Ok, thanks, now I understand. I dont like HDR Chun too, specially cause of the accidental air SBK that often comes out when we try to do j.RH as an AA from crouching. The others changes I can live with.
About Blanka balls, they are unsafe on block (or hit) against Vega, Balrog, Dhalsim, Fei Long, Blanka and Dictator. Against everyone else, they are safe.
Yeah, I thought of that as I wrote one of my responses, but I neglected to mention it. Great point.
Right, those two are probably the fastest dashing attacks in the game (fierce version for Dictator). Not sure if other stuff hits a blocked beast roll, but those two definitely do the trick.
this thread gives me a headache
I do think HDR is slighlty more balanced and overall the better game, but there were just certain changes that were completly unwarrented and quite frankly retarded. I also find it funny that I have no issues doing a meaty jap into 720 with gief in turbo yet I can’t pull it off in HDR…same with a meaty c.FP with fei into a rekka.
Right, true. My point is that whether you can hit Blanka back after the Blanka Ball is character dependent, NOT distance dependent. If Balrog or Fei can hit the Blanka Ball back, they can ALWAYS hit it back, no matter what. If Chun Li cannot hit it back, she can NEVER hit is back, regardless of distance.
Of course, there are some VERY specific situations where you CAN hit the Blanka Ball back. If you do it too Meaty against a rising opponent, you can be hit back. If you do it while deep in the corner, you can be hit back.
Again, the main point is that the results of a Blanka Ball, Balrog charge, or Honda Headbutt are not reliant on distances with which you use it. As long as you reach the opponent, it’s safe. Cammy, amongst these four characters, is the only one where there is a qualified “too close” distance to use the Drill.
I’m unsure if i can agree that a move that will always be hit back against part of the cast is more safe than a move that will not be hit back if done correctly.
I won’t take it any further.
I honestly don’t know how I would rank one against the other in terms of which is generally"more safe," but I’m not too concerned with that sort of comparison. In practice, moves that are safe against part of the cast are a little easier to use safely than a move like the Drill. If you are up against a character that the Blanka Ball is safe against, there is no concern at all about it during the fight–the safety is matchup dependent, and you’re set, no more effort or concentration needs to be put toward that aspect of it. Same if you’re up against someone who can consistently punish it; you know every single time you use the attack wheher it is safe or unsafe. There is no gray area.
Meanwhile something like the Drill requires constant awareness of your spacing when using it. It becomes something you get a feel for in time, but it is still something that needs your attention.
(excuse slight derailing of thread here)
I agree that it can’t be done from tournament level events. Too few events, too few matches. However, it most certainly can be done by looking at ranked matches online. There’s a big database somewhere with the results of every ranked match ever, millions of matches.
My vision is to get access to that database, and then there’s a lot of really interesting data we can look at, for example:
Many other interesting things can be done with this data. I don’t know, I’ll see if I can poke around, but I think the statistical answer to all of these questions lies in the data of ranked mode.
I also think that if this data cannot be obtained from live, it might be possible to work with 2DF or GGPO to implement such data. But I think the best source would be XBL, because of sheer numbers.
Isn’t ranked mode busted? Will that not give a false results?
PS. Damn good Chun!
I think it’s VERY important to consider that these character-as-projectile moves all have very different properties:
To single out one aspect of a move as being THE aspect that makes a move great or not speaks to a tuning issue, NOT to the actual mechanic itself.
Almost any move can be made godlike or garbage with tuning: hit box, recovery, range, speed, charge, damage, comboable, etc.
In addition, many characters are designed with a certain feel/playstyle in mind.
And simply “powering up” (or nerfing) with tuning one of that characters moves can often result in that character then being played totally different.
I’m with you Zass.
The data can give you SO much utility.
But unless there is a top dog in the process that shares this perspective, I don’t see it getting any traction anytime soon.
I worked briefly for a gaming company, and the culture there didn’t have the understanding, affinity, or fervor to see data as a valid indicator and tool to craft with.
A lot of the thinking was strictly “copy what style appears to be working in other successful games” or rely on individuals as artists to feel out what works best.
The company I was at was not configured to capture, sort, and record most of the data it generated.
And some bits that were tracked and passed on, were met by the designers (not sure about the other departments) largely with dismissal, failure to connect, or were ignored.
