Player stats, sales, and reception speculation thread

[Kwami]

It can't accurately look up how much sugar is in a Filet o Fish. I'm not worried about it demoting me to secondary status. No, that's not just around the corner.

It's not even close. This so-called "AI" has a few interesting applications, but it's way overhyped.

 

[bladex1200]

It's not even close. This so-called "AI" has a few interesting applications, but it's way overhyped.

I think a big issue with using it to develop video games in particular is that AI is good at approximating answers based on generally-available or specifically-trained data sets; it doesn't do real computation unless provided with particular helper tools to do so (e.g. "When I say 'call this function', this is what you should do exactly" can be added to a prompt and a helper tool function written and referenced specifically so an AI can do real computation without guessing). This is why AI out-of-the-box sucks at figuring out how much sugar is in particular foods, because that's not an answer you can approximate given a large amount of data: it's an exact number, and you're either correct or you are wrong when sifting it out of the pool of references.

When it comes to particulars like what makes for fun game design, the devil really IS in the details. Even if an AI could come up with a design document with mechanics that appear to be appropriate or approximately-correct, you'd need human designers and QAs to break the design and figure out whether the AI's general proposals make any sense and contribute meaningfully to a fun gameplay loop. As an example, an AI game designer might read a bunch of designs and weigh that a common pattern in successful game design is the presence of complex systems that provide the player with dopamine rewards in the form of EXP or unlocks. An AI might therefore constantly be sprinkling these into its design unless deliberately trained or warned against this behavior, because it associates these things with positive feedback (not in a literal sense but like being close in the linguistic or relational web of positive terms due to showing up in positive reviews and commentaries on particular games).

 

[The Civs 6]

I wonder about ai applied to play testing. I know before Open Ai went big they used to build very strong Dota AIs because they could train them on thousands of games.

Conversely Civ 7 could have used a lot of play testing.

 

[aieeegrunt]

I wonder about ai applied to play testing. I know before Open Ai went big they used to build very strong Dota AIs because they could train them on thousands of games.

Conversely Civ 7 could have used a lot of play testing.

All of their games could

 

[Gori the Grey]

Didn't mean to derail this thread into a discussion of AI.

(There's a thread for it in OT, where I've been making the same basic case, for anyone interested. (Against at least two enthusiasts; it's been a lively back-and-forth.))

 

[IntelligentDisk]

I think a lot of us actually want Civ7 to fail because if it fails there's the possibility the franchise will return with a future entry containing features we enjoy. It could also just die, that's a risk with any game failing in a franchise, but I think what many of us don't want is 7 succeeding and becoming the template for future entries (whether because you hate era-splitting, you hate civ-switching, you hate leader/civ detachment, etc). It just seems vindictive and mean to say so, so nobody wants to admit it.

But I don't see what's so wrong with admitting you hate a franchise entry and want it to crater. This is a free market, after all, nobody owes anyone anything product-wise. If the old adage to "vote with your wallet" is to be taken accurately, people who hate the game should be excited and pushing for wallot-voters to not buy it. And people who want the game to succeed should be excited that it is doing well. Which is why in this thread people constantly talk past each other and straight-up lie or cherry-pick or misread data to suit their preferred end.

Don't you think it's sad and overly resentful though? To be "excited" and actively push for people not to buy the game? This is a game which many people enjoy and hundreds worked on, but you want it to fail because it tried to innovate. Civ VII hasn't erased past entries - they're still there to be enjoyed. Isn't it better to keep criticism constructive and not destructive?

 

[SammyKhalifa]

Civ 5 is very strong. When aligned to the release date, it took six years until Civ 6 surpassed Civ 5. In 2030, it could still have ~10000 concurrent players.

Now looking back, there is a pattern: Civ 1-3 were classic civs, "classic trilogy". Civ 4-6 are modern entries, "franchise expansion trilogy" where expansion is not referring to the game but expanding the franchise to casual players and modern gaming culture. Baba Yetu, hexagons and cartoon graphics. Civ 7 is again a breakaway.

Civ III has "cartoon graphics." I mean, at least I think so, and with hardware limitations, etc., considered.

 

[itix]

Civ III has "cartoon graphics." I mean, at least I think so, and with hardware limitations, etc., considered.

