Jump to content

Mister Wu

Members
  • Content Count

    43
  • Joined

  • Last visited

Community Reputation

6 Neutral

About Mister Wu

  • Rank
    Veteran Battle Trainer
  • Birthday August 26

Battle Pokémon

  • Main
    Charizard

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Here we are for the March 2019 top 5 charts of Wii U online ranked matches! And there are two surprises too! Well, one actually, the other is just a consequence of the first report being written in April 2018, thus meaning that the March 2018 charts are available for a comparison! POINTS Again, there are quite a few changes compared to the chart of last month: On one hand, yuukatubasa is still at the top, with an increase of 649 points compared to last month; on the other hand, perea81 had a more significant increase of points, 4624 points, which allowed them to reach 3rd place. It's also interesting to note how Mellocair, who appeared in the wins-based chart last month, appeared in the points-based chart of this month as well. They got no less than 1870 points this month. Very interesting is the performace of herobrien/jervis6 as well, who obtained no less than 9178 points, ending in 2nd place. Finally, this time even the minimum increase in points of Lashe in 5th place can be estimated, it being 878 points. Since these reports started on april 2018, we can also see the charts of March 2018 to make a quick comparison on the activity: Back then, more than 3 times the points were needed to reach 1st place, even 5th place had a higher amount of points than the current 1st place, although the two values are more similar. WINS The changes with respect to last month here are less drastic than those observed in the comparison of the points-based charts: yuukatubasa, Mellocair and perea81 all reappear from the previous month, with an increase of wins which is 12, 70 and 100 respectively. In 5th place we see the return of Odilonreis, who was last seen in the charts on Novemeber 2018. This time they won no less than 21 matches to return in the chart. The most intriguing change in this chart, though, can only be appreciated when looking at the chart of March 2018: As surprising as it might look, Kevin/spyro22 finally managed to reach the top of the chart after disappearing from the charts for an entire year! No less than 284 wins were needed for the achievement, which is actually the biggest minimum increase seen in the current charts. When comparing the number of wins, the values still tell us that more than twice the number of wins were needed to reach the top of the charts last year, although even with the results of this month Kevin/spyro22 would still have reached 5th place the chart of March 2018, adding to the merits of said results. To qucikly sum things up, the comparison with the results of last year definitely indicate that the expected decline in activity in the online ranked matches effectively happened, with the current activity being possibly one half to one third of the activity of last year. Despite that sharp decline in overall activity, some dedicated players didn't drop the mode, with perea81 and Mellocair continuing to have some matches this month and Odilonreis returning to them after having disappeared last December. The real surprise of this month though was the return of Kevin/spyro22, who managed to finally reach first place after last March they only managed to reach 2nd place. Considering how they had since disappeared from the charts, said return with a result that would have allowed them to appaer even in the top 5 chart of last year definitely stands out, to say the least.
  2. Here are the parameters (statistics), attack and hit tables used by Pokkén Torunament DX version 1.3.3: Parameters: https://docs.google.com/spreadsheets/d/14XWvka5YlWjcftZ2cbVJAVsK12_0CoFrwRwI0dFuKxs/edit?usp=sharing Attack (fighter Pokémon): https://docs.google.com/spreadsheets/d/1Cmpw5Xbvwr73KLMAcnnW5u16RRh9s-zPUdgADud4Wzc/edit?usp=sharing Hit (fighter Pokémon): https://docs.google.com/spreadsheets/d/1dsX-lF-0aAWQjKA_qyoBWVYnQ7zXrET-pqM-UD_s2AA/edit?usp=sharing Parameters: various stats of the fighters and Support Pokémon like HP, size of the Burst Gauge, duration of Synergy Burst, time needed to charge the support gauge, and so on Attack: various data about the attacks of the game, such as damage, scaling, ability to break shield or pierce armor, alteration of the Synergy and support gauges, stun duration, how the attack alters the status of the opponent, what is the heght and type of the attack, and so on Hit: various data about the hitboxes, including size and duration It is important to stress how these tables do not contain data about the moves that use the attacks! They do not contain the total duration of a move, when the attack is used in the move, the attributes that are set during the move as well as additional effects such as recoil damage or reduction of debuffs, and what are the move the current move can cancel to. Said information is much more time consuming to extract. The Attack and Hti tables of the Support Pokémon will come at a later time as well.
  3. Here is the monthly update on the TOP 5 charts of Wii U online ranked matches! POINTS The chart has some pretty big differences compared to that of last month: Only two palyers are returning, RieBear whose points increased by 125, and yuukatubasa, whose points increased by 625. These small increases are already enough to reach second and first place, respectively. Of the newcomers, we can only estimate the minimum points increase of sea6741 in third place, which is 4867 points, an amount only slightly lower than the minimum estimated amount of last month, 5219 points gained by TOMRO_SUN who then disappeared from the charts of this month. WINS The differences with respect to past month are similar to those seen in the points' chart: We see that agin, the number of wins of RieBear increased by 5, while the number of wins of yuukatubasa increased by 11. These are again small increases, yet they are enough to reach first and second place. Since the other players have a number of wins which is less than 606, it's not possible to estimate the number of wins they had in the previous month. One thing to note, though, is that in third place of this chart we have the return of Mellocair, a player maining Garchomp active in September and October 2018. Last time that they appeared in the charts, their rank was A5, so this again suggests that the player has been active all these months. Overall, these results again seem to indicate how many current players reach these charts through short but intense bursts of activity, as indicated by the little increases in the following months as well as them leaving the charts in just a few months, likely as soon as the points and wins obtained in these bursts expire. Notable exceptions can still be seen, that of this month being the return of Mellocair, who was in these charts a few months ago, with a much lower rank. Likely due to expiration of wins and points, this month the numbers seen in the charts are much lower, to the point that it's not possible to make reliable considerations on this month's overall activity in this mode. Hopefully, next month an estimation of the minimum number of wins for the newcomers will be possible, allowing to have a better idea of the actual activity.
  4. Mister Wu

