System Frame Data Explained

[Sly Bass]

OOOOOHHH now I get it !!! If its part of a string theres no recovery on the 1st Punch...... can you use that plus the advantage you get on normal hit to calculate the duration of the opponents hit stun ?....... and from there do you just need another hit to become active before the end of that duration ?

The move details should tell you your advantage after each move you've hit them with.

 

[iHajinShinobi]

This thread should be stickied.

@CyberEvil @Brute @Mr. Wah

 

[Lulu]

The move details should tell you your advantage after each move you've hit them with.

Theorotically speaking..... advantage is calculated by Taking The Hit/Block Stun duration (is that even a static a value?) and subtracting the recovery of the attack from that duration..... right ?
In other words if a punch is 15 (2)18 and its +2 On hit.... does that mean the opponent you were hitting was in a 20~21 Frame Hit Stun ?

 

[XZero264]

Keep in mind that string strikes startups can vary depending on how long into it's delayable window you input the next command. On page 3 the area marked "Followup Interval" tells you the possible points during the current strike you can input an appropriate command.

For example (gonna have to jump to Leon for this, sorry) Leon's and frames at fastest inputs are as follows: 14(2)22 and 11(2)22 respectively. If you look at page 3 in the Interval Followup you notice when you input you get the numbers 16~38F. This is the input window you have to put in a followup command (In this case ). If you input the second immediately you will followup on frame 16 of the strike and get the normal followup frame data as 11(2)22. However if you delay the strike input between the window (16-38) you will add the delay window to the followup strike (for example with a delay I can get to be 25(2)22 or 32(2)22.
On page 3 you will also see "Followup Frame" this is the frame during the delay in which the command was input, it will tell you by giving you a number.

 

[Lulu]

Keep in mind that string strikes startups can vary depending on how long into it's delayable window you input the next command. On page 3 the area marked "Followup Interval" tells you the possible points during the current strike you can input an appropriate command.

For example (gonna have to jump to Leon for this, sorry) Leon's and frames at fastest inputs are as follows: 14(2)22 and 11(2)22 respectively. If you look at page 3 in the Interval Followup you notice when you input you get the numbers 16~38F. This is the input window you have to put in a followup command (In this case ). If you input the second immediately you will followup on frame 16 of the strike and get the normal followup frame data as 11(2)22. However if you delay the strike input between the window (16-38) you will add the delay window to the followup strike (for example with a delay I can get to be 25(2)22 or 32(2)22.
On page 3 you will also see "Followup Frame" this is the frame during the delay in which the command was input, it will tell you by giving you a number.

That explains the dropped Juggles....

 

[jjinkou2]

@Sly Bass i looked at akira frame data xls sheet, the active column greatly simplify this.
but may be i've missed something or may be i've found a typo in this sheet.

for akira
16(2)24
12(2)22

So the addition should be 19 (last real active frame of ) + 12 (start of ) + 1 (phantom) = 32 and the active frames should be 33 and 34. But in your column E in front of it's written 27-28.

Why's that?

note: for 6K and 6KP the active frames are correct though.

 

[Sly Bass]

@Sly Bass i looked at akira frame data xls sheet, the active column greatly simplify this.
but may be i've missed something or may be i've found a typo in this sheet.

for akira
16(2)24
12(2)22

So the addition should be 19 (last real active frame of ) + 12 (start of ) + 1 (phantom) = 32 and the active frames should be 33 and 34. But in your column E in front of it's written 27-28.

Why's that?

note: for 6K and 6KP the active frames are correct though.

Those are two separate moves you have listed there. If it was a string, you would be correct.

 

[jjinkou2]

Ok i understand where my error is. The excel file doesn't give a special notation for . it's written 6P and 6PP that i converted into and . (they are instead and )

I checked with the info window and with your formula, i have the same result 27-28.

Ok now i can do the same for my excel file on kokoro.

 

[Sly Bass]

Ok i understand where my error is. The excel file doesn't give a special notation for . it's written 6P and 6PP that i converted into and . (they are instead and )

I checked with the info window and with your formula, i have the same result 27-28.

Ok now i can do the same for my excel file on kokoro.

Yep. Makes finding unholdables easier.

 

[jjinkou2]

Kokoro's wiki frame data is updated.

Frame data

 

[jjinkou2]

How did you have found the correct distance in the column "Reach" ?

For example is 2.38m in your file , but the info window gives 1.49m. I'm sure yours is more correct and i would like to change kokoro"s file too.

Obviously it's not the sum of (1.88) + (1.49) provided by the info window.

 

[Sly Bass]

How did you have found the correct distance in the column "Reach" ?

For example is 2.38m in your file , but the info window gives 1.49m. I'm sure yours is more correct and i would like to change kokoro"s file too.

Obviously it's not the sum of (1.88) + (1.49) provided by the info window.

In string, add up the reach of each move.

 

[jjinkou2]

it's not the sum of (1.88) + (1.49) provided by the info window.

that gives 3.37 and not 2.38 in your xls file.

it's neither the sum of (1.88) and (1.29 in the info window but in your file it's 1.09) that gives 3.17.