At one point I even put together a spreadsheet indicating user trends and game data on how/where players got equipment, and how we could materially modify that delivery to get users into certain areas of the game where they weren’t going.
Despite very positive feedback from my immediate peers, I couldn’t even get a chat with the key person in charge of that area to present it.
It’s just not in the culture of game developers yet.
Gathering online data unfortunately suffers from the exact same issue as from offline matches: there aren’t enough ranked matches with top players who use more uncommon characters.
Plus, it’s sort of ridiculous to use online matches. How does that relate to HDR offline? First off, there’s the matter of lag. HDR lags enough that certain frame-dependent tactics are no longer as viable as they should be in real life; meanwhile, other tactics that depend on fast reaction time are enhanced online. As a whole, players with worse reaction timing that may not be able to respond within the shorter threshold are hurt more by lag. GGPO has much better lag handling and even then, I’ve seen firsthand that the lag there differs from offline play enough that any accurate rankings can’t be made.
Second, using online ratings just doesn’t make sense. With all the disconnects (some of which cause the dropped player to take the loss instead) and matches against players who are halfway across the world, the more a player gets to the top, the more it’s a luck factor in getting fair matches in. You could spend 2 hours getting a couple of points total against beginners. Then out comes somebody with an insanely high ping, does nothing but “teleport” jump in tick throws all day, and you’re down 50 points (from personal experience).
The other reason the online ratings don’t work is because there’s almost no incentive to keep playing so players at the top play very little and many of the very best offline ST/HD players never really played ranked in the first place. And maybe if the top US scoreboard players had the best tourney results we could reconsider, but that’s not the case (although they’re obviously all great players).
So really, as nh2 said, there’s no way to get a meaningful scientifically-driven character ranking list via any means.
Jab headbutt and fake fireball didnt change anything? You got to be kidding me.
Arcade ST for sure.
There’s tons of changes in HDR where it feels horribly unbalanced and not ST feeling, I wish I could expand on this more, but it’s a feeling you get when you go playing ST for an hour straight then HDR for 15 minutes. HDR would have been really great if they kept the “dice-based” factors in the game (such as dizzies, damage, etc.) The changes look great on paper, but when you’re actually playing, none of the balances “fix” ST (notice how I put quotation marks on fix.)
For example, O.Hawk vs Guile, this is one of my most hated matchups in vanilla ST. Guile has so many tools versus O.Hawk to keep him out EASILY, but if Guile messes up and eats a dive, it helps “reset” the match, partially in Hawk’s favor due to the knockdown, but when we are in HDR, the same matchup, there is no knockdown on dive to help “reset” the match like in ST, and Guile will forever be in the lead in that matchup (because he’s standing and can fb/aa easily.) On top of that Guile has “rainbow” style flash kick where if Hawk even thinks about jumping he gets kicked in the face (and knocked down.)
But of course, there are some characters who were treated very nicely, such as Ken, having O.Ken’s fireball, and DP invincibility, while keeping the things N.Ken had, and adding reversal super helps his matchups a lot, this is a buff that’s actually sensible, and good, but while Bison (who’s already extremely good in Vanilla ST) gets these random buffs that he doesn’t really need (come on fake slide/better st.jab, etc) It’s hard to do a comparison with this because I hardly know anything about ST that isn’t O.Hawk.
HD Remix sounded good (I was totally hyped up about this game for the longest time, now not so much anymore), but in reality it’s just good on paper, and nothing else. I just wish I knew more frame data/how the game works to make a better write up for this. Personally, I think I could even make a better “remix” of ST, and I’ve only been playing ST for 2-3 years. Hawk/Fei/Cammy/Blanka still suck in top play, and characters who didn’t need extreme nerfs got nerfed hard as fuck, while character who don’t really need extreme buffs got buffed for no good reason at all.
YO REAL TALK!
To HDR is the definitive version of sf2 to me. I didn’t play ST and I can understand where the OG’s are coming from because the game has a special place for them… But, if it wasn’t for HDR I would probably done with sf2 as a whole. I read all the changes to the game and it made sense to me. And most important, I just find the game fun.
I think it’s a matter of perspective.
If you are looking for the data to give you a specific answer to an ideal question, then the data is going to have a tough time trying to fit that bill.
Instead, look for the data to give you a more accurate picture on the actual topography of the game.
As Zass pointed out, he would be looking at tiering within subsets: low/medium/high level players.