    How clashes work

    APPENDIX CLASHES WITH FIGHTERS IN THE AIR Clashes with fighters in the air work in the same way as clashes with fighters on ground, with one imprtoant difference: for the fighter in the air the stun of the stronger knockback lasts the same amount of frames as the normal knockback. In the cases observed the stun lasted 30 frames, one frame less than the normal ground knockback stun that lasts 31 frames. 's upward attack (8Y, clash priority of 2) against 's Counterattack (X + A, clash priority of 3) 's Fire Punch (clash priority of 3) against 's Counterattack (X + A, clash priority of 3) CLASHES OF DIRECT ATTACKS WITH GRABS When a direct attack clashes with a grab clash priority is not considered, and the direct attack always wins the clash and connects, even in the cases in which there should be knockback for both fighters instead. 's simple punch (5Y, clash priority of 1) against 's standard grab (Y + B, clash priority of 0)
  5. Mister Wu

    How clashes work

    's 8Y, 6Y and 5Y have a clash priorty of 3, 2 and 1 respectively, neither of these three attacks are able to break light armor, so priority is not used to determine whether an attack will be able to break light armor or not. 0 clash priority is used for special cases like impacts and possibly grabs. It's not used for the usual moves, in any case remember that 0 clash priority can still cause a knockback clash against priority 1 moves, since the clash priority difference is 1 - 0 = 1. You must consider the difference among clash priorities.
  6. Mister Wu

    How clashes work

    As @bryansempai pointed out, there's actually a very funny case in which having a clash priority higher than 2 results in the expected outcome, but the results are not... satisfactory! 's weak high attack (8Y) has a clash priority of 3 and thus wins the clash against 's low weak attack (2Y), that has a clash priority of 1. However, 's 8Y is unable to break Burst's light armor, resulting in continuing his attack and connecting the whole combo! You can see this fun corner case here:
  7. An update: I was wrong about how the attack height is determined: attack height is a coded value found in the attack table, just like the attack type. Furthermore, 's Flying Stance was able to evade standard grabs simply because those grabs have special flags that allow them to only grab opponents on ground and not opponents in the air. I now corrected the posts I made and uplaoded a new image that shows a case of actual evasion due to attributes.
  8. Mister Wu

