Luigifan18
Smash Master
- Joined
- Feb 19, 2015
- Messages
- 3,134
- Switch FC
- SW-5577-0969-0868
The strength of primary Spirits in Super Smash Bros. Ultimate is governed by an RPG system where each Spirit's stats are dependent on its level, which ranges from 1 to 99, and this level increases as the Spirit gains experience points. Now, this thread isn't concerned with what a Spirit's attack and defense values mean or how leveling up modifies those values; a thread for that already exists courtesy of
S
snjwffl
. This thread will be about how leveling up happens; how Spirits gain experience points, what determines how many experience points they gain, and how many experience points each level requires.
Spirits gain levels in three major ways; being equipped for battles, being placed facilities, and being fed snacks and cores. Sadly, Smash Ultimate only displays how many experience points are being gained via the third method, and it doesn't show how many experience points a Spirit currently needs to level up except on the manual level-up menu (and even then only in the snack-feeding menu for a single Spirit at a time), so raw data regarding how experience points are gained is largely unavailable, and as far as I'm aware, the community has not conducted any research to uncover that data. I believe we will be able to find that data by either painstakingly testing experience and level gains in various situations or by hacking and reverse-engineering the game's programming (or, in simpler terms, datamining it) to discover the relevant values and algorithms. I unfortunately lack the technical expertise and technology to carry out datamining, and I don't know how well research will work when most of the variables we seek are hidden from us, but I can hopefully contribute to the effort to work out how Spirits work by organizing the community's information in this thread. By doing so, data on how Spirits level up will be readily available to anyone who may find it useful.
While I made this thread without having any actual data, I do have some theorizing and conjecture to offer based on my own experience playing Super Smash Bros. Ultimate. In other words, I can offer semi-plausible generalizations of the experience formulas the game uses. I'll put them in a spoiler tag for two reasons: I go into great detail into my theories and my reasoning for each of them, meaning that my general formula theorizing will take up a lot of space, and while I will be editing this post to include hard data in the future as people come up with it, I want to keep these theories intact. As for why I'd like to keep my rambling theories intact even when actual data becomes available, I have two reasons for that: to preserve the work I put into writing them, and so I can point at them and laugh along with the rest of you once proper research and datamining proves me hilariously wrong.
Spirits gain levels in three major ways; being equipped for battles, being placed facilities, and being fed snacks and cores. Sadly, Smash Ultimate only displays how many experience points are being gained via the third method, and it doesn't show how many experience points a Spirit currently needs to level up except on the manual level-up menu (and even then only in the snack-feeding menu for a single Spirit at a time), so raw data regarding how experience points are gained is largely unavailable, and as far as I'm aware, the community has not conducted any research to uncover that data. I believe we will be able to find that data by either painstakingly testing experience and level gains in various situations or by hacking and reverse-engineering the game's programming (or, in simpler terms, datamining it) to discover the relevant values and algorithms. I unfortunately lack the technical expertise and technology to carry out datamining, and I don't know how well research will work when most of the variables we seek are hidden from us, but I can hopefully contribute to the effort to work out how Spirits work by organizing the community's information in this thread. By doing so, data on how Spirits level up will be readily available to anyone who may find it useful.
While I made this thread without having any actual data, I do have some theorizing and conjecture to offer based on my own experience playing Super Smash Bros. Ultimate. In other words, I can offer semi-plausible generalizations of the experience formulas the game uses. I'll put them in a spoiler tag for two reasons: I go into great detail into my theories and my reasoning for each of them, meaning that my general formula theorizing will take up a lot of space, and while I will be editing this post to include hard data in the future as people come up with it, I want to keep these theories intact. As for why I'd like to keep my rambling theories intact even when actual data becomes available, I have two reasons for that: to preserve the work I put into writing them, and so I can point at them and laugh along with the rest of you once proper research and datamining proves me hilariously wrong.
- General
- Experience point gains do not appear to directly be modified by a Spirit's current level; experience point gains are relatively constant, with the factor making higher levels take more effort to achieve being an exponential growth in the number of experience points required per level.
