>>21145585By feel.
Like
>>21145671 said, you determine their options and your options, and then gauge each decision's likelihood. In judging the likelihoods, you have to gauge how good each decision is, but also judge your opponent based on any knowledge of their style and past moves during the battle. You're not just trying to make the statistically best move, but the best move based on your opponent (the Poker factor) and like
>>21145816 said, you adjust according to the flow of the battle. For example, sometimes you have two solid choices, but you have another option which doesn't really gain you anything but puts you at less risk of losing some advantage. Or you're at a point in the battle where one person can't inch their way back into the advantage and needs to make a higher risk, higher reward kind of choice (or a Hail Mary that leaves them extremely vulnerable to sealing their loss but could pay off huge and really turn the battle around) so you need to be considering those types of choices or trying to block them from paying off for your opponent.
Risk management is twofold, you not only gauge likelihoods of a positive/negative result, but also the gravity of a negative result.
Always be thinking how each move choice will be affecting future decisions several moves down the road. If it's a 3v3, you should be thinking all the way to the end of the battle on each move.
So it's a lot of statistics being done "by feel" coupled with judging your opponent and the situation you're in.
It's definitely a lot to consider, which is why in serious battles, I often take the entire time allotment to make my choice.
>>21145836Those are things you need to be considering in your decision-making process. And hopefully the choice you landed on has a positive, neutral, or only slightly negative result (that had a high probability of paying off very well for you) playing off of those possible choices by your opponent.
Also,
>>21145741 can help cure predictability.