    How clashes work

    Very simply, this post will deal with how clashes work, now that the value determining clash priority has been found! Clash priority is found in the attack tables, here are the ones of Pokkén Tournament DX version 1.3.3, while here are the ones of the Wii U version. The clash happens when two direct attacks' or throws' active hitboxes collide with each other before colliding with one of the fighters. It's a rather rare event with jabs, while it can be seen more frequently when moves with many active frames are involved (e.g. 's charged footdive, j[X], against 's upward spinning attack, ]X[). What happens in the clash depends on the clash priority of the moves, said value ranges from 0 to 5, with 0 and 5 being reserved for special cases (the latter in particular is used for Burst activation and for the Burst Attacks, but also some moves have that priority, like 's charged Force Palm). SAME CLASH PRIORITY In this case, the clash results in a small yellow circle followed by a knockback animation for both fighters with a stun that in general lasts 31 frames. 's simple punch (5Y, clash priority of 1) against 's simple punch (5Y) 's simple scratch (5Y, clash priority of 1) agaisnt 's simple scratch (5Y) CLASH PRIORITY DIFFERENCE OF 1 This case is very important as the animation is easily confused with the animation of the case of same clash priority, but the outcome is very different. In particular, the yellow circle is slightly bigger, and this time the fighter with the higher priority move will have the usual knockback, while the fighter with the lower priority move will have a stronger knockback with a stun that lasts longer. In the cases I tested with and , the stun of the stronger knockback lasted 42 frames instead of the 31 frames of the usual knockback, leaving the fighter with the higher priority move at a +11 frame advantage. 's belly bump (6Y, clash priority of 2) against 's simple punch (5Y, clash priority of 1) 's simple slash (5X, clash priority of 2) against 's simple scratch (5Y, clash priority of 1) CLASH PRIORITY DIFFERENCE HIGHER OR EQUAL TO 2 In this case, an impact with a big yellow circle can be seen, following that the move the higher priority will connect. 's Dragon Claw (FS X, clash priority of 3) against 's simple punch (5Y, clash priority of 1) 's flying kick (8X, clash priority of 4) against 's simple punch (5Y, clash priority of 1) 's Knock Off (6A, clash priority of 3) against 's simple scratch (5Y, clash priority of 1) 's Ice Punch (8A, cash priority of 4) against 's simple scratch (5Y, clash priority of 1) In short: SAME CLASH PRIORITY → Same type of knockback and stun for both fighters CLASH PRIORITY DIFFERENCE = 1 → Stronger knockback and stun for the fighter with the lower clash priority move CLASH PRIORITY DIFFERENCE ≥ 2 → The move with the higher clash priority connects Additionally, knockback gives 60 Mega values to each of the fighters, while having the clash won by a single move (priority difference ≥ 2) gives 30 Mega values uniquely to the fighter winning the clash. Furthermore, here are two additional aspects that i'll deal with in the appendix: for the fighter in the air, the stun of the stronger knockback lasts the same as the stun of the normal knockback direct attacks interacting with grabs always win the clash regardless of clash priority And that's what there is to say at the moment about clashes, the most important part is being able to recognize the case in which a priority difference of 1 is happening, as those cases generally give a considerable frame advantage but are easily mistaken as clashes with the same clash priority. If it's not possible to tell the yellow circles apart, it's important to remember the clash priority of the moves that are frequently used.
  9. 3) EXAMPLES OF ACTUAL USE Let's now see how the attributes were set in the examples I provided above: EVASION As we'll see below, the preliminary phase of the jump has the attr_jump set for the whole phase, it sets the following attributes: h throw = 1 m throw = 1 l grap = 1 l shot = 1 l throw = 1 Since there is evasion on grabs of all heights, the standard grab (B + Y) is therefore evaded. As an aside, the standard grab has a special flag set that prevents it from grabbing opponents who are not on ground, for this reason, and not because of evasion properties, the standard grab fails to grab in Flying Stance: INVINCIBILITY In the Wii U version, the Burst Attack started in the air has the attr_muteki_all_hml set for the whole duration of the preliminary animation, said function sets the following attributes: h grap = 2 h shot = 2 h throw = 2 m grap = 2 m shot = 2 m throw = 2 l grap = 2 l shot = 2 l throw = 2 Since invincibility is set for all attack types, the counterattack is evaded. RED ARMOR The high stance has the attr_attack_armor_m set from frame 4 of the animation (5th frame) until the last frame in which the high stance button is pushed. This function sets the following attributes: h throw = 3 m grap = 3 m shot = 3 m throw = 3 l throw = 3 It should be noticed how this is a composite of the attr_attack and attr_armor_m functions, composite functions are commonly used in the game to set multiple attributes at once. In this case, since there is red armor on all the throw attacks, the grab is red armored. BLUE ARMOR Fire Punch in Duel Phase has the attr_block_all_hml