- Different Spirits appear to require different amounts of experience per level. I have not looked at many specific values, but it feels like lower-ranked Spirits require less experience points to level up than higher-ranked Spirits.
There are likely four separate XP requirement formulas or tables, one for each of the four ranks; I believe that the developers restricted themselves to just four formulas/tables in order to develop the game more efficiently, though it's possible that certain individual Spirits may have gotten tweaks to the formula for their rank to modify their growth rates.The amount of experience a Spirit requires to level up is tied to its base power, its type, and whether or not it is enhanceable; stronger Spirits require more experience to gain levels than weaker ones, Neutral Spirits require less experience per level than Attack, Shield, or Grab Spirits of similar power (the three main types all require equivalent amounts of experience), and enhanceable Spirits need more experience than non-enhanceable Spirits of the same type with similar power.
- Battle
- I believe that experience points gained through battle are determined through formulas that measure the player's performance alongside either the battle's difficulty or its length and scale, depending on whether it was single-player (Spirit Board/Adventure Mode) or multiplayer (Smash), respectively. Single-player puts more emphasis on a battle's difficulty and the player's performance, while multiplayer puts more emphasis on a battle's scale and scope.
- In Adventure Mode and Spirit Board, I believe that the experience formula is derived from the gold/SP reward formula, which itself is derived from the scoring formula for Classic Mode, with the additional factor of the power of the player's team compared to the power of the enemy Spirit (with this additional factor having the side-effect of reducing the experience gained by more powerful (i.e. higher-ranked and higher-leveled) Spirits). As for why I believe the scoring formula for Classic Mode is used here, it makes sense that the development team would reuse scoring and reward formulas for multiple modes, since Ultimate was developed on a tight schedule; using the same functions (like, say, mathematical equations) for multiple purposes would save valuable time on programming and debugging. As for the Classic Mode score formula which the experience/gold/SP reward formulas are based on, I can't claim to know any of the numerical values, nor do I know exactly what the operations in the equation are, but I think I have a solid idea of what the individual variables are. However, I go into so much detail in describing them that I had to put them in a separate list in a spoiler tag. I will just put a brief summary here:
- Damage dealt
- Time taken
- Victory/defeat
- Perfect game
- Difficulty
- Smash is the game's multiplayer mode, so it has no need for intricate scoring. It does, however, need to calculate how much gold is earned for setting up, playing, and finishing a match, and I believe that this is the formula that is piggybacked to decide how much experience each Spirit participating in the match earns. I believe the variables are as follows:
- The number of characters/players that participated. More characters fighting equals more gold and experience, with CPUs, amiibo, and actual people all being worth different amounts. Super Smash Bros. is a party game at heart; it's designed around a group of buddies screwing around in colorful environments. Thus, the game rewards you for having more people participate.
- CPU players aren't worth much gold or experience per participating character, though higher-level CPUs are worth more than lower-level CPUs (since they give the human players more of a challenge). CPU players can equip Spirits, but those Spirits won't gain experience.
- Figure Players are worth a more varied amount of gold/base Spirit experience per participating FP due to having 50 degrees of skill variance rather than 9; since level 1 Figure Players are exceedingly weak and stupid while level 50 Figure Players are very, very capable of outperforming a level 9 CPU if trained well, a low-level Figure Player is worth less gold/experience than a low-level CPU while a high-level Figure Player is worth more gold/experience than a high-level CPU. Figure Players can't equip Spirits; their stat adjustments instead come from consuming Spirits, much like they could consume Equipment in Smash 4.
- Human players are worth more gold per participating player than any CPU or Figure Player. The innate gold/experience amount given per human player is mostly static, since the game has no way of knowing how skilled a human player is ahead of time. However, I assume that it can tell if a "human player" doesn't actually have anyone operating the controller (i.e. no inputs are being made) or if they're intentionally throwing the match (by self-destructing), and will reduce the gold and experience it gives accordingly. Only human players' performance influences the amount of gold the game gives for a match, and only Spirits equipped to human players can earn experience for participating.