I never played Civ III, but bitmap graphics drawn using a 256-color palette tend to suffer from that unintentionally. With 16 or 32 colors, you had to be more abstract. More colors give more freedom, but not enough, and you easily get a comic book feeling. It needs a lot of talent to get it right.

Civ VI is different; it is intentionally made in a semi-anthropomorphic style. Like, Philip II of Spain appears to be modelled after a dog...?

 

[queenpea]

An update on player numbers and Steam reviews:

The review average has ticked up positively to 47.17% positive (from a low point of 47.14%). In this metric, we are now seeing small movements in the positive direction.

However, at the same time, we have seen a dip in player numbers. The week of October 20 showed a 24 hour peak of 8078. The post-patch high showed a peak of 12,617 during the week of September 29. The October 20 number is more in-line with the low points we were hitting in early September, such as 8,791 during the week of September 15.

 

[itix]

An update on player numbers and Steam reviews:

The review average has ticked up positively to 47.17% positive (from a low point of 47.14%). In this metric, we are now seeing small movements in the positive direction.

However, at the same time, we have seen a dip in player numbers. The week of October 20 showed a 24 hour peak of 8078. The post-patch high showed a peak of 12,617 during the week of September 29. The October 20 number is more in-line with the low points we were hitting in early September, such as 8,791 during the week of September 15.

In my opinion it proves that game and its updates draw attention, but it cant keep players engaged.

 

[IntelligentDisk]

In my opinion it proves that game and its updates draw attention, but it cant keep players engaged.

This seems to be correct. It's the same problem Civ VI saw, only ever seeing week-to-week growth in the first year if there was an update or sale. New Civ games aren't as engaging to most Civ players until years later.

The review average has ticked up positively to 47.17% positive (from a low point of 47.14%). In this metric, we are now seeing small movements in the positive direction.

We're close to having back to back weeks above 50%, for the first time since the opening 2 weeks. This week is on course to having the 3rd best week - so far 91 positive, 87 negative, 178 total, 51.1% positive.

However, at the same time, we have seen a dip in player numbers. The week of October 20 showed a 24 hour peak of 8078. The post-patch high showed a peak of 12,617 during the week of September 29. The October 20 number is more in-line with the low points we were hitting in early September, such as 8,791 during the week of September 15.

I'm confused by "the week of October 20". Wouldn't that usually mean the week of October 20th - October 26th? It has only been one day, a Monday, so it's odd you'd choose this week of one day to use to compare, if that's what you mean. A better comparison is a Sunday to Sunday. Sunday 5th saw a peak of 12,617. 2 weeks later, Sunday 19th saw a peak of 10,404. That's a decrease of 17.5%. That's consistent with the 2 week drops from the previous updates, however it's the highest number out of the past 5 updates.

• 1.2.5: 12,617 to 10,404, 17.5% decrease
• 1.2.4: 11,871 to 9,670, 18.5% decrease
• 1.2.3: 11,758 to 9,595, 18.4% decrease
• 1.2.2: 11,911 to 10,066, 15.5% decrease
• 1.2.1: 11,707 to 9,652, 17.6% decrease

 

[bladex1200]

Don't you think it's sad and overly resentful though? To be "excited" and actively push for people not to buy the game? This is a game which many people enjoy and hundreds worked on, but you want it to fail because it tried to innovate. Civ VII hasn't erased past entries - they're still there to be enjoyed. Isn't it better to keep criticism constructive and not destructive?

Lots of things people hate get worked on with painstaking effort, that doesn't directly contribute one-to-one on their quality. Lots of films, tv shows, books, sculptures, etc contain the blood, sweat, and tears of their creators (and not just individual artists but editors, producers, etc - thousands of faces you will probably never give thanks to even if you do enjoy the product/artwork).

We wouldn't even be having this conversation (or, at least, it would appear to be much more one-sided) if this piece of media wasn't a video game. Nobody begrudges anyone for wanting a Michael Bay blockbuster Transformers movie to fail just because a lot of hardworking people poured their craft into making good CGI effects, explosions, and sourcing good costumes for everyone to wear.

