if someone were reviewing an actual show, they wouldn't dish out * when the participants haven't had time to actually do anything (unless you can find me any examples of anyone doing otherwise).

In such circumstances where a match is really short, the base point really should be DUD.
AnubisPosted on 10/09/04 at 11:01:10

I don't think so.  I dunno how real writers do it, but personally, I agree with TNM's method.  I think that if it's short and sweet and there's no time to screw it up, it should get a *.  That's basically the "base point", and I agree with that.

DUD makes it sound like it sucked, and I've seen plenty of squashes that were at least good for a quick pop or putting someone over.  As I said, Goldberg made it famous.  I wouldn't call any of his squashes DUD, I would put them all at *.  He's no work machine, but he doesn't suck that badly either.  In fact, there were times I wished he'd squashed his opponent that he didn't.

Of course, the one thing I do have a problem with is a WEAK squash, and TNM doesn't seem to mind.  If it's with a punch or kick, it still gets *, and that is just silly to say the least.  Finisher squash or power move squash, sure, but not punch or kick or inside cradle or whatever.
91Posted on 10/09/04 at 12:34:52

If you're happy with it because what you get is how you would personally interpret it, that's fine and dandy. As someone who's generally gone with how the ratings system was interpreted before TNM came along, I'd personally rather go like that.

Actually I wouldn't mind hearing Olivers side, is the whole * thing deliberate and what have you?
HawkPosted on 10/12/04 at 20:06:36

This thread nearly made my head explode a few times.

Am I the only one who had NO IDEA that TNM7 was this complex?
Snabbit888Posted on 10/12/04 at 20:28:25

Yes, Hawk, yes you were. :)
Rick GarrardPosted on 10/13/04 at 01:08:37

Has ANYONE ever actually ever downloaded the TNM6.22 source code and tried to use it to make their own wrestling sim yet?

I think the fact that no one has used the GPL source code to create their own sim yet speaks very much about the complexity of TNM even in it's earlier versions.  It's so complex in it's learning processes that no one has tried to compete with it.  And those that do don't have the built-in learning that TNM does.
americamamushiPosted on 10/13/04 at 09:49:05

On 10/13/04 at 01:08:37, Rick Garrard wrote:Has ANYONE ever actually ever downloaded the TNM6.22 source code and tried to use it to make their own wrestling sim yet?

I think the fact that no one has used the GPL source code to create their own sim yet speaks very much about the complexity of TNM even in it's earlier versions.  It's so complex in it's learning processes that no one has tried to compete with it.  And those that do don't have the built-in learning that TNM does.
thats why it's the wrestling sim to end all wrestling sims

(the fact that our TNMs like to screw us over so much is a testament to the fact that it is superior, because the sim knows it can mess with out heads and we'll keep coming back for more)
;)
Oliver CoppPosted on 10/14/04 at 21:00:34

On 10/08/04 at 00:49:23, Snabbit888 wrote:I have another question regarding this. When a match ends and you continue it using the 'c' button, what does that get counted as when considering match ratings? A nearfall?
Five negative near falls.
Snabbit888Posted on 10/14/04 at 21:03:54

Ah, so quite the opposite of a nearfall then. :)
Oliver CoppPosted on 10/14/04 at 21:07:18

As for the "*" vs "DUD" discussion, the system was originated by Norm Dooley way back (going on a -**** to **** scale) and later adapted by Dave Meltzer to be -***** to *****.

The base score for an average match has always been ** - nothing out of the oridinary, not bad... average.

Like Anubis said, that makes "*" below average but not atrocious. DUD ratings used to be the cutoff point from "below average" to "atrocious".

Of course, everybody can see this differently. It's just the system that I've picked up through 19 years of reading the Observer *g*
americamamushiPosted on 10/14/04 at 21:58:36

So is dud at the very end of the scale or between * and -*
Oliver CoppPosted on 10/14/04 at 22:04:30

It's right between the two.
91Posted on 10/14/04 at 23:02:18

That as may be (and I still haven't seen anyone do it like that), wouldn't it be ideal to have consistency with short matches - they generally score anywhere between -* and * depending on those involved, but how can one ten second match really be better than another ten second match?
americamamushiPosted on 10/14/04 at 23:09:51

but how can one ten second match really be better than another ten second match?
Imagine a 10 second match between Barry Horowitz and the Brooklyn Brawler.

Now imagine a quick match where someone (like the Rock for the sake of argument) just blows away his opponent, whoever it may be.

Thats how I figure it...

... or the sim just likes to mess with ya ;)
Oliver CoppPosted on 10/14/04 at 23:22:16

So what's your point? Would you like a way to influence the ratings?
Snabbit888Posted on 10/14/04 at 23:32:06

I'd vote a big no for being able to directly influence the ratings.  TNM is great because you can pick who wrestles, what kind of match, how long, who will win, etc., just like a real booker, but just like a real booker, the wrestlers have to execute.  The wrestlers still have to wrestle the match and make it happen.  I implore you to leave ratings alone and let the wrestlers decide. :)
americamamushiPosted on 10/15/04 at 17:17:00

Maybe some people would like an option that would stiffle ratings a twinge.  Like a flag that would always make a make under a minute or whatever a -*

... not me.  but some people
AnubisPosted on 10/16/04 at 04:54:15

I think people should just get used to the ratings as the TNM method.  It's like rating movies.  Some do "thumbs up" and "thumbs down", some give star ratings, some give number ratings, etc.

Some wresting critics may go by one star system, but TNM's star system is either the "Observer system" or the "TNM system".
zackarcherPosted on 10/16/04 at 16:14:41

The "thumbs up/down" thing could make a fun plugin.   Like the roman days, we could have the owner of the fed give his opinion of the match.  Thumbs up could be a title shot.  Thumbs down and the wrestlers involved are instantly slain.  ....or fired, whatever.
americamamushiPosted on 10/16/04 at 18:13:44

...of course there'd have to be a flag to make it historically acturate so thumbs up meant die and thumbs down meant live ;)
Oliver CoppPosted on 10/16/04 at 18:54:07

Which, in turn, would be a perfect tie-in for the proposed feature to permanently stop a wrestler from being used ;)