- The length of the match as established by the rules (# of stocks per character (if Stock or Stamina rules), amount of stamina per stock (if Stamina rules), raw maximum duration of match in minutes (any kind of ruleset)). The more stocks and stamina each player gets and the more minutes allotted to the match, the more gold and experience it's worth. I wouldn't be surprised if a Stock or Stamina match modifies the amount of gold and experience it gives if it concludes by last-man-standing instead of time out, and I also wouldn't be surprised if time matches weigh time more heavily than stock or stamina matches.
- Victory or defeat. I doubt that this has any influence on the gold earned from a multiplayer match, but the placement of each player may affect the amount of experience gained by the Spirit that player has equipped — though not to anywhere near the degree of influence winning or losing a single-player Spirit Battle has on how much experience a Spirit gets from it.
- The performance of each participating human player. I doubt that this is anywhere near as intricate as it is for single-player; I think the game only cares about KOs scored, falls suffered, and damage dealt, and maybe the other stats shown on the results screen to a minor degree. I suspect that the gold the game gives is determined through some combination of the performance of all human players, but each player's equipped Spirit gets separate amounts of experience based on the performance of the player they were equipped to.
- Other settings in the ruleset (Final Smash Meter, Stage Morph, Underdog Boost, Launch Rate, etc.) may influence the gold and experience given for a match, but I can't say how or to what degree, and I doubt it's anywhere near as significant as anything I went over prior.
- The number of characters/players that participated. More characters fighting equals more gold and experience, with CPUs, amiibo, and actual people all being worth different amounts. Super Smash Bros. is a party game at heart; it's designed around a group of buddies screwing around in colorful environments. Thus, the game rewards you for having more people participate.
- Facilities
- The Gym grants experience points in real time, even while playing other modes, other games, or if the Switch is turned off. I believe Spirits in the Gym gain 2 XP per second, considering that you can sit and watch Spirits' experience meters fill up in the Gym; low-level Spirits level up rapidly, but high-level Spirits will take a long time to level up (and, again, I doubt the experience gains are reduced at high levels). This is modified by the Spirit's facility-compatibility condition — likely 1 XP per second in poor condition or 3 XP per second in good condition. I could easily be wrong about these numbers, as I haven't checked very closely.
- The XP a Spirit earns for being placed in an Explore facility is dependent on which specific facility it participated in and how the session concluded (completion or canceling). The amount of experience a Spirit gains for exploring is the smallest in the Ruins (Toadette's facility), moderate in the Caves (Charlie's facility), and greatest in the Jungle (Linebeck's facility) — directly in proportion to the length of their sessions (2, 6, and 10 hours, respectively). The experience given for an exploration facility may be tied to its reward formula, but I doubt this to be the case, as it only makes sense for the Ruins due to its rewards being weighted towards SP; I have no clue what sort of correlation there could be between how the game gives rewards for a session in the Caves or Jungle and how much XP the participating Spirits get. It is more likely that each of the three facilities is assigned a specific amount of XP it gives for each completed session, and if a session is canceled, the XP given for that session is cut by a ratio of how long the session was supposed to be compared to how long the player actually let it be. There may or may not be an additional withdrawal penalty on top of that (which is probably a flat value, but could be a variable), since a prematurely ended session gives rather little XP even if it was canceled more than halfway before its proper conclusion.
- If Dojos award Spirits experience points for learning a style, the game doesn't tell the player about it at all. I don't think that Dojos give any experience, but I am open to the possibility that they do.
- Feeding
- Snacks are the one source of XP whose values are completely constant, and one of two for which the game outright tells you what the value is; small snacks are worth 250, medium snacks are worth 1000, and large snacks are worth 5000. However, on top of needing Snacks as material, gaining experience in this way costs SP; 8 for a small snack, 32 for a medium snack, and 160 for a large snack. Snacks are by far the most efficient way to gain levels, but if you were saving up SP to shop or summon, a feeding session can and will set you back quite a bit.