I don't want Civ7 to fail because it tried to innovate, I want it to fail because the innovation it did doesn't suit my tastes and this is franchise I want to enjoy, and probably won't if the innovation continues in this direction. It's not destructive to say that this particular entry has failed, at least for me, provided you are providing salient alternatives that would get you to return for a future entry, which many complainers have done and continue to repeat.

 

[queenpea]

This seems to be correct. It's the same problem Civ VI saw, only ever seeing week-to-week growth in the first year if there was an update or sale. New Civ games aren't as engaging to most Civ players until years later.

We're close to having back to back weeks above 50%, for the first time since the opening 2 weeks. This week is on course to having the 3rd best week - so far 91 positive, 87 negative, 178 total, 51.1% positive.

I'm confused by "the week of October 20". Wouldn't that usually mean the week of October 20th - October 26th? It has only been one day, a Monday, so it's odd you'd choose this week of one day to use to compare, if that's what you mean. A better comparison is a Sunday to Sunday. Sunday 5th saw a peak of 12,617. 2 weeks later, Sunday 19th saw a peak of 10,404. That's a decrease of 17.5%. That's consistent with the 2 week drops from the previous updates, however it's the highest number out of the past 5 updates.

• 1.2.5: 12,617 to 10,404, 17.5% decrease
• 1.2.4: 11,871 to 9,670, 18.5% decrease
• 1.2.3: 11,758 to 9,595, 18.4% decrease
• 1.2.2: 11,911 to 10,066, 15.5% decrease
• 1.2.1: 11,707 to 9,652, 17.6% decrease

The data is from SteamDB -- I am just sharing the most recent information. SteamDB always looks at 24 hour peaks. These are always low in the week, with peaks on the weekends.

Basically, player numbers seem to have reverted to pre-patch levels. That's all it's showing.

 

[IntelligentDisk]

The data is from SteamDB -- I am just sharing the most recent information. SteamDB always looks at 24 hour peaks. These are always low in the week, with peaks on the weekends.

Basically, player numbers seem to have reverted to pre-patch levels. That's all it's showing.

But they haven't yet. They're still 628 to 1,433 higher depending on the day of the week.

• The 3 Sunday peaks in September before 1.2.5: 8,971, 9,217, 9,670. Sunday 19th peak: 10,404 (734 to 1,433 higher)
• The 3 Monday peaks in September before 1.2.5: 6,699, 6,927, 7,275. Monday 20th peak: 8,078 (803 to 1,379 higher)
• The 3 Tuesday peaks in September before 1.2.5: 6,320, 6,490, 6,745 Tuesday 21st peak: 7,373 (628 to 1,053 higher)

I don't want Civ7 to fail because it tried to innovate, I want it to fail because the innovation it did doesn't suit my tastes and this is franchise I want to enjoy, and probably won't if the innovation continues in this direction. It's not destructive to say that this particular entry has failed, at least for me, provided you are providing salient alternatives that would get you to return for a future entry, which many complainers have done and continue to repeat.

We're all very aware though that if Civ VIIs development did get stopped early like many people wish then that could be it for the Civilization franchise. Is this a side effect you'd accept in pursuit of actively wanting the game to fail? Isn't it best to keep criticism constructive and keep pushing feedback that you want the developers to hear so perhaps one day you too can enjoy Civ VII? I think the developers have shown they are listening given what we have seen in 8 months worth of updates.

 

[queenpea]

But they haven't yet. They're still 628 to 1,433 higher depending on the day of the week.

• The 3 Sunday peaks in September before 1.2.5: 8,971, 9,217, 9,670. Sunday 19th peak: 10,404 (734 to 1,433 higher)
• The 3 Monday peaks in September before 1.2.5: 6,699, 6,927, 7,275. Monday 20th peak: 8,078 (803 to 1,379 higher)
• The 3 Tuesday peaks in September before 1.2.5: 6,320, 6,490, 6,745 Tuesday 21st peak: 7,373 (628 to 1,053 higher)

We're all very aware though that if Civ VIIs development did get stopped early like many people wish then that could be it for the Civilization franchise. Is this a side effect you'd accept in pursuit of actively wanting the game to fail? Isn't it best to keep criticism constructive and keep pushing feedback that you want the developers to hear so perhaps one day you too can enjoy Civ VII? I think the developers have shown they are listening given what we have seen in 8 months worth of updates.

