Unless your error rate is VERY high or you are playing with a very low edge (under 0.5%) you will make MORE money playing fast even if you make more mistakes. If you do something like remove your hands from the buttons to take a second look, your speed will probably go down by 30-50%. But mistakes just don’t cost that much. Most errors that involve making the slightly wrong play cost well under $.50 on a dollar game. So in an hour, your errors might add up to $2-$3, but probably not more. But if the play is worth $25 an hour (Such as a dollar game with 1000 hph and 0.5% total edge) and you decide to address errors in the manner suggested, you wind up with an actual edge of between $12.50 and $17.50 an hour. But allowing the mistakes to happen still leaves you with $22 - $23 an hour. Even a $5 an hour error rate is better than slowing down by 30%.
Optimization is rarely about perfection. The same kinds of arguments can apply to the strategy one learns as well. In a great many cases, the time required to look on your phone app to see the right play will cost MORE than the “error” of making the second best play. Some folks also fatigue a lot more if the strategy they employ is too complex, which is about the surest way to create an unacceptably high error rate and may even cause you to play less despite having an edge, which clearly has a large impact on your daily expected value.
QZ