What's new

Detailed Tests on LP/HP Jabs

KOHR

Noob


All parameters and their meaning will be explained in detail below, but it is not necessary to read everything - the larger the number in each table cell, the better.

LP Length: Horizontal Length of LP - The higher this parameter, the longer the length, the better the jab.
LP Priority: Priority of LP - The higher this parameter, the better protection by jabs from standing attacks.
LP Height: Vertical Height of LP - The higher this parameter, the lower the jab is to the ground.
HP Length: Horizontal Length of HP - It makes sense to find out which characters have a longer HP or LP. Take this into account when choosing which button to glitch jab with to defend against standing attacks.
HP Priority: Priority of HP - The higher this parameter, the better HP protection with jabs.

Detailed Description.


To understand the results you need to know about hitboxes, hurtboxes and pushboxes. Hitbox is the area where the attack is active. Hurtbox is the area where the hitbox will hit. Pushbox is an area that can not be crossed (the character model). In UMK3, all three regions are always rectangles. In the screenshot below, for example, roughly labeled Hitbox (blue) and Hurtbox (red).


Obviously, none of these rectangles coincide with the graphical representation of the character, so the previous table contains the real values that are processed by the game.

LP Length.

The longer the LP jab, the greater distance you can keep between you and opponent while runjabbing for your own safety. You can also start jabbing earlier, because your jab requires less distance between you and the opponent.

How this test was performed - the second player selects Scorpion (so that the hurtbox, pushbox and walking speed are always the same), and presses block, excluding the movement of the hurtbox. The first player with the tested character comes close, then the second player (Scorpion) moves frame by frame to the maximum distance, so that the first player touches Scorpion’s block hurtbox with their jab hitbox. The number of frames to be diverted is entered in the table. In fact, it denotes the maximum distance between the pushboxes of the characters on which the impact works under these conditions.

In the figure below, the areas of the hitbox of the attack (blue), the hurtbox (red), and pushboxes (white) are roughly marked. The value entered in the table is indicated by the green arrow. The actual length of the hitbox cannot be calculated because it is invisible, but this test allows you to judge its length indirectly.


While at rest, the characters will either stand close to the pushbox, or if the pushboxes cross, the game will push them apart in such a manner that the characters will slightly push away from each other. When you move two characters to each other, each time they will be at different distances. In the tests, only the smallest of the distances were taken, if this was not done, then in all results where length is measured, there would be a range of values instead of one, and this would also increase the number of tests required. The screenshot below shows the possible distances between the male and female ninja.


If you perform 2 jabs in a row, then the animation of the second one will be different from the first one - the jab is executed by the other hand. However, the hitboxes of both hits are the same.

LP Priority.

At impact, the hurtboxes of different characters are relative to the character's sprite, it can be closer or further from the front edge of the hitbox. In this case, Priority is represented by the green line.


The larger this distance, the more difficult it is to counterattack you from a standing position when you are jabbing. Because of this, the opponent must either attack between your jabs, or be further from the hitbox region of the jab. The hitbox of the opponent’s counter attack should also be longer than the green arrow.

As you can see below, Kabal's hurtbox is closer to the opponent than the Robot Ninja's hurtbox, but Kabal’s priority is still greater, as his jab is much longer. The opponent must be further to reach Kabal while his jab’s hitbox is active.


First, I will briefly describe how this test was performed, then there will be a detailed description for those who wish to repeat it.

In order to find out the position of the hurtbox of the test character, we will need to get hit by a down + LK from Scorpion. Similarly, we will need to test the length of LP, in order to measure the maximum distance of which the down + LK attack works, and simultaneously do the same with the jab of the test character. It was possible to get hit by any other attack suitable of length. The difference from the previous test is that we will measure the impact at the time of the jab’s first active frame - exactly then the hitbox appears in the test character’s hurtbox.