If you would just click around a bit more on the charts, you can see where they reference “week from October 20,” so I believe they are looking back at the week and calculating an average.

Maybe send them an email for clarification?

 

[IntelligentDisk]

If you would just click around a bit more on the charts, you can see where they reference “week from October 20,” so I believe they are looking back at the week and calculating an average.

Maybe send them an email for clarification?

It only shows that on mobile if you change the setting to max for me. No, they aren't looking back at the week and calculating an average, they just list the highest peak from the week. "Week from x" means x plus the 6 days after. You used a week of one day - a Monday. You would have to use the week from Monday 13th Oct to compare as that's a full finished week...

Anyway, as long as you're aware the player numbers haven't fallen to September levels yet.

 

[queenpea]

It only shows that on mobile if you change the setting to max for me. No, they aren't looking back at the week and calculating an average, they just list the highest peak from the week. "Week from x" means x plus the 6 days after. You used a week of one day - a Monday. You would have to use the week from Monday 13th Oct to compare as that's a full finished week...

Anyway, as long as you're aware the player numbers haven't fallen to September levels yet.

I don't think we know that yet? My point is that we have reverted to a pattern (the trend) of declining player numbers and we are around the September levels we were seeing before the patch was released.

I don't think it matters a great deal to look at individual weeks with heightened specificity. My aim was to point to the trend line, which has resumed a decline post-patch.

I guess I should have said that compared to late we are doing slightly better than September but worse than late August?

In the simplest terms: line go down.

 

[Xandinho]

It seems that this past week is the first one since launch in which the reviews are mostly positive (51%, with 77 positive and 73 negative). Of course, it's still very low, but could this be a sign of a turning point on the horizon? Let’s wait and see.

 

[canadamufc]

Some people will just not accept the fact that Civ 7 is failing.

When your newest game has half the playercount of a previous franchise entry released 15 years ago then it is obvious. And dont get started on VI where it is a 1/4. This has been going on for so long that we argue over minute numbers, in the end it's all the same the numbers are horrific.

Like the closest Civilization game to VII in current playercount numbers right now is not even V, it's Civ IV...

Just waiting for them to wrap it up and go back to basics for Civ 8.

 

[IntelligentDisk]

I don't think we know that yet? My point is that we have reverted to a pattern (the trend) of declining player numbers and we are around the September levels we were seeing before the patch was released.

I don't think it matters a great deal to look at individual weeks with heightened specificity. My aim was to point to the trend line, which has resumed a decline post-patch.

I guess I should have said that compared to late we are doing slightly better than September but worse than late August?

In the simplest terms: line go down.

What don't we know? We aren't around September levels. I literally just proved it with this:

• The 3 Sunday peaks in September before 1.2.5: 8,971, 9,217, 9,670. Sunday 19th peak: 10,404 (734 to 1,433 higher)
• The 3 Monday peaks in September before 1.2.5: 6,699, 6,927, 7,275. Monday 20th peak: 8,078 (803 to 1,379 higher)
• The 3 Tuesday peaks in September before 1.2.5: 6,320, 6,490, 6,745 Tuesday 21st peak: 7,373 (628 to 1,053 higher)

Did you not see this?

You don't think it matters a great deal to look at individual weeks yet you're looking at an individual day - Monday 20th.

Here are the past 20 Monday peaks dating back to the beginning of June:

1. 8,078 - 10/20
2. 9,040 - 10/13
3. 9,355 - 10/06
4. 7,178 - 09/29
5. 6,699 - 09/22
6. 6,927 - 09/15
7. 7,275 - 09/08
8. 10,022 - 09/01
9. 8,891 - 08/25
10. 8,554 - 08/18
11. 7,705 - 08/11
12. 8,528 - 08/04
13. 9,348 - 07/28
14. 7,834 - 07/21
15. 8,057 - 07/14
16. 8,945 - 07/07
17. 9,382 - 06/30
18. 9,532 - 06/23
19. 7,540 - 06/16
20. 8,355 - 06/09

The number is better than several across June, July & August. So far, we've reverted back to June, July & August numbers, not Septembers.