IFComp 2023 Predictions

As in 2022 and 2021, I made a spreadsheet to predict the IFComp results based on the IFDB ratings, the number of public reviews, and some information about the games. Basically, this is a simple linear regression model that predicts IFComp scores based on the results of previous ifcomps from 2018 to 2022.

The games are currently ordered based on predicted ranking, using the 2018-2022 model (there’s also a 2022-only model).

The top few predicted games are mostly what I would expect - I’d guess that LAKE Adventure would be first. Although I think there will be some different trends this year - last year, choice-based games overperformed a bit and parser games underperformed, which is not what I expect this year. Also, there seem to be fewer IFDB ratings and reviews compared to last year, so the predictions might be less accurate.

What are your predictions?

14 Likes

Wouldn’t it be better to do prediction after voting has ended?

6 Likes

Probably too early to do any sort of prediction right now, as the reviews are still coming in. After reviewing all 75 on my reviews thread (but not voting) I will give some further evaluations, but this will be done only after the deadline. Stay tuned!

4 Likes

Well I sure hope you’re wrong, but you’re probably right.

2 Likes

Some I am surprised about.

  1. Although I do like the game a lot myself and may be biased, I am still super surprised that PQ is not in the top five. The reviews for it have all been very positive and it also has that perfect 2 hour mark. It is so well implemented, as well. I don’t know. Just saying.

  2. I can’t tell specifically about LAKE Adventure, but the reviews seemed quite {confused about/unsatisfied with} about the game (or maybe nostalgic?). But I don’t know, since I haven’t played it.

  3. How did Milliways get put at 21st?! Have you played the game? (I guess I’m safe to downvote my own game, even if it isn’t a good idea to do so…

Now, just, one more question is
“Who wrote Dick McButts Gets Kicked in the Nuts?”

1 Like

I had given you an 85%! That’s very good, among the top 5 parsers!

1 Like

I know, and that was amazing and really surprising! But I’m so glad somebody liked it :slight_smile:
(I know others liked it, but that much was so great!)

Now just whether you round up or down for the rating…

2 Likes

I’d be surprised if Lake Adventure were number 1, just because historically when serious games have gone toe to toe with funny games of equal quality, the funny has won more often.

Similarly, I think Assembly might rank a little lower because the process it simulates is a little frustrating/repetitive (making and taking apart furniture), although everything else about the game is top-notch. Similarly, I think Dysfluent will be lower, because it simulates a frustrating experience.

I think Prince Quisborne will be higher. That’s what happened to ‘The only possible prom dress’, which was also very long; it had low predictions but high results (could be due to people entering scores in late).

I think Put Your Hand Inside the Puppet Head and Lunium will both be substantially higher (my personal guess is Lunium as the highest rated choice game, except I haven’t played Xanthippe).

11 Likes

I’m just surprised Dick McButt is so low :joy:

3 Likes

And Then You Come to a House… won in 2021 and Lake Adventure bears some resemblance to that game. It would be an awesome winning streak.

3 Likes

If Lake Adventure won I would find it to be very reasonable. I would be surprised a bit though, just because of the additional friction from using an older format or the dosbox thing, and the fact that the game is a lot less ‘lush’ than the other. I don’t think it’s bad at all, just different enough from the last one that I’m not sure it’ll place exactly as high, especially since the competition is pretty intense this year.

5 Likes

Parsers in general brought their game this year.

10 Likes

I agree with Mathbrush that Dr Ludwig seems likely to beat out Lake Adventure due to being more lighthearted—yes, And Then You Come to a House did win, but it wasn’t as heavy as Lake Adventure.

While both seem to have been rated on IFDB by a handful of people who really liked them, I would be surprised if Citizen Makane and Ribald Bat Lady Plunder Quest actually did as well as the predictions currently show. The reviews I’ve seen are mostly positive, but not quite “top 10” positive.

Also based on vague “my assessment of the general temperature of reviews” vibes, I suspect both Lunium and Fix Your Mother’s Printer will place higher.

