2023 WIF AI Progress Report

[eouellet]

Hi Angeldust2,

The workaround one-map scenarios work good, especially the Day of Infamy is quite fun. The Fascist Tide is a bit cumbersome because of re-creating the Transfer Box, but it still works good. I don't know if any other player tried it.

I will have a look at the external app effort, although last time I did programming was literally in the last century!

[dan_hnnng]

I wish we could have the optional rules worked on instead of the AI.

[paulderynck]

It seems to me that 8 or so years ago the majority of opinions expressed were in favor of getting an AI. If you search the forum far enough back you'll find a survey that was done to that effect.

Now it looks like the reverse has become true. A possible explanation is that those who wanted an AI have moved on and the remaining more dedicated MWiF players have found satisfactory ways to continue their interest in the game.

[Angeldust2]

Quite possible. Maybe this is a situation, where adapting to today's reality and even preparing for the near future is more prudent than to stick to a plan, which does not fit anymore? And me and others were under the impression, that Erik Rutins as publisher has exactly outlined such a route in the state of the game thread, in 2019, so only 4 years ago, not?
If I understood him correctly at that time, he committed to first finish
- selected additional optional rules
- one-map scenarios
- ongoing bug fixes of essential bugs
before working on such a rudimentary AI as currently is the goal.

[marcusmccannis]

I'm just hoping the game continues to evolve in whatever direction and is not abandoned.
It is complex yes, but at least there is a community to support the outcome.

[juntoalmar]

It seems to me that 8 or so years ago the majority of opinions expressed were in favor of getting an AI. If you search the forum far enough back you'll find a survey that was done to that effect.

The game was released about 10 years ago. I guess it's reasonable to think that most people that answered the survey thought the AI may take 2-4 years to complete. But now we know that it has taken 10 years to start thinking of an AI for a couple of simple scenarios (if it happens at all).

The knowledge of the game complexity, the problems fixing bugs, and easier matters than the AI (like the convoy system) could probably draw a different result in the survey today.

Personally, I doubt that the people who play the game regularly right now (among of whom I include myself) would care for an AI for a small scenario, because I think the magic of this game is in global war multiplayer.

May finishing an AI for Barbarossa & Guadalcanal attract new players to the game? It's debatable.

As a player I say that developing the AI is against my interests (I'd rather fix bugs & convoy). But from the business perspective I completely understand: Matrix & Steve don't get anything from old players who have been playing for 10 years (well, economically speaking, morally the get the satisfaction of great job done) but only from new purchases. That would be different if this game had a subscription model where the profit comes from playing players, but that's not the case.

[lucliu]

The deadline is approaching.

[rkr1958]

The deadline is approaching.

What deadline?

[lucliu]

What deadline?

Schedule for 2024
The expectation is that the AIO will be playing the Russian side of the Barbarossa scenario fairly well by March 1st, 2024.

[rkr1958]

What deadline?

Schedule for 2024
The expectation is that the AIO will be playing the Russian side of the Barbarossa scenario fairly well by March 1st, 2024.

Got you. That's less than a week away ...

[lucliu]

Are there any updates regarding the 2024 schedule?
Steve was inactive in the past few months and it's already past the due date without any notifications.

[rkr1958]

Are there any updates regarding the 2024 schedule?
Steve was inactive in the past few months and it's already past the due date without any notifications.

Good question. I post the following without comment, so the folks may draw their own conclusion(s).

1. Just checked the beta FTP site. 6.00.00.03 (10/16/2023) is the last update for the beta testers. No notification post in the beta tester forum of that update. It was only "discovered" by a tester logging into the FTP site almost a month after the update was uploaded. Since then I check the FTP site at least once a week to see if a later update has been uploaded. As of a few minutes ago, 10/16/2023 is the latest update.

999-2024MAR-Status-1.png (127.69 KiB) Viewed 1046 times

2. Last login for Steve was Mon Nov 20, 2023 6:67 pm.

999-2024MAR-Status-2.png (46.75 KiB) Viewed 1046 times

3. His last post was Mon Aug 14 2023 11:50 pm.

999-2024MAR-Status-3.png (60.34 KiB) Viewed 1046 times

[lucliu]

Thank you for sharing the information, Ronnie.

[Joseignacio]

[eouellet]

Thank you Ronnie for this info.

This is more pointers for me that there is a serious MWIF fatigue, but how can anyone place a blame for it, after working on this game for nearly 20 years! Very few would have kept going for so long.

I went to read Steve very first post about MWIF, dated May 19, 2005, here is an except:

The learning curve for WiF is one of the steepest and longest for any board game I have played. The brilliance of its design and the extreme pleasure that comes from playing it lies in the intricate weave of its tapestry composed of the map, units, and rules. The AI opponent will have to understand the tapestry as well as the players it competes against or it will just be another "stupid AI".

[Erik Rutins]

Hi everyone,

Steve has continued working on the AI and we've been checking in with him bi-weekly. Without a doubt, the WIF AI has proven more challenging than he expected but he has not given up. He's working through Barbarossa AI decisions phase by phase and when that's done we will have a public beta. I expect there will be a few more internal test builds before then.

Regards,

- Erik

[Joseignacio]

Thanks, Erik

[marcusmccannis]

Well, that's a promising statement.

[marcusmccannis]

Well, that's a promising statement.

[jaeenox]

Hi everyone,

Steve has continued working on the AI and we've been checking in with him bi-weekly. Without a doubt, the WIF AI has proven more challenging than he expected but he has not given up. He's working through Barbarossa AI decisions phase by phase and when that's done we will have a public beta. I expect there will be a few more internal test builds before then.

Regards,

- Erik

Another month gone. Is there any news? Does Steve have any deadlines to meet or is it just an open ended contract to finish whenever? Perhaps he could use some help from a an experienced AI programmer? Does MG have any expectations? Will it be this summer, this year, 2025, etc.