Talk:Hitlag

This page may seem useless with the lag and flinch articles. However, I think there's a big distinction between them:

• Freeze frames are the lag that occurs with attacks.
• Flinching is when a character's current action is interrupted and cannot take action, not the fact that the character suffers lag.
• The lag article was to be split anyway, so maybe now it can only encompass Wi-Fi lag and TV lag.

Toomai Glittershine eXemplary Logic cntrbs 16:14, 5 July 2009 (UTC)

Move

Oppose definitely used more, but i like freeze frames better Poultry(talk) the Team Liquid 07:12, 22 April 2016 (EDT)

I am going to push very hard against this one. "Freeze frame" isn't exactly an unknown term and is actually descriptive of the effect. "Hitlag" may be more popular, but it is extremely confusing when used in conjunction with the more set-in-stone "hitstun" - both terms sound like "the period of time after you're hit before you can do anything", so I think it would be to our benefit to only use one to major effect around the wiki. I also suggest that due to this similarity, "hitlag" doesn't sound like it can apply to damageable items and stage elements, which it indeed does. Toomai Glittershine The Brazen 09:32, 22 April 2016 (EDT)

Oppose. While Hitlag is way more used, as Toomai said, the similarity between that term and hitstun will probably be very confusing to readers, and one of the most important things on a wiki is readability IMO. Nyargleblargle (Contribs) 12:59, 22 April 2016 (EDT)

Oppose, per Toom and Nyargle. Miles (talk) 13:49, 22 April 2016 (EDT)

Oppose per obligatory agreement Toomai. Disaster Flare (talk) 13:56, 22 April 2016 (EDT)

Move (take #2)

This was brought up before, but I think it bears being considered again, as it is still relevant to this day. The term "freeze frame" is a pretty niche and even rather juvenile one. It is my understanding that the only reason we still call it freeze frames instead of hitlag is because of the potential for hitlag to be confused with hitstun. However, aside from the fact that Nintendo is doing a pretty good job at making sure of that themselves, I don't think this is a particularly good reason to use a more obscure term. Of the 3 terms used (hitlag, hitstop and freeze frames), freeze frames appears to be the least commonly used in my experience, and I think it's pretty clear to anyone that hitlag is the most commonly used, especially with Ruben's script viewer becoming more and more widely used as a resource by the community. We tend to use the most common community terms in our page titles, and I don't think there's a particularly good reason for this to be an exception, especially when the alternative we are using is such a poor one. Alex the Weeb 05:47, February 21, 2020 (EST)

Your argument seems to be based on the premise that "freeze frame" is a "poor" and "juvenile" choice. You have provided no evidence of this. Conversely, here are the Google results I am currently seeing:

• Smash Bros "hitlag": 4,880
• Smash Bros "hitstop": 2,630
• Smash Bros "freeze frame": 121,000

SmashWiki by itself is not going to be the cause of "freeze frame" being dominant in search results; something else is displaying a community preference for the term overall. Toomai Glittershine The Chilled 06:49, February 21, 2020 (EST)

If you're going off of the number of Google search results then I'm not surprised the number for freeze frame is so much higher. Freeze frame can refer to a lot of things, such as entertaining Smash Bros. screenshots, but hitlag and hitstop are very specific terms that can only really mean one thing. As for freeze frame being a poor term to use, I guess that is subjective, but given the kinds of things the term freeze frame can refer to, such as childrens' party games, I think there is at least some case for saying that it is a term with juvenile connotations at least. Alex the Weeb 06:55, February 21, 2020 (EST)

I agree with Toomai. Freeze frame, no matter how juvenile it is, is just much more commonly used. In fact, I don't think I've heard "hitlag" outside of this wiki. 72.203.118.154 10:34, February 21, 2020 (EST)

As I have explained before, the Google search results are incredibly misleading. If you want an example of why, just take a look at the differences in content you find from searching for hitlag and freeze frame on Twitter for instance. You'll notice it's far easier to find smash related content for the former than the latter, because hitlag is almost exclusively attributed to smash, whereas freeze frame is not, and even in the context of Smash can refer to other things. However, a consequence of this is that there will be a larger pool of search results for freeze frames due to its range of uses. As I mentioned before, hitlag is also the term that appears in game scripts, and on many technical sites as well. Please stop using this misleading statistic, or personal anecdotes. It is trivial to demonstrate hitlag's common usage in the Smash community. Alex the Weeb 10:53, February 21, 2020 (EST)

I agree with Alex, here. The game's scripts refer to hitlag as hitlag and not as freeze frames. I was denied the term "windbox" for frame strips because that's not what they're called in the game's data. If that also applies to freeze frames, it should be called hitlag. Alex is also correct for saying that hitlag is more widely used in the Smash Community. When people refer to it, they don't say freeze frames.Hitbox Enthusiast Zeck (talk) 14:50, February 21, 2020 (EST)

