Why does Execution matter?

What makes execution so special? I remember in Tatsunoko vs Capcom there was a button that you can hold down to perform a super. Why is it bad to map a specials or supers to a button (or at least a less complex execution method)? Why does execution matter?

For one, there’s a lot of things that are extremely difficult to implement without using execution in some way. It’s harder to implement a wide range of moves if you don’t include execution. Street fighter for example has 6 buttons, then even more special moves on top of those buttons thanks to being able to do command inputs.

I was playing Pocket Rumble for the first extended period of time recently with Pictoshark (Shoutouts to how you won’t punish my sweeps, and keep sweeping at the wrong times in 3s) and something that struck me was just how goddamned annoying it was to input crouching normals when I can’t hold downback to block low, then press the button to do say a crouching sweep.

Games like Rising Thunder, Divekick, Fantasy Strike, and Pocket Rumble inherently limit the number of moves they can implement, and the range of more subtle manipulations of those larger actions possible by removing the execution component that is present in the similarly structured games around them. When you insist, no command inputs, it means you simply cannot implement 5 normals and 5 special moves, one of which has 2 versions, another has 4 versions with 4 followups.
It means you can’t have versions of the same special move differentiated by which button you press, like ryu’s slow, medium, and fast fireballs, or Oro/Urien/Gouken’s differently angled fireballs (Rising Thunder got around this with crow’s fireball toss, by having you press a direction during it, but that’s mildly tight timing, which still requires execution). Without execution you also don’t get cool stuff like Slayer’s backdash cancels (BDC). For example, the input on Slayer’s BDC Bite can be 6321447H or 44632147H. 44 is the point where you backdash, it can be at the beginning or end of the input. The first gives you a near-instant bite with little movement, the second gives you more invincibility frames and moves you backwards. They are applicable in different situations and should be treated like different moves.

Something else I have pointed out is that there is no way to implement wavedashing in smash bros that keeps all the current nuances of the move without it being difficult to execute. The closest thing you can do is have jumpsquat cancel directly into wavedash, but this still takes moderate execution skill to perform. As-is right now, wavedashing has an effect proportional to the time you airdodge off the ground, and the angle you dodge into the ground, allowing you to move more or less distance, and that gradient in distance can be important. Wavedash in place is a fast way to get out of shield or dash dance. To get this fine-grain level of tuning, you need to have the analog stick factor into the movement, it cannot be a single button or button combination. Not to mention Wavelanding and triangle jumping, which necessitates that airdodge is its own button to make all these nuanced possibilities available.

The other thing is, Execution is something that allows players to distinguish themselves. Having an execution gradient across a game allows players early on to play a game one way, then dig deeper and find out it’s actually played a different way. Then dig even deeper and find new ways to play. And you continually integrate these new ways with the old one and ideally end up with a more rich and varied game that stays fresh no matter how long you play it, and no matter how good you get at it.

SFV has kind of a setback there compared to SFIV. SFIV had really hard combos, and a combo system that allowed for a massive amount of intepretation. This lead to a lot of degenerate tendencies, which is unfortunate, but in the clean-up job for SFV, they ended up seriously limiting the potential of the combo system in regulating which moves are allowed to start juggles, continue juggles, and the gravity limitations on juggles. They started learning their lesson more around Urien/Ibuki/Juri, and they have a larger number of potential combos (but Ibuki and Juri have lame damage output and suck in other ways) however the damage was largely done by that point. So in tournaments we see a lot of top players doing the same combos we do at home and not varying them that much between each other, and that’s really boring. SFV improved its neutral and pressure games in many ways because of these changes, but it’s still a letdown overall.

