Take Two Interactive (Rockstar, 2K, Private Division) canceling games, ending projects and laying off 5% of its workforce

[herbal space program]

1 hour ago, Lisias said:

You see, given this margin for errors, 1% is not that off.

OK, so if you cherry-pick the data you can make what @PDCWolf said only a fourfold exaggeration rather than a tenfold.  Do you think I was born yesterday?  I'm a scientist, with 35 years of experience both trying to demonstrate what the truth is and trying to bust people who distort what the truth is to their own ends, having reviewed countless manuscripts in my career.  It could not be more clear to me that both of you are playing fast and loose with the actual data to support your predetermined conclusions. What I want to know is: why? Don't bother with any more obfuscatory nonsense. I am many times over wise to all of that.

 

And since I made rather a strong claim, here is more data to support it: SteamSpy: 16.4%, Gamelytic: 21.2$%, VG insights: 12.1%, Playtracker: 3.85%. Average: 13.4%.  Which of those is the outlier, eh? And why? I can offer a reasonable hypothesis: Playtracker is not based on actual sales at all, but rather models sales based on concurrent player numbers. A game like the KSP2 EA will of course look particularly bad by that measure, because a lot of players who bought it have since set it aside to wait for more content and more bugfixes, if and when those come.  And of course that's 15 months of total sales vs. ten years.  A thoroughly dishonest representation of reality, no matter how you slice it.

Edited May 17 by herbal space program

[PDCWolf]

58 minutes ago, herbal space program said:

That statement is demonstrably untrue: https://steamdb.info/app/954850/charts/.  Estimated KSP2 sales were about 10% of the similar estimates for KSP1, not 1%. Still nothing to crow about, but I have to ask myself, why the tenfold exaggeration on your part?

My bad, missed a 0. Even then it changes nothing about the argument.

[Lisias]

38 minutes ago, herbal space program said:

Do you think I was born yesterday?

Your arguing sounds like that, no doubt.

 

38 minutes ago, herbal space program said:

And since I made rather a strong claim, here is more data to support it: SteamSpy: 16.4%, Gamelytic: 21.2$%, VG insights: 12.1%, Playtracker: 3.85%. Average: 13.4%.

Average? You are using average to support your claim? You still have a margin of error between 3.85% and 21.2%!!

If you cared to check at least ONE of then, you would find a very, very interesting data:

Stats

Copies sold:

557.9k (278.9k - 836.9k)

Gross revenue:

$25.6m ($12.8m - $38.4m)

Source: https://gamalytic.com/game/954850?utm_source=SteamDB

Do you see that two values between parenthesis on copies sold and gross revenuw? What do you think they mean?

You are averaging numbers already averaged, damn it!

Edited May 17 by Lisias
Hit "Save" too soon

[Fizzlebop Smith]

Double mean precision... to handle those pesky floats? Remember how we ALL hate those pesky floats...

Time to hold hands and sing in a circle.

[herbal space program]

8 minutes ago, Lisias said:

Average? You are using average to support your claim? You still have a margin of error between 3.85% and 21.2%!!

What we have is an egregious outlier, with a reasonable hypothesis to support why it is that and so might legitimately be excluded.  But even setting that aside, as you so superciliously pointed out, those numbers in the parentheses are the 95% confidence intervals, even with the distorted distribution caused by the outlier, and what PDC Wolf claimed was outside of even those. Riddle me that. They just apologized and called it a rounding error, LOL. How about you just stop promulgating disinformation?

Edited May 18 by herbal space program

[Ker Ball One]

6 minutes ago, Fizzlebop Smith said:

Double mean precision... to handle those pesky floats? Remember how we ALL hate those pesky floats...

Time to hold hands and sing in a circle.

All these numbers float.  And when you're down here... You'll float too!

Edited May 18 by Ker Ball One

[Lisias]

1 hour ago, herbal space program said:

What we have is an egregious outlier, with a reasonable hypothesis to support why it is that and so might legitimately be excluded.  But even setting that aside, as you so superciliously pointed out, those numbers in the parentheses are the 95% confidence intervals, even with the distorted distribution caused by the outlier, and what PDC Wolf claimed was outside of even those. Riddle me that. They just apologized and called it a rounding error, LOL. How about you just stop promulgating disinformation?

You are claiming a 95% of confidence on an average number between (278.9k - 836.9k) ?

[herbal space program]

16 hours ago, Lisias said:

You are claiming a 95% of confidence on an average number between (278.9k - 836.9k) ?

So you don't know what confidence intervals are? Do you know what a normal distribution is, or a T-test? [snip] ridiculing my statistical reasoning is all I can say! They arrived at that confidence interval by imputing  a normal distribution on those data, including the major low outlier you guys cherry-picked, which is why it is so broad. If they had eliminated it, as they should have on methodological grounds, the distribution would have been a whole lot tighter. Again, please stop promulgating disinformation!