Moving to the specified distance, we get information about the hurtbox at the time of the jab’s first active frame - Now we are standing in the position: hurtbox + length (down + LK). Subtracting from this value already known to us, this is a number, we get the priority value, which differs from the real one by the constant component (the length of the impact is down + LK). The result will always be less than 0, as long as the auxiliary attack is quite long. The minimum value obtained is -17. If we add 17 to the number obtained, then the result of each test can be interpreted as the distance measured in Scorpion’s walking frames (different characters may have different walking speeds), on which the jabs of the test character are longer than the lowest-priority jabs, which are denoted 0.

Next is a detailed explanation of the practical implementation of the test. Down + LK must be performed in a way such that the first active frame coincides with the first active frame of the jab. A simple check shows that by starting the frame count from the moment the LP button is pressed, the kick will be blocked if the second player presses BL on the 11th frame. Likewise, the down + LK will be blocked by pressing BL on the 8th frame. Therefore, in order for the first active frame in both shots to coincide, it is necessary to start counting by pressing the LP button of the first player while standing by the second player, and press the down + LK of the second player on the 4th frame. Using save states and the fact that walking in UMK3 does not have acceleration - walking within, for example, 10 frames in a row is equivalent to walking 10 times, but one frame with pauses between it, you will easily find the right distance.

LP Height.

There are 3 different heights of LP jabs. In some matchups, high jabs do not connect against a crouch blocking opponent, which puts LP run jabs into question in these cases. For example, the character with the best numbers in all categories (except height) for LP jabs - Kabal, cannot run jab Stryker, which makes him change tactics in this matchup and play differently than against other characters. Here I only tested the bottom edge of the hitbox, which covers all ground applications of this attack. The higher the number in the table, the greater the chance for a LP to not whiff against a crouch blocking opponent.



How this test was carried out: Sonya crouch blocks in the corner, the player closely performs 3 jabs at the maximum speed. Let's consider all possible variants.

The highest jabs do not connect at all. The only character with such jabs - Stryker, is not able to efficiently pressure many enemies, as his jab does not hit if the opponent is in a crouch block state. Such jabs are denoted by zero in the table.

Standard jabs connect against everyone except Stryker. In the test, this situation is revealed if 2 out of 3 quickly inflicted jabs connect, and the last jab whiffs. This is due to the fact that the hurtbox of Sonya’s block is higher than that of Stryker, but it is low enough that when the hurtbox moves (if the jab hits Sonya, the hurtbox starts to move), one of the jabs whiffs. Such jabs are denoted by a 1 in the table.

Low jabs connect on all crouch blocking opponents without exception. In this test, this scenario is revealed if all three jabs connect on the crouch blocking Sonya, and you were thrown back by the pushback from the corner. Such jabs are denoted in the table by 2.

HP length.

HP length, all tests are similar to the case with LP. It makes sense to find out which hits the character has longer - HP or LP and take this into account when defensively to defend against standing attacks.


HP priority.

Priority of HP, all tests are similar to the case with LP. HP, in addition to protecting from standing attacks, is used to protect against air attacks, where the height of the upper edge of the hitbox is also of great importance, here this case is not considered.
 
Last edited:

Biggs

FOOOOOOOOOOOOOOOOOOOOL
@KOHR really did an excellent job.

Some implications of this data are:
  • Male ninjas are no longer regarded as having the second longest LP range. Unmasked Sub-Zero has the second longest LP range.
  • Sheeva has the best HP in the game (it is not the longest however).
  • It also appears that Shang Tsung has better jabs than Male and Female Ninjas...
 

dubson

Noob
@KOHR really did an excellent job.

Some implications of this data are:
  • Male ninjas are no longer regarded as having the second longest LP range. Unmasked Sub-Zero has the second longest LP range.
  • Sheeva has the best HP in the game (it is not the longest however).
  • It also appears that Shang Tsung has better jabs than Male and Female Ninjas...
yeah Sheeva's jabs have the highest priority, but that's only one piece of the pie so i cant label them the best yet

off of glitch jabs it's really strong into conversions, especially in anti air situations

but her very large character hitbox makes things awkward though, both for and against her in terms of actually connecting with them and converting

strykers have longer range, but a he has a better sized character hitbox relative to the rest of the cast and can potentially infinite off of his

and that's just one example, all pieces of the pie

i knew Shang and USub's were underrated. me and @Shock talked about that ages ago

kabal numbers lookin super strong
 

Shock

Administrator
Premium Supporter
I know there's other work done on all the frame data. Can we publish a complete guide to all the numbers available? I started working on it character by character years ago and only did a few, but that hard drive died, I'm going to try to revive it soon but I think all the stuff I worked on was covered by others eventually.
 

