Learning SFV-AE: Rookie to Bronze

When Street Fighter V: Arcade Edition was released late last month, I decided I’d finally take the plunge and start playing it online. I did some research on how people tend to practice, and stumbled upon several resources that have helped me a lot thus far.

My training regimen consists of 3 parts:

1) Practice Mode

Joe Munday’s Gief’s Gym is an amazing place to find practice mode workouts ranging from the most basic topics to really advanced techniques. The entire series is available online in the reddit thread linked above, but please consider supporting the author by buying the paperback version on amazon.

In the first part of my SFV training I’ve been focusing exclusively on the first 4 workouts of Gief’s Gym

Below you can find one example of what drilling normals looks like:

You can see how to set up the anti-air practice, as well as 3 basic drills below:

I added another wokrout, namely dashing, so I could get used to doing this instead of jumping to close gaps. After reviewing my replays from the past month, I’m surprised at how difficult it is to consistently apply it in my matches.

2) Casual Matches

While looking at some general tutorial videos on YouTube, I came across the following advice about playing online:

If you want to improve, you need an environment in which you can experiment and try new things without being scared you’ll be punished and lose LP for that. And while some people manage to keep this open attitude while playing ranked, for many players casual matches provide the breathing room to try something new.

So, if you want to improve, you should play a lot of casual matches, as those will pair you up with other Rookie and Bronze league players and give you the opportunity to practice against players of your level. Once you find you’re consistently beating them, it’s time to jump back into ranked.

There is one caveat, though. Casual matches at the very bottom of the ladder (0LP) are a place where you could get a good match, but are just as likely to be mercilessly slaughtered by 3000+ matches and 95%+ winrate smurfs. If you want to lower your chances of being matched with smurfs, the best thing to do is to struggle a bit in ranked until you get to at least 50-100LP (that’s just 1 or 2 wins!) before grinding out practice sessions in casuals again. This seemed to drastically lower the amount of smurfs I encountered (and echoes the advice I read online).

3) Ranked Matches

A reddit user named Truen1ght posted a series of ‘Tips for Players in Rookie, Bronze, and Silver‘. I highly recommend you go save that nugget of wisdom in a notepad file somewhere. As I had absolutely no prior experience with fighting games, I decided to follow his advice on how to approach ranked play to the letter. Here’s what he has to say about working your way up through Rookie League:

Rookie to Bronze

To get out of here, you don’t need much. You need an anti air to deal with front jumpins, an air to air to deal with crossups, and a fast combo to punish the things your opponent does.

  • Learn 1 anti air and its followup. For Akuma, this would be : b ST HP xx H Red Fireball. It’s super reliable and does decent damage. 8-10 of these and you win a round.
  • Learn 1 air to air, and its followup. For Akuma, he doesn’t have a good one, so jump back MP is ok, as it does some damage and comes out relatively quickly. There is no followup for Akuma’s air to air.
  • Learn 1 fast combo, by that I mean a combo that starts with a light attack. When your opponent does something like a heavy special that you block, use it. For Akuma this would be : ST LP > ST LP xx L Tatsu > H Goshoryuken.

Now for your strategy. Just play it safe. Walk, and block. If your opponent jumps, just anti air or air to air him : do that enough times and you win the round outright. Don’t jump unless the opponent is spamming fireballs, in which case you jump in HK or HP, and then use your light combo to finish it off. Let your opponent kill himself, block everything he does, and only punish. Don’t try to start any offense yourself unless the timer is running out (30 seconds or less). Let him do his random attacks and specials, and just punish after blocking them.

I would always do my practice workouts first, warm up some more in casual mode and only move to ranked when I was feeling good in the casual matches. On days were I was getting consistently outplayed, I just skipped ranked play and focused on executing Truen1ght’s strategy better in casual matches.