Edited May 18 by Vanamonde

[m4inbrain]

59 minutes ago, herbal space program said:

So you don't know what confidence intervals are? Do you know what a normal distribution is, or a T-test? 

Do you know the difference between dis- and misinformation? 

As a "scientist" trying to propagate "the truth", i find it rather interesting that you immediately jump to the conclusion that people maliciously spread disinformation rather than making a mistake, regardless of the topic at hand. 

Tells more about you than you might think.

You also seem awfully certain that these outliers should be excluded, with no actual reasoning. "Methological grounds" is not sound reasoning to remove outliers. 

[JunaidWolf]

So how many games have been canceled was gamefreak game one of them

[JunaidWolf]

Do we know how many games have been canceled and if gamefreaks new project was one of them

[Lisias]

12 hours ago, m4inbrain said:

Do you know the difference between dis- and misinformation? 

<...>

rather than making a mistake, regardless of the topic at hand. 

Granted, I my had made a mistake. This is my reasoning.

On my post, I took the "less favorable" KSP2 stats as well the "most favorable" ones, observed the gap and established that we have a huge margin for errors, from KSP2 selling ~3.58% of the KSP¹ copies to ~16.48%.

Quote

• Steam Spy
  • KSP¹ : 3.58M
  • KSP2 : 590K
  • Ratio: 0.164804469273743 , or ~16.48%
• Play Tracker
  • KSP¹ : 6.24M
  • KSP2 : 240.7K
  • Ratio: 0.03857371794871795 , or ~3.85%

I compared the values from the same sources. I used Play Tracker's KSP2 values against KSP¹, then I used Steam Spy's to do the same.

I'm not averaging values from different sources because I don't know their sources neither the methodology they used to reach these values. I just assumed all them are equally competent to report these values, but without knowing their sources, I could irresponsibly use the same mass of data more than one time, contaminating the sample space. So I ruled out averaging the different sources reports on the spot (seriously, this should be a no brainer).

Where is the error on declaring that we have a margin for errors from 3.85% to 16.48% ?

Edited May 18 by Lisias
Tyops are cosnpiring agaisnt me!!!

[herbal space program]

12 hours ago, m4inbrain said:

As a "scientist" trying to propagate "the truth", i find it rather interesting that you immediately jump to the conclusion that people maliciously spread disinformation rather than making a mistake, regardless of the topic at hand.

Immediately jumping to conclusions? I have been reading the endless doom propaganda getting posted [snip], replete with  belittlement and ridicule of any who disagree,  for weeks upon weeks now, and I am tired of it. Anybody who makes utterly false statements like the one I called out above in bold allcaps with multiple exclamation points is peddling disinformation, not just misinformed, period. And I am in fact a scientist, and do care very much about the truth, although I don't really give a hoot what  95-posts-in-ten-years you thinks of me.

Edited May 18 by Vanamonde

[Stephensan]

a good reminder for the people reading, take a break and drink water, the story will continue until the truth is out for the public.

[Vanamonde]

Things were getting overly personal and some comments have been removed. Please back off of the tempers. 

[Bej Kerman]

2 hours ago, herbal space program said:

although I don't really give a hoot what  95-posts-in-ten-years you thinks of me.

I mostly sympathise with you, but I have to point out that if we're measuring others' worth as human beings by post count and frequency, then I outrank you   

[herbal space program]

23 minutes ago, Bej Kerman said:

I mostly sympathise with you, but I have to point out that if we're measuring others' worth as human beings by post count and frequency, then I outrank you   

I'm not measuring anybody's worth as a human being by their post count, just the value of their opinion of me in a weeks-long debate in which they have not ostensibly participated. That and my suspicion that they are in fact a ten year-old sock puppet that has been dusted off to be used by somebody  who feels they need a little virtual backup. As to outranking me, I only lay claim to leadership of the Chimpanzee division! I would never challenge your clearly higher rank among the Kerbals.

[K^2]

On 5/17/2024 at 3:37 PM, Lisias said:

You see, given this margin for errors, 1% is not that off.

This analysis is what "Lies, damn lies, and statistics," phrase was coined about. You've taken the difference between best estimate for KSP1, worst for KSP2, and said, "Yeah, 3.8% is not far off from 1%".

I mean, even that is closer to 10% than to 1% in order of magnitude, and this is after you've basically taken the opposite ends of the bell curve to try to make it work. You're taking 5% of 5% in probability, and it STILL isn't supporting your claim. The lowest 0.25%-likely estimate is entirely too high to support the 1% claim. This is gross misuse of the stats that's either ignorant or dishonest. Neither of these is excusable to continue pushing after you got called on it.