 

[XZero264]

Probably takes into account forward motion. Leon's PL range is up to 4 meters because he takes a huge step forward despite the strike itself being about 1.5m.

 

[Lulu]

Okay so since I've changed characters I now have a use for Page 3 of the move details.... but I don't understand how to read and interpret it properly just yet.....

So anyway.... a Regular Side Step is evasive for 20 Frames and leaves you vulnerable for 3~5 Frames..... right ?

How soon into the Side Step can you cancel into an attack.... I tried using the start up details of a side step attack to determine when the side step ends and the attack begins but unfortunately it seems included in the start up of the attack is the Side Step Duration aswell....

So does anybody know when the Side Step ends and The Attack Begins..... or better yet can anybody help me figure out how to get those details myself ?

 

[YBNDM]

Very helpful, need to study this stuff!

 

[AlexXsWx]

Glad you're jumping into the frame data!

Here is the explanation on strings.

P is 17(3)21, so your active frames are 19, 20, 21. Since it is a string move, there is no need to account for recovery and add in any additional frames.
So since is 14(3)27 we take 21(from the result) + 14( start up) + 1(The phantom buffer frame) to find that the very first active frame of is 37. The active frames are 37, 38, 39.

In order to find an unholdable, Their first frame of recovery needs to be met by the very first active frame in your move. So will be an unholdable if you are +18 because that move is active on the 19th frame and the opponent loses their invulnerability on the 19th frame.

Does this make sense?

What is this "phantom buffer frame" that you've mentioned?

I understand that for majority of attacks it takes 1 extra frame on the beginning, so if I have a with frame data of "10 (2) 15", the active frames will be 12th and 13th.

Consider following example, which also shows how I check frame data:
I go to the endless training room against Zack, and make him perform Zack-beam while being as far away as possible. This gives me constant +87 when I block it (Steam Ver.1.07).
If I follow it up with the , my advantage goes down to 59, what confirms the initial extra frame (1 + 10+2+15 = 28; 28 + 59 = 87)
If I follow it up with (same punch twice), my advantage is 31. Let's see - 1 + 10+2+15 + 1 + 10+2+15 = 56; 56 + 31 = 87. This confirms the two initial extra frames, one for each punch.
However, if I follow up blocking of Zack-beam with a string (frame data of 10 (2) 15, 8 (2) 15), my advantage is 49. Let's check - 1 + 10+2 + 8+2+15 = 38; 38 + 49 = 87. I don't see any room for the second "phantom buffer frame", so what is that?

 

[illninofan]

Alright, so I understand the basics of Frame Data as stated in the game tutorial.

For example, Honoka's 6P (P), has values of 11(2)22.

11 Startup (these frames determine the amount of time from which the attack starts/the character is not moving to when the attack lands or not, frame 1 = start, frame 11 = did it land or did it get blocked?)
2 Active (the time in which the attack is hitting)
22 Recovery (the amount of time it takes to recoil from the attack and she can either defend, go on attack again, backdash etc. depending on the situation)

Then there's advantage and disadvantage, if the attack leaves you at a +2 on block, for example, this means you can move before your opponent does. A -2 is the opposite.

The issue for me is figuring out how to apply these data to become a better player (especially with regards to reactions/reading the opponent), but I think that comes with time and experimentation (as does everything else in the game).

Am I on the right track?

Greater question: who thought to use frame data as a tool in the first place? My silly guess is someone with programming knowledge back in the day broke into Street Fighter 2 and used the information to gain an advantage over his friends. XD

 

[Sly Bass]

Alright, so I understand the basics of Frame Data as stated in the game tutorial.

For example, Honoka's 6P (P), has values of 11(2)22.

11 Startup (these frames determine the amount of time from which the attack starts/the character is not moving to when the attack lands or not, frame 1 = start, frame 11 = did it land or did it get blocked?)
2 Active (the time in which the attack is hitting)
22 Recovery (the amount of time it takes to recoil from the attack and she can either defend, go on attack again, backdash etc. depending on the situation)

Then there's advantage and disadvantage, if the attack leaves you at a +2 on block, for example, this means you can move before your opponent does. A -2 is the opposite.

The issue for me is figuring out how to apply these data to become a better player (especially with regards to reactions/reading the opponent), but I think that comes with time and experimentation (as does everything else in the game).

Am I on the right track?

Greater question: who thought to use frame data as a tool in the first place? My silly guess is someone with programming knowledge back in the day broke into Street Fighter 2 and used the information to gain an advantage over his friends. XD

Knowing frame data helps with frame traps and knowing what moves to use next in situations. It makes mashy opponents appear to be noobs. You can throw punish properly, etc.

 

[Matt Ponton]

Greater question: who thought to use frame data as a tool in the first place? My silly guess is someone with programming knowledge back in the day broke into Street Fighter 2 and used the information to gain an advantage over his friends. XD

The common thought is "Wow, that move is fast! But so is this other move! How much faster is one move over the other? Well, the game runs at 60 frames per second, so as a measure of time we can use frame counting to determine since 1/60th of a second is too difficult to time with a stop watch at human reaction speeds."