Biggs

FOOOOOOOOOOOOOOOOOOOOL
I'm pretty sure someone else has worked on stuff but I can't remember, someone here definitely knows.
Assuming game is running at 60 FPS (Mame, I think Arcade runs at 56) and Player advantage. Frames were counted using Frame Advance starting at the frame the button was pressed and counted until the opponent was hit.

STARTUP
LP 13
HP 13
LK 10
HK 10
dLP 13
dHP 10
dLK 10
dHK 17
bLK 19
bHK 16
clLP 4 (This is when the character begins the throw animation, I believe it is inescapable at this point)
clHP 7
clLK 7
clHK 7
BK 4 (Animation is 9)
cBK 4 (Animation is 9)
 

Shock

Administrator
Premium Supporter
Assuming game is running at 60 FPS (Mame, I think Arcade runs at 56) and Player advantage. Frames were counted using Frame Advance starting at the frame the button was pressed and counted until the opponent was hit.

STARTUP
LP 13
HP 13
LK 10
HK 10
dLP 13
dHP 10
dLK 10
dHK 17
bLK 19
bHK 16
clLP 4 (This is when the character begins the throw animation, I believe it is inescapable at this point)
clHP 7
clLK 7
clHK 7
BK 4 (Animation is 9)
cBK 4 (Animation is 9)
Yes I have this data as well, you didn't find any exceptions? I started going through recoveries, block stun, etc but that was many years ago. Also arcade is 53.2, very strange refresh rate.
 

Biggs

FOOOOOOOOOOOOOOOOOOOOL
Yes I have this data as well, you didn't find any exceptions? I started going through recoveries, block stun, etc but that was many years ago. Also arcade is 53.2, very strange refresh rate.
I haven't found any exceptions so far. I believe things start to change drastically for the recovery. I went back and checked PC MKAK and it runs at 54 FPS. The arcade refresh rate of 53.2 is very weird...
 

Shock

Administrator
Premium Supporter
I haven't found any exceptions so far. I believe things start to change drastically for the recovery. I went back and checked PC MKAK and it runs at 54 FPS. The arcade refresh rate of 53.2 is very weird...
I'm pretty sure RHs have significantly different start ups, like for example Sheeva and Sonya's are at least 5 frames different and most characters are somewhere in between that except like USub who also is probably Sheeva startup. There also might be single frame differences on some close range attacks but that might just be some dedicated priority or collision related issues. Also make sure you're using the character with frame advantage to test default timing.
 

MKK hanzo

Moderator
Yes I have this data as well, you didn't find any exceptions? I started going through recoveries, block stun, etc but that was many years ago. Also arcade is 53.2, very strange refresh rate.
This always puzzled me. Newer mames show 57/60 fps, MKAK goes at 54/60, older mames 53.2/60 and retroarch mame 2003 55/60

Funily by eye the screen refresh of the actual arcade seems to be 60 unless you use an old camera and film the screen and see the black bars passing by.

Save for the stupid speed differences, the 60fps refresh on mks is only on the crap midway arcade treasures 2 compilations (NOT on UMK tho) and console releases (even snes and genesis) and trilogy.

Easiest way to compare? Just let the character tower roll. On ALL EMULATORS AND PORTS you can see some "skips" while the tower is rolling down. NOT on arcade and the aforementioned versions tho,
 

Marinjuana

Up rock incoming, ETA 5 minutes
Can some give me a primer on jab pressure in this game and the throw meta? I get that you can cancel jab with run to do it faster but I don't really get how it comes together. How do I defend against it?