Just answering for myself, but it's because I like the programming/engine aspects. Not the creative aspects of the art/music/level-design/rules/balance/etc. I can use an existing games art/content and just focus on the part I like.

Plus nostalgia.

Some of my guesses:

Programmers are overrepresented both among FOSS enthusiasts and among nostalgic old gamers, meaning there are more people able to and interested in rewriting the engine code collaboratively than there are for the other parts of game-making.

People who want to do unpaid work on the other parts of game-making can do mods instead.

Making a new game is an artistic endeavour that is hard to do without either having an hierarchy making the decisions or being a solo/small group thing, making it somewhat unfit for community FOSS. When doing a remake you have the specification more or less done, you just have to implement it and do the uncontroversial fixes and improvements.

I wonder this too. Few hypotheses

1) great learning experience because you get fast feedback in terms of fidelity to the original, vs. the more amorphous goal of “good game”

2) nostalgia

3) creating an open source rendition is step 1 to the free-for-all modding utopia that people really wanted in the first place

Reminds me of Moonring that recently came out. It's an homage to the old Ultima games, but is new and looks and plays fantastic: https://store.steampowered.com/app/2373630/Moonring/

that was made by one of the original Dungeon Keeper devs btw

Wow wow wow… wow

The outcome and expectations are clear, this makes things infinitely more manageable than a new game. Also it's easier to find knowledgeable team members (when the fanbase was large) who are also easier aligned throughout the project compared to an open ended development.

> instead of, say, creating new games.

The Steam store is utterly littered with indie games on preview, so I don't know how you can say that with a straight face.

> I have never been able to fathom how people put so much effort into remakes of old games instead of, say, creating new games.

I've always held that when I retire, I'll commit to remaking Majora's Mask or Super Mario RPG. (Nintendo recently committed to the latter, but the tone is very different.)

Old games are a means of recapturing childhood nostalgia and reliving the fleeting feelings of youth. New games don't do that at all.

I'm not even a gamer anymore. I have next to zero time or desire to play new games. But I'd happily pick up something old to feel those feelings again, and I regularly watch YouTube videos of old games. They send you somewhere.

Old games have a pre-established fanbase, so releasing some kind of fix/patch/remake for one is almost guaranteed to make at-least some people very happy. The reception for an entirely new game is somewhat less guaranteed.

Because programming is an entirely different skill set to game design, art, music, marketing and all the other skills needed up make a game

Parent's late response to good sibling comments that no one will read:

Many people here seem to think that because one is good at programming they are necessarily bad at everything else like visual art, music, and story-telling. However I think many of such programmers would actually be in a great position to pick up one or even two such side competencies and do amazing things. Many of the old good games probably had programmers doing things outside their core competency. | Especially with the recent advent of AI, I think programmers should no longer think of themselves as narrow specialists but something like directors or conductors.