set from frame 0 to frame 24 of the animation, this function sets the following attributes: h grap = 4 h shot = 4 m grap = 4 m shot = 4 l grap = 4 l shot = 4 Since all the direct attacks are blue armored in that time frame, this simple punch attack (5Y) is blue armored as well. SHIELD The first hit of the counterattack triggers the dm_guard_weak animation that has the attr_guard_freeze set from frame 0 until frame freeze+12 (12 frames after the guard stun animation), this function sets the following attributes: h grap = 5 h shot = 5 h throw = 1 m grap = 5 m shot = 5 m throw = 1 l grap = 5 l shot = 5 l throw = 1 since the second hit of the counterattack hits before guard stun is finished, said second hit is automatically shielded even without the player's input. A notheworthy detail: during guard stun, grabs are set to evasion, this explains why the simple punch (5Y) canceled into Seismic Toss is evaded: Seisimic Toss is active before evasion has ended. REPELLING RED ARMOR In the case of the simple punch (5Y), the function attr_attack_repel is set from frame 0 to frame 7, said function sets the following attributes: h throw = 6 m throw = 6 l throw = 6 since the grab hits in that time frame, it is red armored and then the player is repelled. JUMP Jump consists in two phases: Phase 1 goes from frame 0 to frame 4 During this phase it's possible to cancel the jump into ground moves. Phase 2 starts from frame 5 onward In this phase it's possible to start aerial moves. Phase 1 has the attr_jump set for the whole phase, it sets the following attributes: h throw = 1 m throw = 1 l grap = 1 l shot = 1 l throw = 1 Phase 2 has the attr_jump_start set from frame 0 until frame 8, it sets the following attributes: h throw = 1 m grap = 1 m shot = 1 l grap = 1 l shot = 1 both these functions explain the peculiar evasion properties of jumps. STEP BACKWARD The step backward (also known as backdash) has the attr_backstep function, that in the case of is set from frame 0 to frame 7 of the animation. It sets the following attributes: h grap = 2 h throw = 2 m grap = 2 m throw = 2 l grap = 2 l throw = 2 Stepping back allows to avoid grabs and direct attacks, but not ranged attacks. Surprisingly, this function sets the evasion using the invincibility attributes:
  10. Since I could recently examine this aspect fo the game, I'l lbriefly explain this crucial part of the game: what are the special attributes that the game applies toward incoming attacks from the opponent As you'll see, this includes blue armor, red armor and even evasion. 1) CATEGORIES OF INCOMING ATTACKS First of all, let's see how the game categorizes incoming attacks from the opponent: the incoming attacks can fall into various among 9 categories, based on height and attack type, that are both coded in the attack tables (here are the ones of the fighter Pokémon of Pokkén Tournament DX version 1.3.3, while here are the ones of the fighter Pokémon of the Wii U version) 1.1) HEIGHT The game has three heights: h = high m = mid l = low It is important to note how an attack can actually fall into multiple heights; in particular the coded heights, along with the name we currently give them, are: h = High hm = Mid High hl = not currently categorized m = Special Mid ml =Mid Low l = Low hml = True Mid 1.2) ATTACK TYPE The game has three attack types: grap = direct attack shot = ranged attack throw = grab in this case, each attack fit into just one of these categories. 2) ATTRIBUTES To modify the behavior toward incoming attacks, the game potentially can rely on eight attributes, although it is possible that only six of them are actually used. These attributes are: -1 = standard behavior. The reaction to the incoming attack is the usual one: 0 = set by the unmapped attr_yarare_air functions. Since the attributes are also initialized to this value, it's possible that it works in the same way as the -1 attribute 1 = evasion, set by the attr_evade functions. The attack simply fails to connect, with the hitbox not interacting with the player: 2 = invincibility, set by the attr_muteki functions that are used in BA as well as some hitstuns. The attack can't interact with the player: 3 = red armor, set by the attr_armor functions (for grap and shot attacks) as well as the attr_attack functions (that set it for throw attacks of all heights). Hitstun is removed, allowing the animation to continue after the hitstop, and the damage of shot type attacks is halved. Phase Shift Points are still gained and KO is not prevented. It stops working at the 10th attack or when hit by a piercing attack. In the case of throw attacks, 30 Mega values are gained, the grab fails to connect and the grab anitmation goes on as if the grab was whiffed: 4 = blue armor, set by the attr_block functions (exclusively for grap and shot attacks), the attack deals no damage and causes no hitstun, allowing the animation to continue after the hitstop, furthermore 12 Mega values are gained and the Phase Shift Points are negated. It stops working when hit by a piercing attack: 5 = shield, set by the attr_guard_freeze functions (exclusively for grap and shot attacks). The attack is shielded, even if the player is no longer holding the guard button: 6 = repelling red armor, set by the attr_attack_repel functions for throw attacks (although unused attr_attack_repel_armor functions that set it for the other attack types exist in the code), it works like the red armor, but after the hitstop, the player is repelled backward: 7 = impact, set by the unmapped attr_impact functions, since those functions are not mapped to common states or moves of the fighter, it's not currently possible to know what this attribute does