- The XP value of a Core is dependent on the rank and level of the Spirit that was dismissed in order to obtain that Core, much like the SP you gain for selling it (actually, I don't think high-level cores are worth more SP, but then again, I've only dismissed low-level Spirits). Additional XP is given for feeding a Spirit a Core whose type matches that of the Spirit, but I find Cores more useful for summoning overall — Cores don't give much XP in comparison to Snacks (but then again, I mostly use Cores for summoning, so I haven't tried using a high-rank Spirit as feed fodder...). Thankfully, since Core feeding occurs through the same menu as Snack feeding, you'll know exactly how much XP a Spirit will get from eating a Core (and how much SP the feeding will cost you, though even high-rank Cores are very cheap to use for feeding) before you go through with using the Core that way. I've also noticed that Support cores tend to be worth a lot of XP, probably due to not being eligible for a type-match boost.
- Damage dealt:
- The player gains points whenever they deal damage to opponents during the match; the more damage the player deals, the more points they score. There's probably some formula forming a direct correlation between how much damage the player deals and how many points they get for it (score equals X damage dealt times Y multiplier), but I can't claim to know what that multiplier is because I think difficulty/intensity is also being factored into the score shown at the top of the screen, and I'm paying considerably more attention to important details like the characters, the stage, and everyone's damage than I am to the numbers at the top of the screen while I play Classic Mode.
- I should also note that the player only scores points for damage the opponents take from the player's own attacks; damage that AI characters deal to each other grants the player nothing, even when the AI characters dealing damage are allied to the player. The player also gains no points when opponents take passive damage, such as flower damage, hoop damage, water damage, or recoil damage; the player also does not score points for damage that the AI player(s) sustain(s) from stage hazards, unclaimed items (such as roaming Bob-ombs), or its own attacks (such as Snake's grenades or C4 going off in his face or an AI character stepping on (or being knocked onto) a Motion Sensor Mine they placed). Granted, there aren't many Spirit battles where the player is given an AI partner or put in a free-for-all; this caveat is more relevant for Classic Mode, where that sort of thing happens more often. Either way, it serves as an incentive for the player to aggressively dive into the fray and put themselves at risk (as opposed to the safer (read: cowardly) strategy of hanging back and allowing their AI partner to handle the fight for them and/or letting the AI opponents rough each other up before closing in for the kill).
I will admit that since all of this conjecture is based on my own gameplay experience, and I haven't played many Classic routes with AI partners in them, I could be wrong about AI allies being outright excluded from the player's score. However, even if the player does get rewarded for the performance of their AI allies, it is likely that the player gets less points for damage dealt by their AI partners than they would for that same amount of damage being dealt by the player's own attacks. The rationale behind this score reduction is the same reason that no points are scored at all for opponents damaging themselves or each other; the scoring system is meant to motivate the player.As a final speculation point on damage scoring, given that Ultimate places a lot of emphasis on Assist Trophies being able to take damage and be defeated, to the point where defeating an Assist Trophy counts as scoring a KO in multiplayer, I would not be terribly surprised if dealing damage to an Assist Trophy is also worth points. However, I would also not be surprised if dealing damage to an Assist Trophy is worth fewer points than dealing that same amount of damage to an opposing AI character, and if damge dealt by AI partners does contribute to the player's score, than AI partners dealing damage to enemy Assist Trophies applies both reductive multipliers to the points gained for that damage. It's also interesting to note that while there are more Classic route rounds than Spirit battles with free-for-alls and team battles where the player has AI partners, the inverse situation is true for Assist Trophies — there are several Spirit battles where the opponent automatically gains an active Assist Trophy on their side, (and a handful where Assist Trophies spawn normally as items), while there are quite a few Classic mode battles where Assist Trophies don't show up and Classic battles with guaranteed-to-appear, automatically-generated enemy Assist Trophies don't exist at all.
- Time taken:
- The amount of time a round lasts is factored into the player's score at the round's conclusion, but I would presume that it's factored in differently based on whether the player wins or loses; if the player wins, they get more points for winning quickly, while if the player loses, they get more points for being defeated slowly. Basically, for a victory, the time bonus is Y base value minus X per second elapsed (Y-(XT)), though it is possible that the game only counts seconds beyond a certain threshold to determine how much to subtract from the baseline time bonus (Y-X(floor(T-R))). For a defeat, the game awards a consolation time bonus of X per second elapsed (XT). The main variable (elapsed seconds) is the same for both cases; the difference is how that variable is applied to the overall equation, based on a separate variable.
