How to Review Your Matches
This is something that I've noticed a lot of newer and even intermediate or advanced players struggle with a lot, and so I thought I would make a quick doc/list on how I personally break down reviewing my (or anyone else's) matches, and do learning from them. All of this is a built practiced skill from playing this game for 3+ years being the only HoS player in the US, so its one that I have learned and evolved by necessity. This is not the end all be all guide to match review, nor do I present this as the "best" way, this is just what I personally do and how I think about it. This guide is also very abstract and rambling and not very fact based, I apologize, its very hard to externalize these thoughts as its not something I've really had to teach others about in a long time.
Step 1: Have a purpose
Guilty Gear is a very complex, nuanced game, and goes very fast, so you need a way to limit the scope of what you are looking at and for. The way to do that is to have a goal, or goals, to look for and at specifically when you go into the match review. These can be as vague as "Why do I feel like I never have the life lead?" or "Why can't I pin down my opponent" or as complex and specific as "Why do I keep losing exchanges after blocking Ky Lightning Sphere and what can I do about it?", but the important part is to have one or more goals so that you can narrow the scope of the match down to manageable levels. This is also something a lot of newer players can have trouble with, after all, its hard to know what you don't know. If you are stuck for a topic, you can always use generic things that are always good to study and can serve as a springboard into different specific goals, such as "Where did I get hit and why?", "Could I have used my meter better?", "Did I give up pressure or advantage when I didn't need to during the set, and if so, why or what can I do instead?", "What was my locational heatmap throughout the set, and do I like the kinds of positions I ended up in the most?", ect.
Step 2: Determine How you are Viewing
Match reviews can be of any length, from looking over a single Bo3, to watching a marathon 2 hour set. When approaching a match analysis, determining how you want to review the match can add a lot to what you get out of it. How you choose to watch the match generally can and should change with your goals, but there are some easy ways to break it up. If you are watching a set with a goal that is set long and interaction agnostic, such as "What was the general flow of life lead in the match?", "What did my meter progress look like over each game?" or "What was my movement like over the set?", you should watch the match at 2x speed. This is because (ideally) your specific focus doesn't really care about the specifics of the match, but instead should be a general overview, and by playing the match at 2x speed, your brain has a much easier time ignoring all the minor details and instead focusing on the overall patters of the set, seeing a much faster, prolonged concept view of the overall state of the match and set. If your focus deals with adaptation or changes over a period of time in the set, such as "Why did my strategy at the start of the match stop working midway through?", "What kinds of adaptations did I make in the set and were they the right ones?" or "Why did the match seem to get so much easier/harder at X time?", its good to break the match up into segments to watch. Watching the start of a set for a couple games then jumping to the middle gives a large gap in-between the play just witnessed and the current play, and can help highlight evolutions in the metagame of the set that might have happened gradually over the set, but become sudden shifts when jumped to. Generally for these kinds of goals, I study the first 2/3 games to establish what the metagame stated as, skip to the middle to see if any adaptations have been made on either side or if the meta has shifted, and then skip to the end to see what the meta finalized as, and think and analyze the path it went through to get there. Finally, if your goal is specific interactions, you should fast forward through the match until you get those spots, and start the analysis there. If your questions are general things that happen every game ("What were my meter spend habits?" "Why did my opponent get away so often?", ect) those are ones where you should watch at normal speed, and break each game into its own segment. Watch the game as a whole at normal speed and take notes on points you want to go back to review, but just keep going. Then, once you have your impressions and notes for the game as a whole, go back into the game and stop by the spots you noticed and do a spot analysis there for each.
Step 3: What, How, Why
Once you are in a situation to start analyzing, there should be 3 main things you focus on: What, How, and Why. To break each of these up: "What" refers to the physical, actual aspect of what you are looking at, ie: what happened. "What" should be able to break down the situation at hand, and should answer the immediate or surface answer of the target scope. This is the first step to the bigger picture of the interaction, just identifying exactly what was going on or happening. Since I know I'm not really able to explain this part well, I've included some examples, in brackets [Scope: "Why can I not pin down my opponent" - What: several times when pressuring in the corner, I would go for redashes into 5K, and my opponent would get out during these points, Scope: How do I keep DI in range for a threat after a combo at lv1? - What: After combos ending in SV, DI gets a free tech, I need to find a way to keep threat here because this was the place she got away the most, ect]. Once you have identified the "What", you can move onto the "How", which is how the what came about, or how the interaction played itself out. If the "What" is the physical aspect of the analysis, the "How" is the inner workings of it, why an interaction resolved the way it did, what caused the situation to unfold the way that it did, and so on. This is often the meat of the analysis, and where you will get to the core issues. For example, if your scope was "How do people keep disrespecting me", and you found your what was "I am getting mashed on with certain moves that keep working," the "How" here would be mechanically, how do those moves keep beating you, and how does your opponent keep getting away with them? Finally, there is the "Why" of a given analysis, which is informed by the meta context. The "Why" is an overarching analysis of why the situation even came up in the first place, taking into account the events leading up to the situation or that caused the situation, why it might have kept occurring, why the situation played out like it did, ect. This part of the analysis should take into account not just the immediate events, but the flow of the match as a whole, what has happened up to that point on both sides of the set, and generally why the situation got to be what it is. As might be obvious, there can be several points of analysis depending on your scope and how long what you are reviewing is, so you will almost certainly go through this multiple times. As you hit more and more points of analysis, a broader picture should start to form as an answer to your question of scope, and you can go onto the next step, solving it concretely.
Step 4: Test Your Conclusions and Cross Reference
Once you have finished your analysis, the next step is to figure out what to do about your overall question or scope. Based on the analysis you have done, you should have plenty of ideas now about what to do regarding the question or reason you started this analysis, so take some time to write them down. Then, go into training mode or a replay or even other more skilled player's match footage, and test your conclusions. Jump into the discord and ask your peers about your ideas or your conclusions and see what they have to say. If your ideas were mechanical interactions to solve certain situations, try them out in training mode and see what they handle and if they work, and what the opponent can do about them. If it was a general issue or broad scope, do some theory fighter about and see how it might play out, or look for the situations or overall scope in another's match videos and how they approach it, and how their own answers work out for them. In short, this step is just about testing your ideas and making sure that they are solid and robust, and checking them against other resources. This process will both refine and sharpen any ideas you have, get rid of ones that seem promising but turn out to be half baked, and even open you up to new possibilities or ways of seeing the game that you hadn't even considered.
Step 5: Make an Actionable Plan The last, and arguably most important step of this process, is to make an actionable plan off of your conclusions. Once you have a firm answer or answers for your original question or scope, you should come up with a plan to implement it and repeatedly try it. This can be as easy as having found an answer to a situation that was giving you trouble, and just having a plan of "When I see X situation or feel it coming, I am going to do Y" or as complex as "I am going to try to maintain X distance from Y character for the whole set, and not do A, B, and C movement patterns", but it should be something that you can consciously and discretely work on, and see how it works. The goal of this step should be to make your conclusion second nature to your normal gameplan, so you no longer have to think about it, and have grown your game by at least one level, having evolved past your original issue, or at least one aspect of it.