Not what character “is the best under top tournament play” (he would need top tournament data for that), but rather how the online scene breaks down.
You can assess what players perfer in terms of format:
Yes, you can break down charts on which characters are the most effective using online data.
But you have to sort that data by player skill (or other criteria) to get a more specific picture.
(and even then, you are not working with top tournament data here)
To get top tournament character charts, you have to either use tournament data or “peer review”. (peer review = players use their opinions to populate the chart)
What I’m saying is that any match that occurs online already falls under a taint of un-usability. If a player can’t effectively utilize his optimal, lag-sensitive strategy that works in an offline matchup, then how can any online data be trusted to avoid the same problems? And yes, I’ve experience this firsthand, having played the same matchup against the same good player, both using the exact same tactics, and losing bad in an online tourney but winning easily in an offline tourney within the span of 2 weeks. I simply couldn’t react and time my input-strict tactic well enough online.
And how do you determine top players when some of the undisputed offline top players of the US are sitting at something like 1/0 right now low on the scoreboard? Right now, if all the data were available, it would only be able to answer zass’s questions of how does a 4800+ X do against a 4800+ Y. And even then, the answer would be weak because many matchups still won’t have sufficient data. But how is any of this meaningful in the first place when nobody cares about tiers specifically suited for online play? I’m assuming you think online approximates offline but since I’ve seen notable different results between the two, I don’t think the connection is close enough that a character ranking would be accurate based on these results.
And peer review doesn’t follow hard stats, which is why everyone has been careful to specifically refer to the scientifically/statistically-driven ranking being impossible, not the conceptual ranking. However, theorizing the character ranking among top players is pretty much the only accurate ranking tool we have.
Hey Zass what’s going on with NW ST guys for the 5on and 3on events? Maybe you could post in one/both of those threads let people know?
Ehhhh I’m not sure that the one indicates the other. Guile is popular (at various skill levels including up high) well beyond his ease of use. Sagat otoh easy to use yet unpopular. Some characters are simply more interesting than others yet it’s not something people would agree on in a forum discussion.
This one I’m more on board with. To some extent this kind of analysis is possible with SF4’s arcade ranking system in japan and it’s pretty fascinating. It’s not as specific as what you’ve got there but, like, if you look at the top 100 say there might be yaaay amount of Bisons in the 100th-50th place but less than half that amount in the 49th-1st place perhaps at the very highest level Bisons tricks stop working. (or could be people who use bison tend to not give a shit about BPs)
Damdai used to have a matchup chart built into the 2df website I dunno there was some kind of problem. Even before he got rid of it the numbers were mostly gibberish if he wails on Honda players with Fei Long all day its going to look goofy. Too bad remix didn’t have more sophisticated stats though.
BTW that earlier post saying “no way jose” to zaspacer is part of this kind of thesis I’m thinking up about how elusive a proper tier list really is like even god doesn’t know what the tier list is I dunno I’ll talk about it some other time. I mean well Ganelon pretty much nails it here:
Yeah simply asking Mike Ross about SF4 Honda’s matches is always going to be better than counting up his wins (tourney or otherwise.)
You’ve got to keep two sets of data, one for offline play and one for online play. (I figure someone has already pointed this out, yeah?)
If you really want to get anal about it, you keep multiple sets of data, grouped by connection quality.
Net lag has such a significant effect on the game that I don’t think that there’s any way around this.
Whoa, don’t plum your text, man, some of us don’t have the same background color as you.
RH FK smells like ass, man.
I never even read anything that explained why Sirlin thought that Guile needed that. Far as I know, he made that change simply because he felt that the RH FK was redundant or useless.
Chun-Li’s offensive crouch + forward flip kick was also useless (according to Sirlin, anyway), so why didn’t he make that more powerful, too? (Woulda been nice if he’d done something with it, to offset all the nerfs with which he burdened her.)
I appreciate his sentiment in wanting to give the RH FK something to do, but I still don’t understand why Guile needed a flash kick that hits opponents who jump, dive, or drill in from far away. Were these an unfair threat to him? (I’d like to hear from seasoned Guile players.)
EDIT: I re-read Sirlin’s “Guile changes” page, he justifies the new FK as protection vs. Claw and Dhalsim. Good deal, thumbs up – those fights were definitely a b!tch for Guile.