I’ve currently played a total of 109 matches (20SW/17W/72L or 33,94% winrate). Of those, 84 matches were casual (15SW/11W/58L or 30,95% winrate) and 25 were ranked (5SW/6W/14L or 44% winrate). After match 107, I made it to Bronze league, where I’m currently sitting at 567LP.

Apart from the obvious lesson here (You’ll lose. A lot!), those numbers aren’t all that important. I’ve found that the ranked strategy of trying to block everything and anti-airing jump-ins works quite well on other Rookie players. You can see what that looks like here:

Because of the way LP works though (the reward for winning is a lot higher than the penalty for losing), just being able to deal with other Rookie players is enough to propel you upwards to Bronze league.

The next step

Looking at my replays I noticed quite a few things to work on.

I have a lot of trouble with two types of players. The first ones are people who use a similar defensive strategy as mine, but are consistently better at blocking and anti-airing. While I managed to take the first game, this Nash adapted quickly and completely shut me down in game 2 and 3:

I’m not sure on how to approach this type of opponent, and I imagine I’ll continue to struggle in these matchups until at least Silver league, when I start expanding my neutral game options.

The second scenario in which I tend to struggle is when playing opponents that just rush me down and take advantage of inconsistent blocking (I push too many buttons or step forward at odd times) and late anti-airs.

I guess the answer to this second type of players is simply practising more and improving my block and anti-air consistency.

Other bad habits I should work on:

  • I miss most opportunities to tech throws (especially after an opponent aggressively dashes up to me or pressures a lot).
  • Most of the super combos I launch are accidental, and as a result I almost never hit confirm them first, wasting lots of meter and leaving me vulnerable to counterattack.
  • I have a reliance on sweeping after the c.HP anti-air. I should look for a better combo to punish after I hit an anti-air.
  • I use tatsumaki as a gap closer/attack which is slow, easy to block and punish. I should stop using it on its own, and only utilize it in combos.
  • I miss most opportunities to do the 4-hit target combo (the only combo I can reliably reproduce almost 100% of the time) after hitting a jumping HP.
  • I get caught in a crouching blocking position by overhead attacks a lot. I should stand up more while blocking.

SFV Streamer Highlight: Chris_H_Nguyen


While looking for a good Akuma stream, I stumbled upon Chris_H_Nguyen. He’s very relaxed, even when grinding ranked games. The fact that he’s playing the Ambition’s Lofi Hiphop Playlist certainly helps with that.

Chris continually engages with his viewers in discussions about tech, frame data and tests lots of things in training mode whenever it is mentioned in chat. He’s never worried about looking bad, but just focuses on improving. That’s why he actively invites viewers to correct him if he makes a wrong assumption somewhere.

I started watching Chris at the tail end of his climb through Ultra Diamond, and as of yesterday, he officially hit Master rank. He’s confident he can hit Grand Master soon, and I tend to agree, seeing as how he recently got 65th place at EVO Japan 2018. I’ll be there when it happens :)

Below you can find the last 4 streams before he hit Master league:




2018/02/12: (hits Master League at the very end of the stream)

SFV – Streamer Highlight: AMKIDD

Image result for amkidd

Where were you when AMKIDD became the first US Ryu to hit grandmaster league in SFV-AE? You can read about the achievement in this reddit post.

I went through some of his VODs and have started watching him stream and he is the perfect embodiment of both the street fighter mentality and his main character Ryu in particular (though not quite as boring as the latter). He doesn’t look for excuses anywhere, just works really hard and is always actively trying to improve on his game. On top of that he’s very entertaining, is great at commenting his play while he’s streaming and drops some dope rhymes on a regular basis. If you’re trying to learn the game (especially if you’re doing it on Ryu), he will serve as a great example of a player with priorities on improvement, not winning, resulting in an extraordinarily tough mental game.

Below you can find the VODs of his streams that led up to him hitting grandmaster:





Started working with TL.net writing staff

 I’m a huge fan of StarCraft community website teamliquid.net, so after years of reading their content, I applied to start doing graphics work for them. They said yes! So that’s exactly what I’ve been doing for the past 4-5 weeks.