 

The fact that KSP2 likely (as in, P>90%) sold in a year more than 10% of KSP1 sales in over a decade, is a great indicator of how KSP2 was on as good of a track financially as PD hoped it would be as an EA title. Yes, they needed full marketing push and a console release to actually make it a financial success, which would require it to be in a better shape than it is now, but the risks involved are minimal. A competent team can finish it. Even if you don't think Intercept was running the team competently (highly debatable, but even if), PD just needs to put someone they trust in charge of Intercept, finish the project, and rake in the revenue. Even compared to T2 budget, which is orders of magnitude higher, you're really asking, "Should the company have invested another $20-$30M in a project they've invested ~$50M to have a guaranteed return in $50M-$200M range after console sales with no risk?" The answer is yes. This is the exact situation for which PD exists as a division within T2.

There is zero sense in canceling the KSP2 project based on performance of Early Access. All of the reasoning for the layoffs and projects going on ice are in the financials that were just presented, which show monumental losses suffered by Take Two during the 2024FY. Yes, they took in record revenue, but they have way over expanded in 2022-2023FY, as shown by the huge increase in their operating expenses, far outpacing the growth of revenue. For a company of T2 size, the only sensible response is to shut down basically everything PD is working on that doesn't have imminent release. Since KSP2 wasn't planned to go 1.0 in 2025FY, it got the axe. It's that simple. It will be a little bit more expensive to restart the project in a year, but the company is bleeding money now and the loans are absurdly expensive. A year from now, federal interest rate is expected to be more reasonable, and a bit later in 2025, GTA VI ships. If nothing comes up earlier, there is basically zero chance the KSP2 isn't back in development then. And smart money is on T2 having some skeleton crew working on it from now until then - not to make the progress so much as keep it patched up and retain experience. And we see evidence of that in the movements in the game's repositories.

All the nonsense about, "Take Two finally got fed up with mismanaged project," is misinformed, ignorant, and incongruent with any of the available data, from the financials, to the game's performance, to how mismanagement has been historically handled not only at T2, but at other publishers in similar situation, to messaging around the situation, to the fact that there is still some sort of activity on the game's repositories. The amount of double-think you have to get into to believe that T2 has just been twiddling thumbs for the past four years as Intercept ran amok, and then suddenly, having seen the not yet released game make something in the $10-$20M revenue, decided to just axe the entire studio for mismanagement letting go of absolutely everyone involved, is bordering on flat Earth conspiracy absurd. You have to both believe that T2 was happy to sign off on ~$50M in costs without checking in at all, and then suddenly start carrying when it turned out that it has only cost them 80% of that, decided that nobody at the studio was worth keeping hired, and fired absolutely everybody. This hypothesis doesn't survive the bare minimum of critical thought. Especially against the Occam's Razor of, "T2 just lost over $3B and needs to cut anything non-essential," which is what actually happened, and what T2 said happened, and what everyone in the financial circles reported has happened.

[Lisias]

20 minutes ago, K^2 said:

This analysis is what "Lies, damn lies, and statistics," phrase was coined about. You've taken the difference between best estimate for KSP1, worst for KSP2, and said, "Yeah, 3.8% is not far off from 1%".

I mean, even that is closer to 10% than to 1% in order of magnitude, and this is after you've basically taken the opposite ends of the bell curve to try to make it work. You're taking 5% of 5% in probability, and it STILL isn't supporting your claim. The lowest 0.25%-likely estimate is entirely too high to support the 1% claim. This is gross misuse of the stats that's either ignorant or dishonest. Neither of these is excusable to continue pushing after you got called on it.

I'm not trying to make anything work.

I just mentioned that 1% is not that far off from 3.8%, the bottom line of the estimation. I didn't said the dude was right on that 1%, I just said it was not that far off the 3.8% bottom line, I didn't claimed it was a new bottom line. I just said what I said, it was not that far off, without any judgement of value.

I fail to understand why so much grudge and vitriol about something that may be, at worst, a misunderstanding about what I was meaning to say.

Again, I just said that with a so gross margin for errors (from 3.85% to 16.48%), that 1% was not that off. English is not my native tongue, but as far as I understand saying that something is not THAT off is still meaning that this something IS off.

As far as I seeing, you being (apparently at least) a native English speaker may be being ignorant or dishonest about the representation of what I'm saying.

-- -- -- POST EDIT -- --

Apparently, this drama started with this post:

 

Edited May 19 by Lisias
post edit

[K^2]

Just now, Lisias said:

I just mentioned that 1% is not that far off from 3.8%, the bottom line of the estimation.

It's literally closer to 10% than to 1%.

3.8% is 38% of 10%

1% is 26% of 3.8%

And then you claimed that it's not that far off, "given this margin of error,"  with the entire margin being on the opposite side - towards 10%.

Again, this is either ignorant or dishonest. At best, your successful argument that you understand how percentages and errors work would make it the latter instead of the former, and I'm not sure what you're gaining with that.

