Then your opponent has a weak serve.
Not unless the serve comes in quite slow, which you can see in 3.5 and some 4.0s. And if you aren't crushing it with pace, the net player can easily get a decent shot back unless they've fully committed to crossing. Realistically, the serve isn't a consistently forcing shot until 4.5 (around the time when players learn to consistently do 1-2 combinations with the serve).
It's all about probabilities. Would you rather give up the 20% and gain leverage over the 50%? Or cover the 20% and play the 50% out?
Say the chance of them going down the line is 50%. The odds of them making it are 20%, and your winrate when poaching is 60%. You win 40%+30% = 70% and lose 10%+20% = 30% of the time. Assuming independence and all that statistical crap. Let's say they win going into the alley 40% of the time, and you win with poaching 60% of the time. Poaching still gives you a 60% winrate (30%+30%). Say you didn't, and your odds of winning goes to 50%. Your winrate goes to 55%. However, I argue that poaching should give you more than a 60% winrate when they go crosscourt, and they probably shouldn't be making 40% of balls into the alley unless the serve is weak. Therefore, the actual numbers should be higher. And even consider the fact that, even if they went down the line 100% of the time, you're still winning more than you lose. We can make the math more complicated and cover every possible situation and be more exact, but the bottom line will still be that you win way more giving away the alley and playing the center than you lose unless you lack the ability to hit a decent shot. If we're talking about sitters, the whole thing changes because even 3.5s can go anywhere with that without missing (or at least make more than they miss).
Yep. Math never fails. RNG however, that one's a *****. There's also pot math to take into account for this though. I understand the basics of it, and am too lazy to read the more advanced theory then do the calculations. Winning the hand is always in your favor. Whether it's financially worth it for you to do this every time is another story and depends on the pot and investment sizes. Although the only way it's worth it for the other guy (since if I remember correctly, outside straight is only 25% for the river) is if their all in is pretty small compared to the pot, meaning your investment size would also be small relative to the pot, making it favorable to you financially as well. This is also, of course, ignoring all the mind gaming crap in poker.
Not necessarily. Using the same analogy, if I all in on the blinds with pocket kings or aces, I'm putting myself at risk for very minimal rewards (the blinds). But the fact is, it's so high percentage, I will win far more than I lose, and will have always be net positive in the long run (even though if I lose once, I would lose far more than I could've gained in that one instance). The idea is to bully someone into making a bad choice. If they fight you (going down the line), they'll probably lose. If they let you have your way (poach the crosscourt), they'll probably lose.
And poker is comparable to a lot of things. You aren't only playing one point in tennis, you're playing dozens of them in just one set.