  11. Here are the top 5 charts of ranked online matches of the Wii U version of Pokkén Tournament! POINTS The results are different from what was seen last month: Only Mario.Luigi.taku returned, with 500 points more than they had last month. Furthermore, only in the case of TOMRO_SUN we can estimate the minimum gain in points, with it being 5219 points, since the others' points fall below the person in last place last month. WINS Even here there were significant changes compared to last month: Only Mario.Luigi.taku returns, having won no less than 10 matches. Furthermore, only with TOMRO_SUN we can calculate the minimum number of matches won last month, which is 122. In general, these numbers suggest that the spike in activity seen last month was likely due to the holidays, although it's difficult to make an appropriate estimation due to the newer people in 3rd, 4th and 5th place having less points or wins than the person in 5th place last month. One more time, the next month will be needed to see the actual activity of the people who just entered the charts, considering the lower numbers seen in these charts it is nonetheless to be expected that the activity of the next month will be similar to the activity seen in the case of TOMRO_SUN.
  12. The feeling when I almost forgot about this Here is the monthly update on the top 5 charts of online ranked matches in the Wii U version of the game! POINTS When compared to last month, there was a noticeable increase in points: While artofspeed and vince returned, the former gaining no less than 17362 points and the latter gaining no less than 574 points, the other three were new. The most impressive case is that of Mario.Luigui.taku, who gained no less than 35540 points and ended up in first place. Very interesting is the return of Jessica Parra, who had disappeared from the top 5 charts since last August. Finally, in 5th place there was a newcomer, Shosei. WINS This time, the wins-based chart perfectly matches the points-based chart. Comapred to last month, only vince returned, having won no less than 5 matches last month. This match between points-based chart and wins-based chart means that the placement in the chart wasn't heavily favored by the relatively middle rank of taku, Jessica Parra and Shosei (B rank). Indeed, taku won no less than 469 matches last month and Jessica won no less than 451 matches. Shosei on the other hand needed to win no less than 41 matches last month to enter these charts. Since these are lowest bounds for the estimation, with the actual achievement possibly being significantly more important (up to 642 matches won last month in the best case for Shosei), it's important to be careful before considering the achievement of Shosei less impressive. In any case, the lowest bounds of the estimation were higher than those seen last month, suggesting an increase in activity, which is after all partially to be expected due to the Christmas or Winter holidays. It was in particular quite interesting to see the return of Jessica Parra, who had disappeared from the charts since last August. The results of the next two months will allow us to see what was the actual effect of the holidays.
  13. As usual, here are the top 5 chart of the onlnie ranked matches in the Wii U version of Pokkén Tournament, to have an idea of the activity in this mode! POINTS When compared to the results of last month, something interesting can be noticed: While first place still belongs to luffy who gained no less than 4775 points compared to last month, all the other places now have completely different people, with an increase in points that is expected to be greater than that of luffy except for ririri in 5th place, considering how last month they all had less than 36293 points. WINS Even in this case, the changes involve 4 people in this chart: While in the case of luffy the increase in number of matches won is 87, in all the other cases the people in the chart are completely new. Banisher1 and Vince are also in the points-based chart, confirming their relatively solid performance last month. an interesting thing that can be noted is how Odilonreis and ryuujiid have less wins than mojucohi and Mellocair had last month, potentailly indicating that each month the wins that are several months old are removed from these sums. In any case, considering how last month they had 626 or less wins, Banisher1 and Vince won last month at least 326 and 209 matches, respectively. These are numbers that are higher than those observed last month. In general, different people were mainly active this month, showing an increased activity compared to what could be inferred last month. Among them, Banisher1 and Vince stood out. Since the wins-based top 5 charts of this months suggested that each month the matches won that are several months old are no longer counted, it's important to point out that this estimation is likely to underestimate the actual number of matches won each months. It is possible that the activity in this mode is greater than what these charts show.
  14. Mister Wu