9 Likes

Ooh, thank you so much for this. I never forget how generous this community is, but I sometimes forget how freakishly brilliant so many of you people are. This data set combines clever and kind in such a cool way.

As the Day of Results draws ever closer, I’m slowly going battier and battier even though I was pretty confident (having scored precisely 15th with a similar game two years ago) that “Bali B&B” would place somewhere between 10th and 20th place. I feel much calmer (for the next five minutes) seeing the predicted result of 17th.

This is also fascinating stuff, and I can’t wait to see where it’s utterly wrong this year.

12 Likes

I love this community so much. SPREADSHEETS FOR EVERYTHING!! I feel so seen.

Predictions seem like a random walk for me personally. Only 5 of my personal top 10 made top 10 last year - and a few were DEEP in the ranks. Based on a cursory look, seems likely to repeat this year.

I am not a number; I am a free man, OR, why you should never rely on ...

Though that does seem to advocate for taking a healthy dose of salt with my reviews!

See this is how I feel about Into the Lion’s Mouth! That hit me at precisely the right moment in this review cycle to completely charm the pants off me.

9 Likes

In 2022, I had 5 of my personal top 10 make it there and 4 of my top 10 placed from 30-39(!) So there is definitely some quasi-randomness and I think there should be, to account for personal tastes. Though on the other hand, in 2021, I got 8 of the top 10, and the two I missed on were really close. So things can be elastic! 2022 seemed to have a lot more weirdness/experimentation, though I enjoyed working through all of both 2021’s and 2022’s entries about equally.

It probably shouldn’t be like sports journalists voting on the top 10 college teams in the nation where they have a whole season of results to look at, though preseason rankings might be a better comparison.

I’d be interested in seeing a general list of people who’d voted over the years, if they wanted to share general data.

10 Likes

I’d be interested in seeing a general list of people who’d voted over the years, if they wanted to share general data.

Personally, I don’t record my scores except on the competition page, so I wouldn’t be able tell you what score I gave on a game last year, let alone years ago. All I have to remind me of the games I enjoyed are the reviews I’ve written. Most years I make an effort to distribute my scores along a flat distribution, as many nines as there are sixes and threes.

5 Likes

I think I have a suggestion for you… hear me out… SPREADSHEET those things! :]

11 Likes

Now that the voting is over I feel like I can talk more freely about predictions!

I think Dick McButts is going to rank higher than this spreadsheet suggests - perhaps even slap bang in the middle, which would be poetic.

The reason I think this, is because I think that while attempting to create the ‘bad route’, the author has inadvertently stumbled on making something that people actually like. It’s strayed from “so bad it’s bad” into “so bad it’s good” territory.

3 Likes

Dick McButts, meh. I played it, fairly early in the competition, but chose not to write a review (even in the private author’s forum) because I didn’t want to add to the large amount of publicity it had already gotten. I don’t know if I was seeing the “bad” version or not. I played several times and discovered (as others did, I’m sure) that I could survive longer and longer by making a different series of choices. Which casts the game as a “trial by dying puzzler”. That’s a style I wouldn’t automatically condemn, but in this case the ending was always the same crass punchline. I reached the back of the restaurant, before finally rage quitting. I have read that there was some additional content I missed involving Hitler and a parade, perhaps the first real test of the “Hidden Nazi mode theory of IF” (If you need an explanation for this reference, read the “Wheels within Wheels” review that comes on this IFDB page which is associated with Victor Gijsbers story “The Game Formerly Known as Hidden Nazi Mode”)

Ultimately I gave this game a low middle score (not a 1 or a 2) for a couple of reasons. First, because the game has some nice bells and whistles, and the author obviously spent more than a few minutes writing it. But second, the structure of the game (a randomly generated “bad route?”) suggests the author was deliberately angling for the Golden Banana, and I’d like to soften the standard deviation. Let that award go to a game that was more sincere.

5 Likes