At Pairs and teams scoring a Non-vulnerable game is worth 400/420. A Vulnerable game is worth 600/620. It follows that if you can somehow give away less than either of those two scores you will have returned an actual plus for your side. If the opponents contract to play 4 Hearts Vulnerable they will make 620. If your side is not vulnerable you can go 3 down doubled for a score of -500, an actual profit of 120. However - be careful. The loss of 500 is only worth it if the Vulnerable game can actually make! And even if it can, you must hold your losses to that -500 mark, as if you go down for -800, you just turned the loss from 620 to 800 instead. This is not good. Don' be fooled into thinking that all sacrifices are necessarily the correct thing to do. It is probably as true anything can be in this game that if you are the Vulnerable side and the opponents are Not Vulnerable, then to sacrifice is almost never the correct thing to do.
The crucial point I am making is that how far you bid will depend not just on what cards you hold, what partner bids and what the opponents do but also what the score is. This applies to all forms of the game - Rubber, Duplicate and Teams. This concept also applies to the part-score battle, when both side have roughly an even number of points. This is particularly true at Duplicate, rather less so at Teams or Rubber. The maths is the same: if they contract to make 2 Hearts for -110 to you, then you can go for -100 in 3 Clubs or Diamonds (say). This profit is so marginal that at teams scoring it is a flat board, but it's very important at Duplicate, where the size of the profit is not an issue. if you beat all the other pairs by 10 points, you get the same Matchpoints if you beat them all by 1000.
However - Rubber Bridge players should beware the siren song of the sacrifice. Do not be lured unthinkingly to those rocky rocks. Let us assume a rubber score where you are not Vulnerable, but they are Game and 60. If they get this game they get a 700 rubber plus their Game and overtrick scores and any other scores above the line you might have gifted them on the way. So. Thinks. If we can sacrifice for anything less than 700, you think, apparently rationally, then we are quids in. WRONG. Ok let's say you duly sacrifice in this dubious enterprise and go for -300. Aha! you might think, saved 400! WRONG. The reason is that when the dust has finally settled they will STILL be Game and 60 and will gratefully receive the 300 you just donated to them. Now, having invested that 300 in trying to save the rubber, will you now protect your investment by sacrificing again and again until you finally do get a game? And how will you feel if, in this extended rubber, your opponents find a slam, thus giving them another 1000+ points into the bargain? Initially you were worried about chalking up a -900 rubber but now you find yourself scoring about -3000 instead. The way to avoid this common pitfall is to know when to give up. Minimise your losses by getting out of the rubber as cheaply as possible and move on to the next when hopefully your luck will change. Do not flog dead horses. Cut and run.
This is why - before you even look at your cards - you must check the vulnerability. Your cards don't really matter. The Vulnerability is utterly crucial.