For most people, it is not because they think it is good. Let us face it: it is not really suited to what most people want to use it for. M.A.G.U.S. is undeniably a product of the early 1990s, which is partly good and partly bad. Good because, under the banner of the “story revolution,” people wanted to offer more and more room for story and, of course, roleplaying. But with a somewhat unfinished system, there was not enough time even to finish many things, let alone test them or edit the book properly. The problem is that perhaps there was not even time to gather their thoughts.

That produced the situation at the Rúna camp where players asked Csanád Novák about wizards and hydrogen bombs, prompting VP's big boss to angrily declare that hydrogen did not even exist on Ynev. Most players did not believe this from creators whose credibility had been battered in their struggle against competitors. Yet where there are primordial elements, primordial water is an element, and breaking water down into hydrogen and oxygen makes no sense. So Csanád Novák was essentially right, except...

Except that the MAGUS creative team had so little credibility left that nobody would believe them even if they said 1+1=2. And when official creators themselves treat the previous team's work this way, unified, considered systems are not usually the result. From then on, M.A.G.U.S. became famous for mountains of house rules and the fact that there was no idiocy someone would not include in a game, even as a PC. It is almost inconceivable that anyone would seek both the originally intended roleplaying approach and this mess in the same game. 

There is one small problem, though. However many problems players may have with M.A.G.U.S. or the M.A.G.U.S. groups realistically available to them, the alternative still ends up looking worse. And when they take out some edition of the big green book, several things happen. Part of it is biology: a conditioned reflex. Holding the book and dice already releases hormones; pleasure and happiness arrive, and the addiction gets a little feeding too. Long live M.A.G.U.S.! Another part is that memories really do return. That is the minimum they can expect.

The bigger problem is that when “d20” appears instead of “k100,” they do not ask why d was not good enough for the Hungarian word dobókocka, meaning die... Obviously, because of AD&D. And just as Dungeon Master became Kalandmester, which does have some advantages, D could be replaced with K here too. The truth is that in 1993 few people were online, IRC was hardly widespread, and it did not occur to anyone that d was used not only by many systems, without legal problems, but was also better understood by programs because it was standard. Yet being different for its own sake was not necessarily good for M.A.G.U.S. even then. At the time, though, presumably out of fear of legal issues, it was necessary to insist that M.A.G.U.S. was not (A)D&D.

When the backbone of your identity is that this is not AD&D or D&D, comparisons with AD&D and D&D immediately suggest themselves. Under normal circumstances, that is a comparison M.A.G.U.S. could not win. But it means nothing that a badly edited, unfinished system cannot be better than the established, evolving AD&D, even if we ask people who get the most positive results from M*. Because with hard work and help from many influential voices in Hungary, D&D can lose that comparison. It can provide a worse experience than M.A.G.U.S. or many other systems. 

So D&D's third edition arrived, bringing development under an open license and plenty of other things. My opinion was decidedly mixed, and remains so to this day. While plenty of people in our little country declare that they “run Alien in the Vampire system,” and sometimes try to prove D&D 5's merits by using 5E-compatible homebrew systems such as Mass Effect 5E, the biggest innovation in third-edition D&D was the OGL. The ecosystem forming around D&D, d20 System, and the many games built on it sometimes competed against individual games. That is why fifth-edition D&D also has the OGL and an available SRD. 

When someone is proud of how much they can replace in D&D, and how much they need not even read, we gradually reach the point where they are playing a 5E-compatible homebrew rather than D&D, and forgetting what the OGL and SRD are about. They are about having an alternative within the ecosystem when D&D gets something wrong. Having another option when a particular D&D edition fails to meet someone's tastes or needs. And if another game contains just a few good ideas but an unfinished underlying system, letting it become a game that receives most of its system ready-made while contributing its own good ideas to the ecosystem.

The dual health system would have been no worse if it had been packaged with, say, a d20 treatment of gladiatorial combat in Toron instead of appearing in Star Wars d20. Nor would it have suffered if they had included a map of Ynev and promised several similar small publications in that world, while providing ideas about how to fit it into other settings. The OGL, open source, and the ability to “fork” the common foundation into a game of your own have considerable advantages. 

If someone pays attention to these possibilities of the OGL and finds D&D's proper place, then D&D works for certain kinds of game. It promises more than it delivers, and in that respect it is bad, but the ecosystem stands alongside it, and there are situations in which D&D is the best available choice. The average M* player, however, will not see this side of D&D. The person trying hardest to persuade them will be a fanboy or fantroll who does not understand these advantages and opportunities. Someone who does not understand that needs differ and the same thing is not best for everything. Someone who does not understand the needs of the people they are about to GM for. And often they focus precisely on what differs most from M.A.G.U.S., making them much less likely than average to meet the preferences of the M* fan they are trying to convert.

Despite the pushiness and everything else, other D&D fans portray that fanboy as good, so after trying D&D many people understandably wonder: if this is good D&D, what is the bad kind like? Obviously, the same happens with the Storyteller and Storytelling systems and a few others. Disappointments like these, and the unpleasant experiences and anxiety that accompany them, gradually form a conditioned reflex. Partly to the appearance of d20 or a particular brand. By then the person is already nervous, their stomach clenches, and you can guarantee they will have a bad time. And of course the conflict described above earns them reassurance from people whose company they enjoy more than that of those who made them try D&D. 

Clutching the green book, however, immediately brings the relief of being able to try to enjoy themselves without that kind of harassment and suffering. M.A.G.U.S. can provide that pleasure automatically. Perhaps one day D&D 6 will be the world's best game; perhaps it would be perfect for this person too. But this conditioned reflex and the prejudice built from experience will still stand in the way when it comes to trying it. After recommendations like these, they no longer want to try anything else either. That is precisely why many people stick with “only M.A.G.U.S.,” regardless of whether they actually enjoy playing it. 

This holds whether D&D is wrong for them but the open-source foundations could make them happy, or whether a different introduction would have made the existing D&D their favorite game by now. It also makes no difference whether M.A.G.U.S. is a good or bad choice for them. Or even whether they play M.A.G.U.S. in a way that can work well, or are justifiably dissatisfied with results caused by their own decisions. 

Most M.A.G.U.S. players now insist on “k100,” with all its flaws. They are right that while a d20 gives you differences of 5%, and probabilities smaller than that are fundamentally difficult to handle with 1d20, 1d100 has finer resolution. But in many cases, no finer resolution is needed. Where it is needed, people become interested in rare events that should occur only once in several thousand rolls. A hundred-sided die lacks sufficient resolution for that. Multiple dice and moving away from a discrete uniform distribution would be essential here. 

Even D&D has moved away from that to some extent by now. Many people regard certain M.A.G.U.S. problems as universal. Others have found that they appear in far from every group: many are the fault of GMs, players, and bad examples circulating in the community, and M.A.G.U.S. works very well if you avoid those mistakes. It can even support life-path campaigns lasting several decades without the supposedly typical flaws appearing. But these longer life-path campaigns invariably mean sticking with an old system.

That system still becomes outdated. The fact that many new systems are worse in many respects does not mean they do not improve in detail, or that we do not discover new possibilities and solutions. Old systems cannot benefit from those.