One of the things that new SF players take for granted is that an input will result in the same thing happening every time. This isn't an unreasonable assumption most of the time - if you input throw, you want to get a throw every time, and most of the time you will. However, occasionally there will be some ambiguity in how the game engine interprets inputs based on your interaction with opponents, and this can result in one input giving different results in different situations. Inputs that can give two different results based on in-game character interaction are known as option selects.

Let's examine some of the most common option selects in 3rd Strike, and what makes them useful:


Crouching throw tech - d. + LP + LK

Earlier, I mentioned ambiguity in how the game interprets certain inputs; one such example includes the throw input while crouching. The 3rd Strike engine doesn't allow you to throw someone else while crouching - try inputting d. + LP + LK to verify this if you like. However, it does allow you to throw tech while crouching, with the exact same input. There, the d. + LP + LK input has two different results based on what your opponent does:

Opponent tries to throw you: you get a throw tech.
Opponent doesn't try to throw you: you get a cr. LP. This is because LP has priority over LK if both buttons are pressed and throw can't come out for reasons like the one listed above, that you can't throw while crouching.

It should immediately because obvious why this is a useful option select - instead of potentially whiffing a throw in the second scenario, which is often what your opponent is trying to bait you into doing for an easy punish, you get a fast low poke that can interrupt your opponent's rushdown. Instead of just covering a throw tech with a standing throw, you also cover other options like tick throws, hesitation moves, etc. This is why crouching when going for a defensive throw tech is often recommended.

Of course, there are still vulnerabilities to this option select - no offensive or defensive option is ever foolproof. If you use this too frequently, your opponent can simply bait it and parry the low jab or space themselves in a manner that allows them to punish it with a move that outranges your low jab. You can always mix things up by throwing in regular standing throws, which will punish parry fishing, or simply jump to escape throws and many other offensive pressure chains. The key, as always, is to avoid falling into easily identifiable or punishable patterns.


SGGK - high/low parry, followed immediately by kara-throw

Another ambiguous situation arises when parries are combined with kara-throws. Typically, kara-throws involve cancelling a normal into a throw so quickly that the initial move doesn't connect. However, after a successful parry, the 3S engine only considers the next valid input to determine what comes out. In the case of a kara-throw, the next valid input is the move with which the throw was kara-canceled, so instead of getting a kara-throw you get the move with which you kara-canceled it.

By far the most well-known example occurs with Chun-Li's kara-throw with close st. HK. Ordinarily, one would use far st. MK to kara-throw, as it gives considerably more range. However, after a successful parry, close st. HK is a far more useful move than st. MK, as it can be comboed into SA II. Therefore, the idea with st. HK xx LP + LK is as follows:

No successful parry: Kara-throw comes out.
Successful parry: close st. HK comes out, which is comboed into SA II if you have meter.

Again, this puts your opponent in a difficult position - going for a throw tech is risky because she can simply step out of range and do a standard SA II combo, but throwing out fast low pokes like cr. LP or cr. MK is also risky because she can parry into st. HK with SGGK. Incidentally, this is an excellent counter to the previous option select, since both possible results are either neutral (throw tech) or in Chun-Li's favor (parry low jab, st HK xx SA II).

Below is an example of the SGGK option select in action (occurs at 0:39, 1:57, and 3:30):

D.E.D - meter based option select

Thus far the option select discussed are possible due to ambiguity in how the game interprets your inputs - a kara-throw not being interpreted as such, for example. However, there are also option selects that don't rely on quirks of the 3S engine, just on basic mechanics such as how meter works, and the fact that you get more meter for a move that connects than one that is blocked. This option select only works if don't have a full meter yet, but are reasonably close (~70-90% depending on the normal used for the option select, and the size of the meter).

To explain the D.E.D option select, I'll use a hypothetical situation and some made-up numbers. Say you're playing Chun-li, and you're currently at 85% meter. A blocked b. + HP gives ~10% meter, whereas a b. + HP that connects gives ~20% meter. (These numbers are probably not accurate, but bear with me). What happens if you do back fierce into super without confirming in this situation?

Back fierce is blocked: You're at 95% meter when you attempt to super, so nothing happens. All you get is a blocked b. + HP.
Back fierce whiffs/gets parried: You're at 85% meter when you attempt to super, same as where you started, so nothing else happens.
Back fierce hits: You're at 105% meter when you attempt to super, so the super comes out, and it combos because b. + HP connected.

The end result in this situation is that super only comes out if back fierce actually connects, so there's no need to confirm it in the first place. You can just bang it out without confirming and be perfectly safe in doing so. This is a particularly nice option select because there's no real down side to it - you might get parried, but then it's not any worse than just having a normal parried in any other situation, at least you don't waste a meter in the process.