    About grab crushing and grab teching

    Now that I had a more in depth look at the data, I discovered the following: every attack has various properties that determine evasion, armors and even grab crushing, in particular the most common and relevant for us are: attr_attack_repel grab teching attr_attack_repel_evade grab teching and evasion attr_attack grab crushing attr_attack_evade grab crushing and evasion attr_evade_all evasion only these properties are follwed by the starting frame and, in case it is defined, the ending frame, with 0 being the first frame. This means, however, that the window in which grab teching happens is not fixed, but varies depending on the attack. As an example, here are the properties of the basic weak and strong attacks of in Duel Phase (Wii U Pokkén Tournament version 1.3): 5Y attr_attack_repel 0 7 attr_attack 8 17 6Y attr_attack 0 9 2Y attr_attack_evade_h 0 9 attr_evade_all_h 10 8Y attr_attack_repel 0 3 attr_attack_evade_hm 4 15 attr_evade_all_l 16 5X attr_attack_repel 0 7 attr_attack 8 21 6X attr_attack_repel 0 7 attr_attack 8 17 attr_attack_evade_l 18 25 attr_evade_all_l 26 50 2X attr_attack_repel_evade_h 0 7 attr_attack_evade_h 8 17 attr_evade_all_h 18 8X attr_attack 0 13 attr_attack_evade_l 14 15 attr_attack_evade_ml 16 29 attr_evade_all_l 30 46 attr_evade_all_h 47 54 From these data we can see two things: The grab teching window was set independently of the actual impact frame: it is 8 frames both in the case of 5Y and 6X that are i19 and i27, respectively The grab teching window wasn't set to a fixed value either: 8Y has a 4 frames teching window instead of the usual 8 frames
  15. Good thing that this ranking doesn't count the lost matches nor the lost points, and only sums the won matches/points...
×