VFDC Ranking Project?

Hey guys, I was responding to a thread in the SC forum (specifically, this thread) and came across this "ranking project" that the guys at SC.com are working on (linked here).

It's basically a theoretical quantification of match-ups between characters, with the basic underlying assumption that characters who have the most advantage overall in match-ups can be considered as "stronger."

Traditionally, VF ranking tables are determined by actual, live combat between highly ranked players in a multiple, round robin fashion. Wins and losses are tallied to determine rankings.

My question is...is the theoretical approach used by SC.com possible for VF? I think it might be too hard, but thought this might be a good point of discussion anyway.

 

I don't think it's worth trying - if I can assume we're talking about actually performing an experiment with VFDC members...stats from such a small user base are less meaningful.

Even if we got 50 "good" english speaking players in a room together, and then get them to play a massive round robin...the gap between the best of the 50 and the worst of the 50 would be huge, and would skew the stats.

Somewhere, somehow, this is all on vf.net in japan, people have posted before e.g. who loses to which characters the most. Those aren't even fair indications of who is strongest though, chibita just happens to take nosedives when playing e.g. vanessa more than anyone else, but I don't know that vanessa is tops because of that. Also chibita is going to do fantastically well in such a round robin that it would seem to indicate that lion is the best.

---
The creed "solve every problem by watching movies solution" -
I could maybe take a look at the win% of every video-captured player I have, notate who they're playing, and then base some kind of quick judgment on that - i.e. "the highest akira rating ever was 91%, the highest pai rating ever was 87%" blah blah... but I think that's more an indication of who's popular and who gets recorded.

 

I think people have thought about doing a purely numerical analysis about each character... Probably have to dig up the post though. I know me, llanfair, and the goldens had talked about it in the thread.. You can probably see if that's something worthwhile..

/versus/images/graemlins/laugh.gif

 

Yeah, I'd even started a little something. Assign an arbitrary weight to attack speed, rec, damage, frame advantage, hit level, circularity, etc...
The hard part is calculating damage potential. I'd just lump direct combo damage under damage, and use some multiplier to determine damage for combo starters that need mC or MC.

Lastly, how does one scale things so that number of attacks in your movelist doesn't affect things much - otherwise vanessa is #1 and goh is last.

 

I think CreeD is right. For the most part, there's always a kick-ass player for each character, but those guys would have their weaknesses, which would reflect on everybody.

Also, the argument about Chibita making Lion look like the strongest character holds true. Something similar could be said about Akira, Jacky, and Lau, b/c of the sheer numbers of strong players w/ these characters, while Brad, Shun, and to an extent Aoi would look weaker, due to less representation.

However, there is info on VF.NET, although I'm not sure if its' what you'd want. They have the "records" section, which has the Top 10 highest wins, losses, and matches... no ratios if I remember correctly though.

 

I wouldn't mind seeing the "which famous players beats/loses to X character the most".

really, the point of these exercises isn't to establish who seems to be best (we all can kinda sense that or have similar top 5 lists) ...it's to see who's underrated. If we see that 7/10 badass japanese players win 20 percent less vs vanessa than they do against others, we can infer something useful.

 

Hang on, I think you guys misunderstood me.

SC.com's approach is purely theoretical. I.e. they ask themselves, "How much more advantage does Taki have over Astaroth? Is it 5, 6, 7?" They assign scores for all the match ups and tally totals to decide rankings.

I'm asking if that approach is possible for VF. I.e. do we know enough to say Shun has X number advantage in a match up with Jeffry?

 

I'm asking if that approach is possible for VF. I.e. do we know enough to say Shun has X number advantage in a match up with Jeffry?

No, cuz for the rarer matchups, the odds that we've seen them are low and we can't even make an educated guess as to even which character owns the other... much less by how many games. I have no idea if I'd say for sure that jeff owns shun or vice versa. I have exactly 4 shun vs jeff clips.

I have 59 Kage vs Akira. /versus/images/graemlins/blush.gif

 

Full agreement with everything you wrote there, Myke. I did think that despite WCMaxi's assumptions/perceptions about VF he has projected VF positively, which is nice.

One way I thought this thread could be useful would be if someone chipped in with, say, "Lau will beat an equally skilled Goh 7/10 times, and here are the character specific reasons why." Then, someone can dispute that hypothesis with other reasons.

By way of example, I learned quite a few tidbits of information on SC2 from the arguments in the ranking project thread there.

 

Sups,

Well I'm not going to claim I know everything in VF4:e, however as someone with scripting experience and someone that's studied it's depth, I do understand how it operates.

I'm by no means good yet, but at least intermediate. (I still suck ass at break throws etc).

At any rate, I think you all have some mis-nderstandings on the point of an RP. First, in Japan they don't base them on RR results, it's based on theory assuming the players are both great and completely even. Simple as that. Since humans are the ones that debate the points, it will always be bias to a region. When there were more popular in Japan you'd see specific locations doing their own, mags doing their own, etc. That's why I posted the link to the Challenger RP. Naturally, the more humans that debate the points the more bias you can work out. That's why I make it an open debate on SC.com rather than a closed one with me (someone who knows a lot more about that game's system than a lot of the other debators) making all the rules.

