The Magpie Takes the Train Postmortem

I thought I’d record the basic process of making this game.

JJ Guest selected the prize I offered in 2018 for creating a game based in his world. This was the second time I had offered that prize, and because I hadn’t had enough time to finish my last such game to my own satisfaction (The Origin of Madame Time) I gave myself two years to finish this one.

The original design concept was suggested by JJ, with the Magpie riding a train car in a time sequence that required replay to get down right. The Magpie would change costumes during dark sections when the train was in the tunnel.

So I tried to make an outline first of obstacles to overcome. My concepts were that there would be the following difficulties:
-A rich person who had a ruby carved like a rose (based off of a series I read as a kid about a sapphire carved like a rose). This person would have 5 senses you have to block to be able to capture the rose successfullly.
-A dog who became suspicious of you and had to be placated.
-A rival in disguise
-A sharp-eyed attendant who required you to be in the right costume

Originally, there was no conversation at all and no outside section. JJ suggested replacing the dog with a bird, and that led to a parrot. I decided to create a few conversational topics for the parrot.

Then beta-testing began, and that led to enormous changes, mostly in that people said ‘If you can talk to the bird, why can’t you talk to everyone else’? Which led to more and more addition of conversational topics until they became a significant chunk of the game’s code, despite almost all of them being completely useless.

I also discovered that having several forms of liquid in the game was a terrible idea, as beta testers had the wildest ideas of things they could do with them. That’s why the game has a (also completely useless) drink mixing setup where you can take, for instance, vodka and grape juice, mix them together, offer it to everyone individually (with variations in each person’s responses based on the ingredients), pour it on everyone, drink it, or try to mix it again (with disastrous results). Absolutely none of it is useful, though, except for pouring it on one thing.

Testers found so many bugs with this game. At one point, the Viscount’s hat was able to hold your suitcase and all your clothing, and then destroying the hat removed all of your items from the game forever. For another set of testers, the attendant never made it into the train with you, but managed to comment on everything anyway.

The attendant, in my mind, looks exactly like Missy from Dr. Who. The antagonist, the Viscount, was entirely inspired by a wonderful Arsene Lupin story where he is repeatedly foiled by someone in a train car, only to later discover that his enemy is actually an infamous murderer. Cornelia Hogg is based on the Rockefellers/Carnegies.

Working with JJ on this game has been great. He was able to give me many tips regarding what the Magpie should look and act like, and commissioned the cover art for me. I believe that this game has generated more interest in this comp than my previous games because of the association with JJ’s brand and for his contributions to ‘marketing’, and for that I am grateful.

This game wasn’t designed to win, and won’t win, as it is too short. This was intentional, as I was finishing my Choice of Games novel during the writing of this game. I’m offering the same prize again, but for in two years from now, as next year I am entering a ‘trying to win’ game, which will be large and complex.


I’m not so sure :slight_smile:
Taco Fiction was also quite short but won in 2011. In any case, it will certainly do well.


Actually, this is the game I’d pick to win (the other top contender, in my mind, being The Impossible Bottle). In the nearly unanimous opinions of the reviews I’ve seen (and in my own opinion), there aren’t any real negatives to it aside from its length, and I’m not sure that’s even a substantial drawback. Judging from previous years, I’d expect the eigengame for the competition to be one that’s parser-based, about 60-90 minutes in length, is puzzle-based with moderately difficult but well-motivated puzzles, and has a tone that’s lightly comedic but not outright absurdist. Aside from its length, Magpie, fits the bill.

Of course, it should be said that Magpie is an exceptionally well-designed and -executed game that was a joy to play, regardless of how it does in the competition.


I’m taking note of this. And I love the word “eigengame!”


Yeah. I didn’t get a chance to play this one, but everyone I have seen just loved it. Obviously there are a ton of other games which were lauded for this cool thing or that, but there’s always “but then I didn’t feel this…” or that. Magpie seems to be short but deep, and I think @mathbrush’s comments about the work that went into “extracurricular” systems like all the conversation systems and drink mixing attest to that.


I loved the drink mixing aspect, and tried every possible combination. I’m also glad you included all the conversation options; NPCs who won’t talk to you are no better than machines when implemented in IF.


I agree on the ideal length for an IFComp game, 45-90 minutes is my personal preference. That feels like enough time to accomplish something significant, without overstaying your welcome. I probably shouldn’t feel like this, but if a game is listed as an hour long game on its comp page, but takes me two hours to get through I get grumpy. And if the game is a two-hour game then I expect quite a bit more of it than other games.

Magpie is my pick for best puzzles from beginning to end. There are other games (The Impossible Bottle or The Eleusinian Miseries), that had cleverer or funnier puzzles, but sometimes they also frequently had frustrating or incoherent puzzles that required hints. In Magpie I struggled with the puzzles (not too easy), but solved them all eventually without hints (not too hard) in less than 90 minutes (just right). My favorite games are the ones that I can do with very few to no hints, but don’t feel like a cake walk (The Wizard Sniffer is an excellent example).

I’d be disappointed if Magpie didn’t finish in the top 10, but the competition this year is really fierce so who knows.


I want to point out that the two Magpie games have taken first and fifth in their respective IFComps, as well as winning two of the last three Miss Congeniality awards. This is impressive success for a (thus far) two-game series!

While a lot of this success is due to J.J. Guest’s and Mathbrush’s game-writing skills, I think much of it is also due to the Magpie himself. J.J. Guest has created a very interesting character here. He’s even won his own XYZZY award for best PC.

The moral? One way to write very good games is to center them around a compelling character.

(Not that I have taken this advice to heart in my own games. :grinning:)


I agree. I’m working on my new game and I even thought, ‘Could I make this MC as compelling as the Magpie?’

And then I thought, ‘Nah.’ So it’s back to a bland author avatar for me lol


Well, an interesting central character isn’t the only way to write a good game! :grinning:


They’re certainly beating us Stiffy Makane authors.


I suppose Stiffy Makane is compelling in his own unique way…


Do I smell a possible Magpie anthology game?



I’m going to repeat below what I said to you in an email at half past two in the morning:

I’d prepared this whole load of long-winded twaddle about The Magpie Takes the Train to post here on the forum, but then I had a bunch of mushrooms and I realised that what I really wanted to say was WELL DONE! AND THANK YOU, THANK YOU, THANK YOU for a wonderful tribute and a wonderful bit of fun. I love you, mate.

Oh, and Mr Plotkin - an anthology, you say?

What a great idea! You’re on!