It’s disingenuous to assume (as Sirlin typically does) that players in tournament will execute every single advanced technique 100% of the time, so therefore they should be made so easy that everyone gets 100% consistency. It’s like saying that every tournament player will get the 100% most optimal punish off any given hit like it’s a combo video. Not only do flubbed inputs happen, but players differentiate themselves through which combos off which starters they’ve mastered, as well as movement tech, specific setups, and so on. This is a part of why players each play the game with their own style. Being able to figure out what your opponent is weak at and exploiting that weakness is a standard part of competitive play and it doesn’t vanish at the highest level, because no player has perfectly balanced mastery of everything. Maybe one player is really good at okizeme setups, maybe a another has a strong neutral game. It’s not just a matter of flubbing X% of the time, it’s also a matter of confidence. Speedrunners have this issue too, they don’t always feel confident about a trick, sometimes they feel more confident, and they can judge to a degree whether they’ll get it that time.

Additionally, in fighting games and other games, many inputs are specifically designed to be balanced based on their ease of execution. The 360 input with Zangief is a classic example. The light punch version in SFV has more range than many pokes, and deals 180 damage, which is more than some weaker combos, like Ryu’s c.MK xx Fireball. His HP version does 240 damage, which is more than most up close combos. And they both have only 5 startup, and are unblockable. They’re balanced because in order to use them, Zangief needs to input that 360 motion, which takes real time. The fact that it takes real time not only adds something similar to startup, but it constricts where he’s allowed to do it to avoid startup. He can buffer that motion during other attacks to conceal the startup, which changes where you’ll be watching out for him trying to command throw.

Guile and other charge characters are balanced on a similar basis. Their charge moves are given more power compared to normal special moves because they have that charge limitation on them. There’s a relationship between the power of the move and execution required to pull it off. When you have a relationship like this and it is proportional to the effectiveness of the move, then players feel this natural reward for pulling off hard moves.

L canceling in Melee is another example, as is Parrying in 3rd Strike. These mechanics exclusively work and are exclusively strategic because they are hard. If the window is wider, then the need to read the situation and correct difference in timing is completely removed. You wouldn’t need to read hit/whiff/shield. You wouldn’t need to read when they’ll attack, and how many times.

Put simply, people enjoy doing things that are hard to do and provide a reward for accomplishing them. People enjoy improving in consistency at things. This is the root of where fun comes from (in terms of what actually triggers dopamine release, favorable outcomes from typically inconsistent outcomes). The real design question shouldn’t be whether such mechanically difficult things are included, it should be “how hard is too hard?” “How good is too good?” The answer to that isn’t the same for every game, every move, or every character.

I think that to make a good game, you need to have some things that are harder than others, some things easier. There needs to be a skill gradient that players work their way up. Give players room to specialize. Give them room to learn. And in many cases, it just plain feels better kinaesthetically to perform this tight difficult motion.

Sirlin sees all the value in Daigo knowing Justin would use his Super. However if anyone could do that, if simply predicting the Super was enough, then events like that would be commonplace and boring. Chun Li’s super would be significantly weaker. Of course Sirlin is also on the 3s parry hatetrain, so whatever.

In fighters like Guilty Gear Xrd, you have Ramlethal whose dauro move turns green and does more damage if you just-frame the button input. In Marvel 3, a ton of supers do more damage if you mash them. In SF3 and SF2, a ton of pummel throws do more damage if you mash them. In nearly all 2D fighters there’s safe jumps which let you get in a meaty attack without risk of retaliation. If you’re going to do a crouching medium kick in street fighter with Ryu, there’s absolutely no downside to always buffering in a fireball every time. Reward on hit or block, no penalty on miss, no penalty for messing up the input. If you land a tatsu with akuma, you’d be dumb to not throw in a followup, because either you hit them for more damage, or you miss and they’re knocked down anyway. And should buffering into hadouken off a low short be automatic? Should there be a safejump command that does it for you? For that matter, why have inputs more complicated than quarter circles at all?

Yeah, it is a matter of feeling good, we play fighting games for the challenge, in part the challenge against the opponent, and against ourselves. People enjoy pulling off difficult combos, some more than others. Fun is variable success, in all its forms, head to head or individual. Good fighting games are a mix of both, not allowing one to destroy the other. Having that tight link between input and response is a good feeling, we shouldn’t toss it out just because it’s arbitrary from a decision standpoint, because it’s not arbitrary to enjoyment of the game.