The point is as Ice-9 noticed it that is incites dialog. VF players (in the US) are very hard to talk to... and I've noticed a lot of the drama you guys get here, no sense in getting into that, but it's the closed minded nature of the old guard here. None-to-friendly. Anyway, as I was saying, there is no perfect fighting game, and the only possible way to balance one is to make every character identical. Now, you can say it is "well balanced", but never "perfectly balanced"; that's just unrealistic and ultimately counter productive.

By talking about match-ups we learn more and sometimes counter points are made that reverse people's views on a specific match. People love to argue and this is a productive way to do so.

In the case of Goh/Lau I don't know them well enough to debate, I just have heard many times in Japan that one is bad. Goh has problems using sabaki on some of Lau's heavy shit or something... I dunno.

But as an example, I think it'd be easy to debate Aoi v Sara (I use Sara as main, Aoi as sub-main). If you think about it's a huge bulk of Sara's moves and specifically annoying FL attacks fall to YY. Granted, there's a lot more to that fight etc, but you're not doing a project so I'm not going to waste my time.

Point being, they incite discussion which leads to more people understanding what's going on in the game.

 

Ah for the record...

I used to be hardcore in VF2, hated 3 (*shrug*), tried 4 but sucked ass at it (didn't really try), and now play a ton of Evo.

For SC... I loved SCI, and don't like SCII. SCII is and will be a lot bigger than VF, that's just how it is. One thing you guys can do to change that is change the way you look and new people. It's better to ignore people with "wrong opinions" than to go out of your way to strike them down. Just really unfriendly here and I'm sure I'll eat a lot of shit for saying so... but then again, this ain't my forum, someone just asked me to explain RPs in this thread.

 

Thanks for taking a second to throw us some info on how the ranking chart was worked out... I'd thought it was round robin because one of the japanese guides had made the exact same sort of chart and had based it on actual RR results.

Doing it by just posting opinion is pretty much like any other tiering thread, but I guess the main difference is people can get into specific matchups. We can all probably guess lau is better than goh, and saying so isn't useful, but citing specifics of a lau vs goh matchup (i.e. goh can sabaki the second punch of many canned combos etc) will help players for both characters.

Anyway, I wouldn't object to seeing this sort of thing put up, but who has the nerve to make the first post? I think only 1 or 2 people in NA are good enough at the game that I'd trust them with posting the matchups, and even those people have much less experience than the japanese. A 90% usa player with 1000 games doing the original would be nice but I'd always think "this should be a 90% japanese players with 10,000 matches".

re: Community unfriendliness - I dunno what it is, we get that same comment a lot on the board. It's probably true but I can't figure out why. Maybe because the VFDC crowd is older? I dunno... I'm working on my own habits anyway /versus/images/graemlins/grin.gif

 

I've never seen charts based on RRs. I have in actual game centers over there, but not for actual ranking.

Well if you guys do make one I'd read it. There's a lot I want to know on how to deal with certain match-ups. Aoi used to be my main... but yeah... she just has too much trouble. Using Sara makes it obvious for me now.

 

A bit OT but I wanted to point out that the reason you're most likely having difficulty with Aoi is that she requires the player to not only have a solid grasp of VF concepts but being to apply them all too. She does her best at disadvantage (-3 to -6) so being able to do the defensive techs plus using her special moves are extremely important. Also, you need to know the special properties of her moves to be effective. Things like how to escape all throws from YY stance using 1 command and doing pounce after elbow reversal are key points. Regarding bad matchups, her only really difficult battle is with Akira due to and but she can still scare him by crumpling many of his common attacks such as and

 

Naw, I know all that. Though you have to admit her being at disadvantage on hit all the time is really shitty for her. Forcing someone into an exact style isn't wise in developing a game. Other chara in VF are a lot more free. Aoi has to reverse and YY. I still love her style and will use her, but I'm not going to pretend that it's not an up hill battle, doing so means I don't realistically approach the matches.

You should come fight Shota again so I can get in / watch.

 

Aoi's definitely not a scrub character. She takes a bit of getting used to, but the pay-off is good. The YY and reversals are not necessary to win, but they do help if you know when to use them.

I've seen loads of matches with no reversals, and maybe on YY, and Aoi can still own them.

 

[ QUOTE ]
Shou said:
Things like how to escape all throws from YY stance using 1 command

[/ QUOTE ]

Sorry to continue OT, but what are you referring to? I was under the impression that the only options from YY were to 1. drop the stance, the recovery of which can't be cancelled; or 2. dodge, which is cancellable only by P, which comes out too slow to stop throws.

 

I think Shou means, though I'm not certain, that you only need to input one throw escape if you're dodging in YY stance and are thrown. Since, while dodging, they can only perfom a side throw on you, the only input needed to escape is or and + depending on which way you dodge.

 

So you're suggesting attack, +++,,+? Doing this in training mode, with the CPU set to respond with a throw, results in a side throw animation, but no escape. In order to get the side throw escape, you still have to input the displayed command direction for the throw. Maybe a training mode bug, or my fingers are too slow.