I also agree with Alex. It is officially refered to as hitlag within the game's files and not as freeze frames. I understand that freeze frames may have been a more common term years ago when people's knowledge of Smash was lower as a whole but it is not used as much anymore due to being not fully accurate. It's the same reason why FAF has been a more accepted term than IASA in more recent years as IASA is not always an accurate term whereas FAF is.User:SuperSqank (talk) 16:06, February 21, 2020 (EST)

I would like to clarify that "hitlag" is not used in any game data, but rather "hitstop". The "hitlag" on any data websites is fan-assigned.

Other than that, I don't have any other argument than "hitlag and hitstun are too similar". Toomai Glittershine The Sharp 18:51, February 21, 2020 (EST)

Resetting the weird indentation thing since my thoughts are quite elaborate. After that trainwreck of a discussion in the Discord I thought I'd leave my support for the change. Hitstop would be my preference as someone from the FGC but hitlag is absolutely superior for its prevalence across resources and Smash lingo alike. Anybody getting into competitive or learning mechanics will only wind up being confused by the wiki's arbitrary terminology. Someone using Ruben's Calculator or Data Viewer could see the terms, want to know more, come here, and find that themselves confused. The scripts of the game even say different (Meshima's param sheets specifically). If you worry about hitstun/lag confusion then simply put a "not to be confused with" note up top like with other pages. The introduction along with the header is enough to explain what the concept is.

The guidelines of SmashWiki clearly state that the common fan terms are used if they are available. I believe Toomai is, intentionally or not, misrepresenting evidence here. As Alex states, you'll hear hitlag virtually anywhere else. Twitter results would show a large difference between hitlag and freeze frames. The FGC, too, would show freeze frames just...isn't a term. At all. The sample from the search results should also be taken into consideration, as these are small occurrences for a decade-old community. The wiki's job is to provide the most common terminology, and anybody listening to commentary, at least mid level players, or anywhere in the community, would hear hitlag. Look at those tournament streams on Twitch with 50k+ viewership, which will clearly use hitlag.

I also want to mention that hitlag is used on virtually every instance of where freeze frames should be on this wiki. While the wiki wants to reduce confusion, it is clearly failing in this regard. Make a bot to replace the instances or simply adopt hitlag as a term. While it serves as a redirect akin to super armor, why would you have this everywhere on the wiki but direct it to somewhere else? Does this not confuse readers?

I see absolutely no reason as to why freeze frames has remained a term on this wiki for so long. All it's doing is increasing the wiki's disconnection from the community's defined terms. --Plague von Karma 19:00, February 21, 2020 (EST)

I'm rather indifferent on this one for now. Even though hitlag is the more common term and the one I'm more familiar with, it can get confused with hitstun, which refers to this and not freeze frames which Nintendo just wonderfully decided to call "hitstun". OmegαToαd64 02:39, February 22, 2020 (EST)

Althought the terms could be confused, I find it hard to believe someone would mistake hitlag for hitstun while viewing its page as it has a very clear description of what hitlag is. If anything, it would correct their previous mistake if there was one to begin with.Hitbox Enthusiast Zeck (talk) 17:44, February 22, 2020 (EST)

Just want to bump up discussion a bit, as I feel this is an important thing to go over. I feel the arguments given in opposition to the change are also not answering anything given in support. We aren't going to get anywhere if this is just dropped. --Plague von Karma 22:05, February 29, 2020 (EST)

Even though I initially opposed moving, I am now Neutral. While I have heard Freeze Frame more often, I don't think it being confusing with hitlag is a valid reason to not move it.

Also, probably off topic, but this is also [[an issue with almost every single discussion going on right now. If I may give constructive criticism, this wiki needs to get a handle on things. 72.203.118.154 00:23, March 1, 2020 (EST)

Care to elaborate? It's not really constructive criticism if you don't actually give us anything. SerpentKing 23:07, March 8, 2020 (EDT)

History merge

Can someone with the appropriate rights merge the history of this and Freeze frame? The page was cut-and-paste moved instead of a proper move. PikaSamus (talk) 17:35, November 6, 2020 (EST)

Maximum hitlag

According to the Smash 4 and Ultimate param spreadsheets the maximum hitlag is 30 frames (I've confirmed in Ultimate at least). Does anyone know if the other games have a max, and if so what it is? Based on the article linked on the main page and it's publishing date I assume it was added in Smash 4, but I'd like someone to confirm. --CanvasK (talk) 12:58, May 12, 2021 (EDT)