Many people feel that fighting games should be these high-level competitive pure decision-making exercises, but these games are played in realtime. There is always a dexterity element, always a timing and reaction element simply because they’re in realtime at all.

People enjoy the process of working to hone their skills and pick up new techniques.

Sure combos have tradeoffs, but if someone wanted to they could easily substitute a combo system for a menu that gives you a few different pre-strung combos with meter/damage/positioning/knockdown tradeoffs between them, and vary these based on the starting move. But that’s obviously not as interesting.

You could have it so multishines are automatic on people’s shields as long as you hold down B in smash bros, but not only would that not be nearly as hype as someone making the 1 frame link continuously, but it would make the multishine option outright broken, beating shields effortlessly.

The benefit of having an executive barrier is also that players are all at different levels of execution from one another, people work in the lab to stretch just a bit further than the day before. The combination of reads, understanding the game, and execution are part of what make fighting games interesting. If you remove that factor, then you won’t end up with as interesting or as deep a game, because the depth of these games is very deeply tied to their realtime execution element.

The thing to learn from games like rising thunder is, no, there wouldn’t be more high level players. Execution barriers don’t hold people back from being high level players, you’ll still get your ass kicked by the same people who were better than you before. The thing to learn from games like Smash 4 and Brawl is, if you remove the execution barriers and all the tricks that people have to master using execution skill, you don’t end up with more people engaging in that high level play. Instead you end up with nobody being consistent enough at the game to form a high level playerbase in the first place. The top players shift around and nobody has that space to really develop themselves over others at all.

One answer is because then players will do combos that produce damage proportional to their skill, if it’s easy then people will either all do the same combo, which is boring, or combos will get ridiculously long at higher skill levels.

Another answer is because it’s more direct control over the character. The sensation of direct control, especially in hitting tight frame windows, feels really nice. It’s related to “Game Feel” or Kinaesthetics. If you have everything cancel into everything, or stick huge buffer periods on everything, or input shortcuts, and so on, then things feel really loose and like the response to your inputs is sloppy or delayed. Like compare trying to dragon punch on wakeup in SF2 or 3, to SF4. You have a 1 or 2 frame window versus a 5 frame window with input shortcuts. In SF4 you can just mash reversals, it doesn’t feel like a tight refined input, it feels like you’re slobbering on the stick practically, where in SF2 or SF3, you gotta do exactly the right motion at exactly the right time, and that feels really tight by comparison.

And the last answer is because, pulling off hard stuff, even if it’s small hard stuff like needing to link instead of cancel attacks, is really fun. The neurochemical phenomena of fun is essentially being able to profit off finding patterns in inconsistent phenomena. By having things be hard, by having a lot of hard things, people always have new things to find patterns in, to get consistent at. So games take longer to master and not everyone gets stuck in the hellscape of being consistent at everything except winning tournaments. Sure we can get these high-minded, “everything should be in the decisionmaking phase” ideals, but if that’s the case, then the game gets solved a lot faster, because people don’t discover new ways to use the characters as much, and people get bored when they learn how to follow the optimal set of decisions for their characters. Yeah, everyone can be a “high level competitor”, but being one matters less, and nobody can actually stand out from the crowd.

Games don’t inevitably become solved, they inevitably get closer to being solved, but there’s a ton of unsolved games, and skullgirls (along with nearly all fighting games except like divekick and smash 4) is one of them. Making games difficult to solve, having them be complex enough that people can’t just pick it up and play in the most optimal way possible, is related to what makes games interesting in the first place. Games are ostensibly about improving consistency. About learning things, picking up on patterns, and putting it back into your performance. Making a game complex and difficult to solve is a part of that (though it still has to be a good game underneath). If everyone can solve the game, if everyone can play optimally, like in tic tac toe, then the game straight up dies. The process of being able to pursue higher skill continually is what keeps games interesting for high level competitors (and low level ones too). It’s about the spirit of continual improvement, of always having some new way to stretch out.

