The best copy protection is one that breeds such mistrust in a crack, that the user goes out and buys the game instead.
Having a protection check right at the end of the game is useless, quite a lot of people would be happy that they got that far (assuming they ever did) and the interest in the game would be lost. They are not going to go buy the game just for one level and probably an ill thought out and unrewarding end game sequence.
Exile NEARLY got it right.
Hook got it wrong.
Robocop 2 got it right.
There were two seperate fixes for Robocop 2, for some people that would have been too much. Though the game was correctly cracked in the end, there would still have been enough distrust in some people thinking that they didn't want to play the game and keep getting stuck because of a cracking error.
Robocop 3 got it right.
By preaching the myth that the game was virtually uncrackable, plenty of lamers actually believed it was true, so they all went out and bought the game. It didn't matter that it was in fact 100% cracked first time, the myth of the protection was better than the protection itself.
A games protection needs to kick in not so early that a cracker can spot it but early enough into the game that the user still has a keen interest in it. A cracker will not spend a week playtesting a game to ensure its cracked properly, but a user will still be interested in a game a week later, so any copy protection scheme should activate after an approximate weeks worth of playing.
This is what I did for Premier Manager 3 Deluxe for Gremlin Interactive. I know for a FACT its never been 100% cracked in its life, with the advent of MMU's everywhere, its probable that it would be cracked a lot quicker, and the way it was designed was purely for those people that knew the game inside and out that would know that there was something wrong, and would therefore be more likely to go and buy it.
As for crackers not showing an interest in keeping on doing 100% versions, its true. In the days of Skid Row and Fairlight, there could be anything up to 5-6 games a DAY to crack. You dont have time to keep revisiting old cracks, because when the information gets back to you that XXX game isn't cracked properly, it could well be a week later that you hear about it, and the chances are, you reasonably suspect someone else has spotted it and recracked it anyway. In the case of Robocop 2, Marc spotted the mistakes over two days and rectified them, but the damage is already done. No sooner have you sent a version to your mail traders that supposedly works, you then tell them TWICE more that you have versions that REALLY DO work now!
In the case of Robocop 2, as a protection it was hopeless, but its implementation was what saw it succeed. Sometimes the easier protections are a little more tricky than people anticipate.