Now, although it's nice to just be able to do something like back fierce xx super without confirming, you could always just confirm it too, so that's not where the real value lies. The reason this is so good is because you can use it to do combos that would normally be unverifiable, like cr. LK xx SA III or close st. HP xx SA III, both with Ken. Ordinarily close st. HP is an unverifiable normal, and you'd need two cr. LK's to confirm the combo into super, but if you're paying close attention to your meter, then you can just throw these out safely. You gain range on both of these D.E.D option selects since you'd need to be closer to combo two cr. LKs than just one, and to do the MP, HP target combo instead of just HP.

Here is an example of D.E.D in action. Ken does the normally unverifiable cr. LK xx SA III combo by taking advantage of D.E.D (occurs at 24:56): 

Option selects are crucial to becoming a better, more consistent player, and mastering them will help you out in every aspect of your game.

Another fairly simple post, this one is intended to identify the most commonly used supers, and separate them from the ones that aren't used in competitive play very frequently. Below is a table containing what I think most high level players would consider the "viable" supers with each character. For many characters, there is 1 super that is clearly the best; however, some characters have 2-3 that are used frequently enough to warrant mentioning, and this is reflected in the table. I did arrange in the supers in my personal order of preference - I would generally use "1st Choice S.A." more frequently than "2nd Choice S.A.". However, whenever there are 2 or 3 viable supers, the choice of super is usually determined by matchup-specific considerations, so it's not always a good idea to pick the "1st choice" super. If a super art isn't in the table, it's generally not considered viable in competitive play.

Below the table, there is a short description of why each super in the game is either considered viable or not. Green text means the super is listed in the table below as viable, red text means it's not.

This post will be fairly short, as it doesn't concern mechanics or character specific tactics. For players who have the basics down and are trying to learn more advanced tactics, it often helps to know which players are considered to be elite with each character, as they can go find footage of these players and see what high level play with those characters looks like - mixups, spacing, etc.

In the table below, I've provided 1-3 players who are widely considered to be among the best at their respective characters. Some notes about this table:

1. This is no way intended to be a comprehensive list. Just with Yun alone, there are at least 7-8 famous, world-class players - K.O., Mester, Ochibi, Nitto, Yakkun, Boss, Momochi, Matsuda, etc...and that's just with a few seconds of thought. All of them are worth watching, I just don't want the list to get unwieldy or out of hand.
2. The players aren't arranged in any particular order either - "Player #1" isn't necessarily stronger than "Player #2", they're totally randomly arranged. This isn't a "top 3" list either - there may be other players with that character who others would consider stronger or more worthwhile watching.
3. The list is somewhat biased towards players who are active today, since it will most likely be easier to find footage of such players, both at present and in the future. There are still some older "retired" players on the list though, either because very few people play that character at an exceptional level, or because they were sufficiently well-known during their active period to be worth mentioning now.
   
4. This is list also biased towards characters that players used competitively today - for example K.O. has a very famous Yun and used to use him in competitive play very frequently until 2005/2006, but nowadays he mostly plays Yang, so I put him down for Yang and put down other players for Yun. Same goes for other famous older character/player combos like Boss/Yang, Tokido/Urien, etc. Of course, if you see a video with K.O. Yun or Boss Yang it's a safe bet that it's worth watching anyways.
5. If in doubt, Kuroda is a safe bet for virtually any character. He released DVDs of himself playing every character, so if you can find footage from them or get a hold of them, then they're pretty much guaranteed to be worth it. I didn't list Kuroda aside from Q because I didn't want this list to just be peppered with his name on the majority of the characters.
6. For the characters with only 1 player listed (Remy, Twelve), there is 1 player who outdistances the rest of the competition by such an enormous margin. In addition, they're both low-tier, fairly underplayed characters, and I don't know who else to put (I'm happy to take suggestions). I didn't include Sean because there are no notable Sean players AFAIK. He's pretty much a shittier version of the other shotos in every respect, so you could just go watch a good Ryu or Ken player and have a pretty good handle on how to play Sean as well.
   
   

Also, a tip for beginners watching videos of these players in the hopes of getting better: don't get hung up on the flashy stuff. Yes, it's really cool when MOV breaks out some insane lightning legs setup, or Hayao does a standing Gigas, but these things are only a tiny part of the reason these players win so frequently. Focus on more fundamental things like spacing, rushdown mixups, defense/okizeme, etc. You can spend all week trying to master a true kara/Jiro demon, but you'll just get stomped before you even reach 2 meters if your fundamentals aren't solid, and a demon by itself only does ~40% damage, so it won't win you a round or a match single handedly. Same goes for kara palms, double-fukiage 100% stuns (although these do win games single-handedly), and other flashy, tricky stuff in the same vein.

Happy viewing!