Zorcey
Smash Journeyman
- Joined
- May 12, 2015
- Messages
- 371
I've started work on structuring my mixups more concretely, and I'm considering some different ideas about data collection. For example, my current strategy when poking an opponent with Dtilt for the first time is to dash back and take note of how they respond. Once I have an idea of how they react to the threat of the poke and the poke itself, I become more aggressive by approaching out of it, covering options I noticed they took previously.
However, I'm considering the possibility of approaching out of my Dtilts right from the start of a match, taking the risk of a mixup based more on meta tendencies than player-specific ones at first. I figure that best-case I get a strong opening conversion, and worst-case, while I may get punished for losing the position, I get psychological leverage because my opponent will be incentivized to choose the same option again because of their previous success. It's becoming increasingly common to cover Marth's dash back, so I think it has to be used much more sparingly, and this seems like a possibly strong alternative. But it could also be that approaching before you get player information is playing Marth's hand poorly, so I'm debating.
Overall, my question is what you think about the viability of approaching on meta-based information rather than player-based, and collecting data COMMITTALY instead of noncommittaly.
(I also wonder about its viability versus top players, who understand positions well enough that one victory won't easily tempt them to pick the same option more than maybe once or twice in a row.)
However, I'm considering the possibility of approaching out of my Dtilts right from the start of a match, taking the risk of a mixup based more on meta tendencies than player-specific ones at first. I figure that best-case I get a strong opening conversion, and worst-case, while I may get punished for losing the position, I get psychological leverage because my opponent will be incentivized to choose the same option again because of their previous success. It's becoming increasingly common to cover Marth's dash back, so I think it has to be used much more sparingly, and this seems like a possibly strong alternative. But it could also be that approaching before you get player information is playing Marth's hand poorly, so I'm debating.
Overall, my question is what you think about the viability of approaching on meta-based information rather than player-based, and collecting data COMMITTALY instead of noncommittaly.
(I also wonder about its viability versus top players, who understand positions well enough that one victory won't easily tempt them to pick the same option more than maybe once or twice in a row.)