I'm fairly certain it's 30 in all games. That being said, I seem to remember that being mentioned on this article at one point, but it isn't there anymore...Alex the Weeb 13:08, May 12, 2021 (EDT)

Turns out it does: "can last as long as half a second" (third paragraph from the bottom under "Formula"). I was looking for 30 frames the entire time and missed it. I checked freeze frame's history and its been worded that way since 2009, so I guess it's been around since at least Brawl. --CanvasK (talk) 13:31, May 12, 2021 (EDT)

Impact of spirits

As pointed out by Meshima a couple weeks ago, using spirits (or having them enabled) have an effect on how long hitlag lasts. Since then I've been testing this and have collected the data in this spreadsheet. Player-count multiplier, staleness, hitlag multipliers, spirit abilities, and spirit series bonus have been noted and avoided to simplify testing and avoid muddying the results; that is I tested as 1v1, no stale, 1.0 hitlag multiplier on all attacks, and spirits used had no 1.1x series bonus and had no abilities or had abilities that didn't affect the moves used.

I haven't worked out what the modified equation is but I have worked out a few other things:

• Hitlag is capped the same as non-spirit battles
• Crouch cancelling reduces the max hitlag to 20; ×0.67 of 30
• Electric multiplier is closer to 1.25 than 1.5 when using spirits
• The standard equation is used until 15 frames of hitlag are sustained, then the modified equation is used (10 frames with CC; ×0.67 of 15)

Using the above findings, I think the standard equation can be written as INT(MIN(30, INT(INT(d×0.65+6)×1.5))×.67) (hitlag multiplier and player-count multiplier are somewhere in there but I didn't test those). I don't have a Twitter account, but if someone wants to pass this on to Meshima or any others then feel free; perhaps someone with access to the code will be able to find values that match. --CanvasK (talk) 08:35, May 19, 2021 (EDT)

Using the data you collected, and substituting some of my own, when plotting the threshold damages we see that the new multiplier is around 0.2, close to a third of its normal value. https://1drv.ms/x/s!Aq5gC8H25VMYdnXmCq5y8aPOBJE Alex the Weeb 08:41, May 20, 2021 (EDT)

I came across a value of 0.2 early on and it was just barely off in some cases. The standard value is 100% accurate and I didn't want "just barely". I'll do some retesting because that early finding was also with early testing so I may not have been as thorough as needed. --CanvasK (talk) 09:45, May 20, 2021 (EDT)

It appears as though there was some confusion on the article itself, and my message on the talk page was intended to demonstrate that the growth rate of hitlag does reduce to around a third of its normal value. We will likely never know the actual value, and in fact there may not even be a specific multiplier param, if the game uses similar methods to the hitstun suppression used for balloon knockback. Unless new information arises, sticking to a rough conversion, i.e. keeping the article in its current state, seems to be the best option. Alex the Weeb 14:16, May 20, 2021 (EDT)

Correct. There wasn't an edit summary so I made a rash assumption, but after a while I realized what you were going for and thinking on it I went back on it. And I agree, I'd rather not add more to that section until a more solid answer is found. --CanvasK (talk) 15:39, May 20, 2021 (EDT)

Cleanup

The "Formula" section, at the very least, is in need of a cleanup. In no particular order:

• "while m is a series of multipliers based on certain factors, including:" indicates that everything below is applied where "m" is. Several issues:
  • "In Melee, the intermediate result is rounded down..." is not part of "m". In fact it happens at the start, the opposite of where "m" is
  • The "player count modifier" in Ultimate is applied at the start, not at the end with "m"
    • The shenanigans with spirits probably isn't applied at the end either
• There is no mention of the order the modifiers are applied in nor if those results have any rounding
  • At the very end of the page it is said that the electric modifier is rounded down, but nothing for crouch cancelling; the order also matters as doing one before the other causes frames to be skipped
• The "player count modifier" portion is lumped together with the unrelated shield modifier
• The formula for Brawl and Smash 4 incorrectly says "+4" when it is "+5"
• "d" is said to be rounded down, but my testing shows it isn't

I've gone and made adjustments in my sandbox and the changes I've made can be seen in this diff. I've made it clear what modifiers are used, where they are, and where rounding is (Instead of using brackets I could instead use "FLOOR(x)" if that is more readable). I've also changed the "player count modifier" table; takes less vertical space, matches tables used on pages like stale-move negation, and is more flexible for future games. The testing I did for this can be found in my spreadsheet linked in the above discussion. I'd like to hear other's thoughts on these changes before I put it into the page. --CanvasK (talk) 18:32, June 9, 2021 (EDT)

Bump. I've updated the page with Smash 64, the spirit formula, shield multiplier, and some notes. --CanvasK (talk) 11:22, June 15, 2021 (EDT)