Below you can find some articles for which I made graphics:



I hope to start contributing articles as well in the near future. Will keep you posted!

Learning to Draw 01: Basic shapes (and getting ahead of myself)

I’ve always loved to draw, but never received any formal training. Because I’m getting a bit more serious about improving my drawing skills, I started taking an online class. Rich Graysonn goes over the absolute basics that make up the fundamentals of drawing in How to Draw and Sketch for Beginners.

So far, I’m mainly doing lots of basic shapes exercises:

20180213 Pulling Lines20180213 Rectangles20180213 Squares20180213 Triangles

But then of course, I got a bit antsy and way ahead of myself. A good friend of mine made a series on house dust mites a couple of years back, and ever since seeing those I’ve wanted to do my own series on insects. I’ll gradually keep drawing insects throughout the course. I’m not unhappy with how the first beetle came out, but there’s always things to improve on. Hopefully I’ll be able to do so by the time I’m done with the course.

20180213 Beetle

SC2 Streamer Highlight: DeMusliM

DeMusliM largeBenjamin Baker, aka DeMusliM, aka The Devil Terran has been a part of SC2 since its inception. He was a staple in the earliest tournaments, but he’s been focusing on streaming more than tournament play for the last few years. He was recently picked up by team Wind and Rain, though, so maybe we might see him at more tournaments again?

His stream is super fun to watch, he’s often memeing around with his viewers and on more than one occasion I’ve seen him facetime other pros to have a beer together. Support Benjamin, drop by at the stream. You might even see him break out the ol’ sunglasses:

DeMusliM will be at IEM Katowice 2018, but he got paired with Stats in the first round. Here’s what he had to say about that:  

SC2 Build – TvP – SpeCial’s Ghost Rush


During the Corsair Cup S11 Round 3, 2017’s breakout Terran SpeCial played a proxy ghost rush into double cyclone cheese build in his series vs DnS.

Venipenny posted the following build on Spawning Tool: SpeCial proxy ghost into cyclone followup (TvP Cheese).

special ghost rushAnalysis:
This is a cheese build. You will be behind in economy and have to deal a lot of damage to be even in the midgame. Ideally, you just deal game-ending damage with this build, otherwise you will be playing from behind.

You should try to lift your barracks into their main, and hide it in the fog of war. Immediately start a tech lab and when the ghost is halfway done, send the scv that built the barracks to the natural to build a bunker right next to their nexus/pylon. This will draw units and attention to the front of their base. Then you can run the ghost into their main mineral line and start killing probes. Hold off on cloaking immediately. Even when their stalker/zealot comes back to the main, you can take a few hits before turning on cloak. This allows you a little bit more time to kill probes, as well as giving you more energy to stay cloaked longer to be able to get away.

If Protoss doesn’t defend the initial ghost rush well, this can be an auto-loss for them, but as they get better in defending with shield batteries, it will usually not deal game-ending damage. As your expansion will be later, it is imperative that you do a followup attack. That’s why you absolutely have to get both ghosts out in time, so that you can have them start charging energy. By the time the first two cyclones are at their natural, you’ll have enough energy for 1-2 EMPs. This can instantly deplete the shield battery as well as damage the Protoss units’ shields enough to be able to push in and kill the Nexus at the natural expansion.

I asked DeMuslim about this on his stream (05/12/2017) and he said he really disliked the build, because protoss can just react correctly with shield battery in both mineral lines and it doesn’t work at all. Stalker will force cloak too soon and you don’t get enough out of the build.

Seems like this was an interesting experiment, but it honestly shouldn’t work.


Silver is a bit of a cult-classic among RPGs. Unfortunately, because it was released around the same time as Baldur’s Gate 2 and Diablo 2, it never really got the widespread attention it deserved. The game was also released on SEGA Dreamcast in 2000. In recent years, the original PC version had become buggy and downright unplayable on modern machines, but there is now an updated version available on Steam.

