Mixup calculator

(topic: )
Stand
Crouch
Mid
Low
Best strategy: JavaScript must be enabled
Average payoff: JavaScript must be enabled

This is a basic calculator to answer the question, “What is the payoff for this mixup, and how often should I go mid/low?”

The inputs are the damage you deal—or receive in punishment if negative—in each case. The default values are for Lee’s f,F+3 / slide mixup.

Clicking the tabs changes the labels. This has no effect on the calculation, but makes putting the numbers in the right place a bit easier.

This calculator is helpful to get an idea of how strong a particular character’s mixup is compared to other characters, or compared to other mixups they have. It’s also helpful for finding the optimal strategy.

It can’t properly take into account frame advantage or oki, nor can it take into account a situation more complex than a simple 2-by-2 payoff matrix.

If you’re interested in doing a more comprehensive exploration, have a look at The Gambit Project.

Factoring in oki

You can attempt to factor in oki and frames by adding the expected payoff of the mixup to the cases that give another mixup. For example, for a Kazuya hellsweep / f,F+4 mixup with no walls, you’d start with 0, 22, 34, -75. This gives a payoff of 5.71.

You can add 5.71 to the hit options and subtract 5.71 from the blocking options (assuming your opponent also has a mixup with the same payoff), giving a new matrix of -5.71, 27.71, 39.71, -80.71. Now the payoff is only 4.16! The big difference is that the mid is no longer safe on block.

Note this assumes that the large frame advantage from f,F+4 on hit and disadvantage on block guarantees a mixup, which isn’t totally accurate.

You can repeat this process as much as you like. With a mixup payoff of 4.16, the new altered matrix is -4.16, 26.16, 38.16, -79.16. This one has a payoff of 4.53. Now the altered matrix is -4.53, 26.53, 38.53, -79.53, giving a payoff of 4.44. You can keep going until the payoff is stable, or be happy saying it’s about 4.5.

This is still an incomplete picture, since the option to stay grounded hasn’t been considered, nor has f,F+3 having much better payoff than f,F+4 for a mid.

For staying grounded, this is a strictly worse option. The best case is getting hit by hellsweep, taking 8 damage and ending the oki. This is worse than the average payoff of the mixup, so strictly worse than getting up. On top of that, f,F+4 will deal 16 damage and keep them in the vortex, and the best strategy is to go mid a lot more than low.

For f,F+3, this makes the payoff much better. Instead of taking a mixup on block, and only getting 22 damage and another mixup on hit, f,F+3 is almost neutral on block and gives a full juggle on hit. The matrix in this case is 0, 75, 34, -75, for a massive payoff of 13.86. If you then add the oki in, assuming they keep getting up, the payoff goes as high as 25.

Now you might wonder: well, if f,F+3 is so good, why ever do f,F+4? Because if you just did f,F+3, the option to stay grounded suddenly looks really good. Sure, the hellsweep does 8 damage, but f,F+3 misses entirely, and both of those are a lot less than the mixup payoff now. So you need to do f,F+4 to get people to actually stand up.

This is an imprecise method, but gives a good enough estimate to be useful.

Analysis

The payoff for ordinary mid/low mixups is a lot worse than typically thought, usually 4–6 damage. In these cases spamming mixups isn’t a great strategy, because the price you pay to enforce them isn’t worth their payoff. Pokes, counter-hits, and movement are where you’re going to get a real advantage. They are of course the best way to deal damage to an opponent who’s entirely locked down.

This can change quite dramatically, though, with powerful lows and safe launching mids. If the payoff also includes that same mixup, that’s when you get a terrifying vortex.

When it comes to pokes, how often you should go low depends a lot on if your opponent is consistent with low parries and where the wall is. In cases where a low parry leads to a full combo with wall damage, the risk is huge even on a checking low like generic d4. In these situations, you may have to almost half how many lows you throw out. This also means that if you aren’t low parrying much right now (compared to block punishing lows), you need to get on that.

For most characters, without parries you can go low with pokes up to 28% of the time. With parries and no wall, it’s around 20%. With parries and full wall damage, it’s still around 15%. That’s entirely based on the mixup—crushes and interrupts haven’t been factored in, for example.

The best time to enforce a mixup is into a tech roll. This is because not only can the opponent not interrupt it, but it’s much harder for them to low parry.

The fairly low payoff on most mixups means that throws are a lot stronger than you might think. Even pro players don’t break throws 100% of the time, and one getting through does a lot of damage. They’re obviously risky, being duck punishable, but they’re a far cry from useless.

It also means that in most cases, guaranteed damage is best. For example, a lot of players think the oki from Lee’s slide is worth more than 8 damage, but it’s actually just a high risk, high reward choice. Against correct defence, the average payoff of the oki is much less than 8 damage.