However if you can simplify everything down to the optimal punishes, the optimal options, the optimal decision trees, then the game ends up getting repetitive, because you know it all already. You’re no longer improving. The fun cycle of picking up on something then steadily improving in consistency until you have it down stops. The draw of a truly deep game is being really far away from knowing everything about the game, the draw of always having something new to master or work on, and decisionmaking isn’t the only skill in the world, nor the only one that competitive games should ever test. We shouldn’t segregate competitions between pure execution and pure decisionmaking, mixing the two is great, and allows them to recombine in ways that test both of them to even greater levels.

I’ll admit that L canceling doesn’t add a lot to the game, it’s not a very deep mechanic compared to others. You could just halve aerial landing lag and not lose very much of the interesting elements that make smash what it is. The other thing is, you could disable L canceling for me, have it be automatic for yourself, and I would still probably kick your ass (unless you’re like a power ranked player in your region or something). I’m sorry that this sounds like bragging, but L canceling isn’t everything. There’s a lot of other strategic aspects of the game, and people who are bad at L cancels right now can get good at those other things, can outperform their opponents without using L cancels. L cancels give people a skill to expand out into and become more consistent at to get an advantage over their opponents. L cancels require people to recognize if they’ll whiff, hit, or hit shield. L cancels make it so sometimes people are vulnerable and sometimes they’re not, which you can exploit.

Meteor Cancels have nothing to do with reading or playing against your opponent either, they’re purely a reaction/timing test, press jump 18 frames after being hit, don’t press too soon or you die. Reaction Tech Chasing too. There’s TONS of elements like this in Smash bros that are pure optimization without any element your opponent can interfere in. And that’s alright. It’s alright to let people work to become more consistent at an element for a pure statistical advantage in certain scenarios. Not everything should be made flat-out automatic, or you lose that decision-making aspect of, “my opponent is weaker execution-wise in this area, I can exploit this” If I know my opponent can’t L cancel consistently, then I can abuse them in different ways than my opponents who can. If I know they can’t do the mew2king angle, then I can cover different recovery options without fear. If I know my ice climbers opponent doesn’t wobble often, I can be a bit less afraid of getting grabbed. Even though all these things are pure executional things that don’t involve your opponent, they still affect the way you strategize and can adapt to your specific opponent.

It’s nice to have games like Dive Kick and Rising Thunder where it is easy to execute. It is nice to have variety in the fighting games available to people so that people who are not as skilled have something to pick up so they can learn, and which also explore the unique strategic and design spaces those low execution barriers offer. However not all games should be this way, and we shouldn’t aim deliberately to make future games closer to this standard, we should aim to more or less stay the course, developing games that replicate the successes of the original ones and try new things so we can have a diverse environment.

What do you think of execution barriers for command moves in single player beat em ups like TW101’s Wonder Liner? Are they necessary to keep the game balanced and less spammy or does it not matter because you’re just pitted against AI?

Wonder liner is designed the way it is so you can perform a bunch of different functions with the same mechanic. So you can circle around people or objects on the ground, and so you can select different unite morphs. But it’s not just selecting unite morphs, it’s selecting between several different ones AND how big they are from your pool of wonderful 1s.

It’s also there because drawing shapes is more fun than tabbing through a menu, even though it’s slower.

Like yeah, you don’t need to worry about balance in that way versus AI. The player can’t really be overpowered or underpowered, and something like this isn’t necessary to balance the player’s options against the AI. You could balance morphs relative to each other based on how difficult it is to input in that system, and W101 chose to do that by making Sword one of the weaker unite morphs, but it’s not the biggest deal in the world for a single player game. If you could instantly pick any unite morph, then the game wouldn’t be significantly different than it is, but it would lose out on the multifunctionality of the wonder liner, and it would be harder to specify the size of the unite morph (which is important because that’s resources for the multi unite morph), and it would be slightlly less interesting than getting good at drawing shapes.

If you believe crazy david sirlin logic, then singleplayer games are the only place where arbitrary execution barriers make sense, since it’s not interfering with the raw decision-making that multiplayer games should be about.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s