The game is set in the world of Jarrah. David, the protagonist, is introduced when his wife Jennifer is kidnapped along with all the other women in his village to satiate the desires of the evil sorcerer Silver. While trying to rescue his betrothed, David gradually learns of Silver’s true goals, and how those do not necessarily involve a fun time for the rest of Jarrah. The story itself is not all that special, but it serves the rest of the game well. Many of the dialogues are well-written, and there is a very rich cast of memorable NPCs.

The game combines two different techniques to craft its own unique graphical style. The player, the NPCs, enemies and most of the interactive game objects are represented by a low-poly 3d model. While these obviously look rather dated, their animation is still surprisingly good. The main graphical feature, however, are the monumental, pre-rendered backdrops. The image quality of these has faded a bit over the years, but the designs are still amazing.

Controls and UI
The controls are quite different from a lot of its contemporaries and still feel fresh and responsive. Silver uses an elaborate input system of mouse clicks and swipes that allow you to perform several different types of attacks. As long you perform the correct input command, you can slash, sweep and lunge in any direction as well as block incoming attacks. (But why equip a shield when you could dual-wield swords? Madness! Madness, I say!) The system is pretty intuitive as well, so it really is a major bonus and one of the main reasons the game still has a dedicated fan base today.

The sound design is excellent. In many regions of the game there’s no real background music, but there’s always some appropriate ambient sound (e.g. chirping birds, leaking sewage pipes). When the background music does appear, it’s always dramatic classical music, which really adds to the epicness of the scenes. The voice acting is outstanding, which lends most of the NPCs bucketloads of character.

There’s combat with mobs of enemies, boss fights, puzzles and a bit of exploration in the game. Combat is the main course, though, and it’s usually a bit on the easy side. The enemy AI isn’t very aggressive and as a result combat tends to be rather easy. Some bosses are a bit more challenging, but overall the game is definitely on the easy side.

When a PC-game is close to two decades old, things starts going wrong. The original PC version had become literally unplayable, but the steam version runs very well on modern machines. However, I have experienced a crash or two. Both times this happened right before a save point, which was a bit frustrating. Because of the spacing of the save points, you rarely have replay more than 10 minutes, though.
More annoying is the fact that at a certain point in the game there’s a timing-based puzzle. Because modern machines have increased the frame rate, what used to be a fairly easy puzzle has become a veritable nightmare. I passed it after around 10 minutes of trying, but I’ve read accounts of people being stuck for hours. There’s only one such puzzle I know of, and with some help from YouTube you should be able to complete the puzzle too.

Even two decades after being published and in a world with an incredible amount of competitors, Silver is still worth your time. The controls are still innovative, the graphics hold their own, the music is as good as ever and the few times the game did glitch out never prevented it from being a really nice experience. The game is super cheap so you have no excuse not to play it.

Simple Lamp Animation

I modeled and animated a simple lamp to learn about rigging and kinematics. Below you can see the complete process.

First I modeled a simple lamp, using both box modeling and bezier curves:

wip lamp wirewip lamp renderwip lightbulb materialsThe light bulb was also modeled through bezier curves. It was given a blackbody material to more realistically represent the colour tone of a light source when interacting with another surface (see the slightly warmer colour on the inside of the lamp)

wip lightbulb wirewip lightbulb renderI then rigged the entire lamp and constricted its movement to rotation around a single axis depending on the type of joint.

After the basic rig worked, adding reverse auto kinematics made all the bones in the rig respond when only a single one was being manipulated. This gave the whole rig a more realistic feel to it.

Next up was animating the scene:

I found the scene to be more dynamic once I also animated the camera:

After adding some simple textures I rendered the entire thing:

A simple 6 second animation at 720p 60fps still took well over 12 hours to render. I sincerely hope I will not have to render Full HD or 4K projects any time soon.

I’m glad I finally learned the basics of animation. I had been putting that off, but it’s really not that difficult to get a working rig.