- I will admit that I am adding a layer of complexity to my theory on how the player's score is calculated by assuming that the time bonus is calculated differently based on whether the player wins or loses the round, but I have a good reason for assuming that this is the case, as well as the specific difference in how the main variable (elapsed seconds) is applied for each case. Basically, by making the time bonus inversely proportionate to the duration of the round if the player wins, but directly proportionate to the duration of the round of the player loses, the value of the time bonus is reflective of the player's skill for either outcome. For the victorious player, a quick match is indicative of a decisive victory in which the opponent was thoroughly overwhelmed, outperformed, crushed, and dominated. For the losing player, a long match is indicative of tenacity in the face of relentless doom, forcing the opponent to put out an incredible effort and endure a prolonged, painful struggle in order to put them down and claim victory.
- Of course, it should go without saying that under realistic circumstances, a time bonus for being defeated should never be greater than a time bonus for winning, even if the match goes on for several minutes, so each individual second doesn't count very much.
- Victory/defeat:
- If the player loses, their score for the round is divided by some flat value (my guess is that it's cut in half), as punishment for an exceedingly poor performance (even if the player elects to continue, which isn't possible in a Spirit battle). Being defeated and not continuing will end Classic Mode on the spot, denying the player potential points and rewards in the route's remaining battles, but Spirit battles are strictly single-round affairs; the player is simply denied rewards outside of experience for losing.
- I would like to use this bullet point to remind the reader that this is not a thread about Classic Mode; I'm only discussing Classic mode scoring because I think experience gained in Spirit battles (the subject of this thread) is based on the same formula, and losing a Spirit battle grants your equipped spirit considerably less experience than winning, which is why I think defeat imposes a penalty on the round score.
- Perfect game:
- If the player wins the round without taking damage, they get a "Perfect" bonus added to their score after the damage and time score calculations.
- I’m not sure if the perfect bonus is a multiplier to the sum of the damage and time scores, or if it is a separate value or variable added on to them; I’m also not sure if it requires not being hit at all or if it just requires ending the round with 0% damage (or if it can be obtained with more than 0% damage as long as the player did not get hit by any attacks or stage hazards (meaning that all of that damage was either already on the player's character at the start of the round or sustained passively (hoop damage, water damage, flower damage, recoil damage, etc.))).
- Difficulty:
- The difficulty of the round serves as a multiplier applied to all other parts of the formula. I believe that for Classic mode, the difficulty variable is equal to the current Intensity (treating 0.0 as some nonzero number below 0.1, so as not to prevent the player from scoring any points at all), but for Spirit battles, it's either considerably more complicated or considerably simpler.
- It's possible that the difference between the power of the player's Spirit team and the power of the enemy Spirit is used to decide the value of the difficulty value in the score formula, but it's also possible that the power comparison is a separate variable and that the difficulty value is either fixed (for the Spirit Board) or dependent on the trinary Easy/Normal/Hard selection (for Adventure Mode).
- I think the power comparison is a separate calculation used to modify the score (and by extension the battle's rewards) simply because the scoring formula is already quite complicated and I don’t think the developers would want to make any of the equations the game uses to make calculations more difficult than they absolutely need to be — a typical Smash game already makes a few dozen incredibly complex calculations per second. (Seriously, take a look at the knockback formula, then realize that the game runs it every time somebody takes a hit, on top of adding the damage value (which itself has numerous possible multipliers, additions, and subtractions which can be applied to it) of the attack that just landed to the damage value of the struck player, and that it can be running these calculations on several entities at once, and that these entities are incredibly high-polygon 3D models being animated at 60 frames per second. It's amazing that there isn't any lag offline! Considering how much processing the game is doing at any given moment, it makes sense that the developers would want to make the calculations the game has to run as as ergonomic as they can in order to avoid situations where the frame rate might suffer due to processor overload.)
Last edited:
