Methodical approach to experimenting in games?

I was curious if anyone can share any insights on having a methodical approach to experimenting and improving in games.

The way I see it, most average players (including myself) just play and improve mostly via imitation/copying and random trial and error but there has to be people out there who have a systematic approach… Perhaps the ones that innovate and become the one that other people copy and imitate.

Preferably I’d like to hear from people who can get consistently good at any game they approach regardless of genre AND know how to explain how they did it (as opposed to just doing it by “feel” or “intuition”). Or even from people who play the same game (say Street Fighter), but can pick up new characters quickly and systematically.

Back some time ago when swapping around with random characters I used Akuma for awhile, this guy who was also using Akuma sent me an invite and after a few matches he improved greatly. You can always send invites to players you consider good online, though they’ll only accept if they like you.

Take a character and figure out concepts of things that you think may work, or might be good/broken, then test those out. Basically, if you think that something might be good or broken if a character has it, test it out.

Thanks for the feedback guys. I was hoping for something a little more in-depth, however. For example… These are some of the ways I experiment right now when I pick up new characters in SF4

  • Experiment: try to win matches using only normals

    • why?
      • learn the ins and outs of your normals
      • learn to defend yourself using only normals (specially anti-air)
      • learn how to control spacing with normals
      • normals are generally safer on block/whiff than specials
  • Experiment: try to throw opponents as many times as possible

    • why?
      • learn surprising ways when throw attempts actually works
      • an effective ground game isn’t complete without knowing how to throw effectively
      • if you know when to throw, you will also know when opponent wants to throw you. improves throw techs as well.
  • Experiment: try to win matches mostly maintaining a position right under the opponent’s jump-forward landing

    • why?
      • shuts down most lower-level players who don’t know how to “get in” without jumping at you
      • learn how to use your movement as well as the opponents to control spacing
      • learn basic zoning skills than can be built upon


I’ve seen people advocate this sort of segregated approach, but I don’t personally think there’s much value in it. A well rounded player gets good by using all the tools at his disposal in tandem not building up individual skills and then piecing them together. Focusing on certain areas is one thing (I get jumped in on a lot - I should focus on anti-airs, I get frame trapped a lot - I should focus on patience) but I don’t feel it should be used to the detriment of your overall game.

My thoughts on being able to approach every game is to take a mechanical approach. ALL FGs run on certain concepts (even between 2d - 3d, although the gulf is somewhat larger than 2d - 2d : 3d - 3d)

For instance, frame advantage is good. How can I get frame advantage, does a move give it to me? Or maybe an assist? Once I have it, what sorts of mixups or pressure or whatever else to benefit my game can I apply? Which ones are safe? How far away am I from the opponent after landing it? People who are truly innovative understand these nuances very well and can apply the rules and numbers to develop their ‘tech’.

This goes for any competitive thing really. Once you know the rules you can change the game.

Airdash Academy has some good high level ideas of how to approach getting better at Fighting games. I’m not up to date, but the first ones I watched are really good, and now that I’ve remembered, I’ll be sure to watch the rest.

Specifically, the thing I remember is him suggesting is…

  1. play long sets
  2. pay attention to why you lose
  3. spend time in training mode attempting to find a solution, or watching videos of people playing those matchups looking for how other people deal with those situations
  4. play long sets attempting to implement what you’ve found.
  5. repeat step 2.

The difference between this and just watching videos and stealing cool shit is twofold, first, you’re not just learning flashy shit because it looks cool, you’re specifically getting better at the things you have trouble with, and secondly, because you have something specific goal, you’re more likely to get more out of the training mode time, and more likely to be paying attention to the matches. Note that getting videos of your matches is really helpful if you’re doing this.

Although one thing they don’t mention is you need to pay attention to root causes, not necessarily what causes you damage. Specific example, don’t pay too much attention to execution mistakes, you should always be working on your execution, but just because your last health points are lost to somebody uppercutting through a dropped combo doesn’t mean you lost because of execution, you lost 800/900 damage before that, and you failed to do enough damage to kill your opponent. Look at how you got into that situation, how you took damage, and where you missed damage (which is admittedly much harder) similarly, if you die failing to block a Cammy knockdown mixup don’t just go “Fuck Cammy, how do I block this shit” you need to look at how she got you into that situation in the first place (often the answer is “get better at anti-airing”).

Also, just talk to other players, play other people and ask questions like “what am I doing wrong?” or “what do you get away with vs me that you can’t against better players?”. Focus on what they feel they can do, not necessarily how to stop it, because you want to look up and research that yourself, there could be better answers than what they’ve seen, you might not have the reactions/execution to do what other players do, or you may have better reactions/execution and be able to do something better.

This is most helpful for defense and some neutral game stuff, getting better at opening people up and playing aggressive footsies is different, I don’t have a good methodology to suggest. Mostly I just read Maj’s footsie handbook every 6 months, and watch Valle/other old school footsie based players too much.

Always starrt with footsies, spacing and zoning, as these cross over quite well.
after this, look at how the game/characters are structured; how it prefers to link one move to another. From here, you can learn how to construct and extend combos yourself, rather than simply copying a combo from the pros or from trials.
spend a bunch of time doing this in practice mode, paying attention to how much damage you do, how safe the combo is to initiate, and whether there is any space for your opponent to recover or blockstring (do this by selecting block/block all in practice). Bare in mind that the combos you’re constructing can miss on characters with odd hitboxes, particularly if there are crossups involved, so be sure to try them out on different opponents, from odd distances and in different scenarios.
lastly, once you’ve got the rest down, look at how players are combatting you, especially in matches lost, and try to come up with what you’ll do differently next time. Its easy to give in to your ego and think ‘wow he got lucky’ or ‘that was just a one off’ but chances are that if one person finds a way in, other people will find the same way, so come up with a solution, rather than passing it up as a coincidence.

I have to strongly disagree with this, not only is this the right way to go in fighting games but in life. you have to develop the tools before you can use them, and to develope that tool you need to focus on it

I never realized the responses I had gotten on this thread. 2 years later I am still interested in the process of learning how to learn faster. I’ll go ahead and respond to some of the comments here

I think this is great advice however it’s something I imagine isn’t easy for most people to realize. If other noobies have had the experience I did, It’s easy to get caught up in thinking you lost to “obvious thing A”, when in reality you lost to root causes concerning “subtle things B and C” .

The problem I think though is, as a noobie, it’s hard to “pay attention” to the root causes if it’s so subtle that you may not even be aware it exists!

As a concrete example, one of the things I used to lose to a lot (and will still lose to if the player is good enough) is mid-range zoning. At the time, I thought I knew what zoning was and would recognize it if I saw it, but I had misinterpreted that it only meant “keep away”, which is not quite the type of zoning I was losing to. I couldn’t figure out for the life of me why my character felt pretty much useless and it took me a while to figure out that there are subtle positional points for every character (even when close to the opponent) where all their options are inferior.

It actually wasn’t until I did this that I figured out I was missing a certain fundamental of space and control. I’ve found that simply talking to other people is a really fast way learn and improve. The problem with that though is that communication itself is a skill, and not everyone that’s good at something knows how to explain what it is they are doing to be that good.

This is what I feel the goal should be and I would still like to hear more thoughts on this matter. How do people go about learning and deconstructing the game’s world that they can construct/invent advantageous tactics, exploits, strategies, etc. on their own?

I like how I’m total garbage at fighting games but already utilized the advice given in the airdasher tutorial and give it to other people as well.
I always advocate learned moving, poking and anti airing before trying to learn combos and shit.
You can methodically work your way from your basic arsenal.

Say you learning how to poke with Ryu’s crouching medium kick in SF4 and now you play against someone who focus attacks that and wrecks you.
Now most people are like “Well I dropped combos and didn’t do enough damage.” but that’s not the cause of your failure when you got hit by 5 focus attack crumples during a match.
Your failure comes from the fact that those were landed in the first place and your response should be to go to training mode and find out how to be ready for that and how to counteract it.

If you went to training mode and realized that the low forward cancelled into a buffered shoryuken beats that option then you made progress and conditioned yourself to watch out and act accordingly to that scenario. Now you maybe meet a player who absorbs the and backdashes making the uppercut whiff or only uses focus where srk would whiff and you’d have another problem to find a solution for and on top of that you’re now playing mind games with your opponent reading their habits and making educated guesses on which option the opponent will most likely choose etc.

Bit by bit you become a better player without realizing it.
Myself and other players probably would suck a whole lot less if they didn’t just practice combos in practice mode but rather look for solutions to specific problems and work on applying them.

Ah thanks. I quoted a lot from robotic elf but I didn’t realize I didn’t actually take the time to watch the videos that he linked.

I found #1, #3, #6 and #7 very informative and is exactly the type of information I was looking for. It goes over a methodical framework for constructing skills, coming up with solutions, and iteratively learning. Good stuff.