*itch you guessing! . . .woo! . . . you was m'fckin right!

I’m not actually basing my points on opinion, but on scientific research into the limits of human reaction time, this is the key point that you’re not getting. You probably are very good at reacting but even you have your limits, and the speed of most manuals surpasses them.

…mmmhhhh… I don’t know what the actual definition of guess breaking is but I’m certain of one thing:

I’m terrible at Combo breaking.

Some characters are easier then others to break In my experinence. But All in all I don’t think I want to learn how to distinguish sadira’s light punch and medium kick.

However,

If they did change the breaking system drasticlly, I would not like it. We don’t need anymore blockout symbols.

If people are guess breaking too much, bloody deal with it. I don’t think you can change it.

I comprehend your “logic and theoretic scientific explanations” just fine. While they may be true, they don’t set enough precedence in the realm of KI’s gameplay to be relevant. I noted my limits, so mentioning it again is kind of redundant. As I said. . We’ll agree to disagree.

You’re being deliberately obtuse at this point.

I’m not the one talking about scientific research in conjunction with Combo breakers. . . lol

Yeah, why let facts get in the way of good hyperbole?

I think you mean it the other way around?

http://www.humanbenchmark.com/tests/reactiontime

Unless you’re consistently hitting around 150 ms on this test - No, it’s not opinion.

http://ki.infil.net/reaction.html

2 Likes

If you don’t think the disadvantages and massive penalties that naturally come from guess breaking recklessly aren’t high enough penalties already, then you aren’t good enough at this game to make any sort of balance suggestion.

Ah nice touch. Is it possible that you would also be able to find the KI requirement of a 150ms break reaction to a manual or light auto-double? They have that stuff in the master mode of the game right. I can’t recall if it shows those kind of intricate details or not.

Yet another audible affirmation. . lol but I’ll take it. This isn’t a balance issue, nor is it a headlining issue thats setting precedence over most of the major issues in the game. However, (Generally speaking) to say someone isn’t good enough based on their opinion from experience, is rather boisterous & ignorant.

Read the page I linked.

Manuals have a startup of 5-9 frames plus maybe 3 frames of hitstop. Light autodoubles are around 14 frames. Then consider that in place of just clicking a binary response, you have to identify the strength of the manual, move your fingers to those buttons, press them, and have the signal delayed by the TV and online input delay.

Experimental data on reaction timing doesn’t suddenly become irrelevant because we’re talking about a video game.

1 Like

Quickly working it out, 0.15÷(1÷60)=9 frames, which is how long you have to react to a light manual at best, since it has 4 frames of startup, then hits on the 5th and the break windows starts, then you have 4 frames of additional hitstop in which to break.

Pretty much what others have said. 150ms works out to 9 frames. Bearing in mind that 150ms is very fast for a consistent human reaction time, it simply isn’t possible to break something like a light manual off pure reaction. Especially since it’s not like the first few frames of the attack are particularly notable or spectacular enough that you’re going to begin reacting at frame zero.

@thekeits has been pretty assiduous in joining these reaction vs read discussions in the past. As memory serves, the rough point is that something like a light linker or even light auto is simply too fast to break purely on reaction. Fortunately, humans are very adept at pattern recognition and “guesstimation”, able to take small sample sizes and make what are often pretty accurate predictions into the near-future. Once you start “looking” for certain things to break within certain windows, when you see them and respond it feels like you are reliably reacting, but really you are mostly predicting, and it is only after the fact that this prediction is vindicated or not.

There’s nothing wrong with this - it’s just how you and all the rest of us are wired. But the others are correct when pointing out that manuals and light linkers are largely unreactable. You have to know (guess) they’re coming, and then respond. Just because you can “see” a light or medium manual doesn’t mean you can react it.

In order for you to truly react to something, I have to be able to do it to you in training mode 100 times, and you successfully break it a high percentage of the time (95% or higher). This is definitely impossible for any manual that isn’t a two-hit normal.

You can even test this yourself if you go to breaker training, turn on only heavy linkers and manuals (turn off everything else, including auto-doubles). See if you can successfully break manuals better than 50% of the time over a large sample set. You won’t be able to.

It’s not enough to be able to break it correctly in a match because you were pretty sure it was coming, and you “saw something” and then you did it. This is largely based off prediction (which is fine, because without prediction you would never break anything in KI, it’s just not reaction). Would you have been able to stop yourself and not break if that thing you saw wasn’t what you were expecting (or better yet, change to the correct break strength)? Almost certainly not, but it still feels like you reacted.

But yeah, as Storm said, people are generally wired to remember their successes and ignore their failures. If you successfully break a manual, you’ll remember that, but you’ll forget the times you locked out on heavy auto-doubles because you flinched with a bad reaction (which happens even to the best players).

Also worth noting that bastfree did Jago’s frame data using the in-game frame tool and found out that the break window for his light auto-doubles lasts about 24 frames after startup, which is somewhat distinguishable from the other doubles and then reactable, a small percentage of the time, by amazing players. This was a much higher number than I thought, and I imagine it’s not this high for every character’s light auto-doubles, but it’s worth knowing the number. Manuals are still definitely out of the question, though. We’re talking somewhere between 9 and 15 frames to react to those.

1 Like

This got way deeper than I imagined lol. I understand the point thats being made here. In all honesty I don’t think that I could do it 100 times in a row in training mode. Even with the frame facts in place it’s still probable that being able visualize & react to something at these speeds may not be able to maintain it’s consistency because of how we work as humans, but it is indeed possible without guessing for (not all) but some people. As you stated we are all wired a bit differently. I personally have no issue distinguishing Jago’s light punch manual at all. If I lock out on this command, it’s because I put in the input a tad too late, not because I was guessing and hit the wrong buttons. By no stretch of the imagination am I saying I’m super human either lol, so please allow me to reiterate my position on this particular matter a moment.

I feel that certain characters in the roster telegraph in ways that are more noticeable than other characters. Whether their start up frames be super fast or not; from a visual/physical consensus;(from my eyes to the button press) a lot of characters in the game telegraph fairy easily and the manuals aren’t that hard to distinguish “for me”. In the beginning before I started playing on a competitive level I used to guess regularly, & then I had to question myself for al of the L’s I started to hold because of it. This is just my personal experience. Through training I’ve found that I’ve become a lot more aware of manuals and their appearance or animation. I also use a rhythmic approach for breaking as I’m a percussionist by gift & trade. The timing for heavy, medium, and light manuals are all different. I use my personal attributes for breaking & it works for me 75% of the time. I don’t want to give the notion that I am breaking everything I see with God like abilities from planet namek lol . . thats not the case at all. My approach to breaking is mysterious in it’s own way. I can only contest that I am in fact not guessing, nor am I breaking everything thats thrown at me either.

Please don’t use a color-change reaction time test to get the real story here. The fastest a human can reasonably be expected to ACTUALLY react consistently to visual stimuli is around 13-14 frames. But this is a fighting game, it isnt a screen changing from red to green. I bet you get hit by Jago’s 19 frame overhead now and then, don’t you? But how can this be if I’m in peak condition and can react in 13-14 frames?!

Well, how many frames into an animation is it before you have any reasonable ability to even discern what it is?

Take Spinal for example. The startup of LK and MK are pretty similar. The startup of HK is identical to his run. The startup of LP looks similar to MP and HP, which are actually identical on the 1st frame and for quite a few frames after. Bottom line, you aren’t reacting to a move’s startup animation on the 1st frame almost ever.

Every time I’ve met a player who says they can react to manuals and break them, I present the science (as others in this thread have). When and if that fails, we take the test.

The test: Go into Combo Breaker Training mode. Set the dummy to do no auto doubles, only heavy linkers, and any manual. This will result in them performing a fully random 1 in 3 chance manual after their heavy linker. Now break these on reaction and keep score. Do a large sample size, like 100. You are extremely unlikely to score over 40%, even if the dummy you chose has heavies on the slower end that are border-line reactable. Odds are you’ll score around 33%, which is exactly what you’d get in a large sample by randomly guessing.

In this test, you have no data about your opponent’s habits or skill level to make reads off of. This is a CPU perfectly doing a fully random manual after a heavy linker.

Record yourself trying this and post it up unedited if you think you can beat science. :smile:

(And to correct my own bad math from last year, most light doubles clock in around ~25 frames, which is actually reactable but because of the reasons above, is still extremely difficult)

3 Likes

Instead of trying to call for change, why not check here for some advice where you can take action and curb guess-breakers yourself in the game?

There’s some advice you can use here to deal with so-called “guess breakers”

We’re way out in the water with this whole science thing. I get it & I can appreciate it. However, this topic runs perpendicular to the agenda of the original post. This is more so about breaking in general & I am basing my skill and experience on breaking in general. I’m not sure how we got to manuals only here. I’m simply saying that IG has done very well to insure that each character telegraphs in there own unique way (Except for a few) & that itself negates the need to guess. I can deal with guess breakers all day np. I do it everyday lol. However I just would like to see a bit more competitive “intentional” gameplay rather than a lot of accidental luck. Btw, I only break manuals about 45% of the time. I never really go for them. Too risky with timing even if I can see them or not lol.

Looks like a good read. I appreciate the link. I’ll definitely check it out. I’m able to deal with guessers no doubt. Lol I just counter break on linkers lol.