[Lisias]

16 minutes ago, K^2 said:

It's literally closer to 10% than to 1%.

3.8% is 38% of 10%

1% is 26% of 3.8%

And then you claimed that it's not that far off, "given this margin of error,"  with the entire margin being on the opposite side - towards 10%.

Again, this is either ignorant or dishonest. At best, your successful argument that you understand how percentages and errors work would make it the latter instead of the former, and I'm not sure what you're gaining with that.

Dude, you need some sleep. And perhaps a good chat with a good friend.

If I have 1000 apples, and someone says me that between 3.85% and 16.48% of them are rotten, I have a loss somewhere between 38.5 and 164.8 rotten apples.

If some other dude by some reason guessed that perhaps about 1% of the apples were rotten, then that guess would be about 10 rotten apples.

Now please tell me: 10 is a number more near 35.8 or more near 164.8?

Anyway, I think that enough is enough. At this point, absolutely nothing that I can say will change anything, so the best option is just to... Ignore. I see no reason to keep feeding the trolls - and some people here are trolling.

Edited May 19 by Lisias
Tyops are cosnpiring agaisnt me!!!

[K^2]

3 minutes ago, Lisias said:

Now please tell me: 10 is a number more near 35.8 or more near 164.8?

Here's how critical thinking works. Imagine that the lower bound was 15.5% instead of 3.85%. Here's your question modified to that situation.

"Now please tell me: 10 is a number more near 155 or more near 164.8?"

And of course, we get the same answer. 10 is more near 155. So we conclude that 15.5% is close enough to 1%

And what if both lower and upper bound were higher?

"Now please tell me: 10 is a number more near 998 or more near 999?"

Of course, it's more near 998. So we conclude that 99.8% would still be acceptably close to 1%

Nonsense, right? So is your original claim and your apples question and for the same exact reason.

 

The number you are comparing 1% to is 10%. And the question is, is 3.85% closer to 1% or to 10%. Do your example with apples again, and have a think about it.

[NH4Cl Enthusiast]

5 hours ago, K^2 said:

It's literally closer to 10% than to 1%.

3.8% is 38% of 10%

1% is 26% of 3.8%

And then you claimed that it's not that far off, "given this margin of error,"  with the entire margin being on the opposite side - towards 10%.

Again, this is either ignorant or dishonest. At best, your successful argument that you understand how percentages and errors work would make it the latter instead of the former, and I'm not sure what you're gaining with that.

By your logic 0.1% is closer to 99% than it is to 0.0000001%. 

0.1% is approximately 0.1% of 99%.

0.0000001% is 0.0001% of 0.1%.

So with your reasoning it's several orders of magnitude closer!

You don't understand that "closer" is not well defined but instead just go ahead guns blazing claiming anyone who disagrees with your subjective interpretation is dishonest or ignorant. Do you often find that people back out of arguments with you? Here's a tip: you're not winning the argument, people just don't bother.

 

 

[K^2]

18 hours ago, NH4Cl Enthusiast said:

By your logic 0.1% is closer to 99% than it is to 0.0000001%. 

Yes. Because "500 people got the game" is closer in every practical way to "Almost everyone got the game" than to "Nobody got the game," which is where your percentages land.

Logarithmic scale is the only correct way to compare the fractions, which is what we're dealing with. That's how you get statistically significant data out of what otherwise would be just noise.

So your "not well defined," applies to armchair philosophers only, while for anyone who actually had to work with data, especially vague data with large uncertainties, the concept of "closer" is very much strictly defined here. The fact that it leaves some lay-persons confused is a problem for the lay-persons. Being uninformed about something doesn't make you less wrong when you say something incorrect. In fact, that's usually the cause.

[NH4Cl Enthusiast]

13 hours ago, K^2 said:

Yes. Because "500 people got the game" is closer in every practical way to "Almost everyone got the game" than to "Nobody got the game," which is where your percentages land.

Logarithmic scale is the only correct way to compare the fractions, which is what we're dealing with. That's how you get statistically significant data out of what otherwise would be just noise.

So your "not well defined," applies to armchair philosophers only, while for anyone who actually had to work with data, especially vague data with large uncertainties, the concept of "closer" is very much strictly defined here. The fact that it leaves some lay-persons confused is a problem for the lay-persons. Being uninformed about something doesn't make you less wrong when you say something incorrect. In fact, that's usually the cause.

[snip]

What you just said is that in your opinion then if GTA 6 sells 500 copies instead of over 10 million, it's closer in every practical way to their sales targets than "Nobody bought the game" which is pretty much where my percentages land. 

But maybe you could share that very strict definition of "close" that apparently is present in your highly specialised field of work? For us ignorant, confused, uninformed, wrong and dishonest lay-person armchair philosophers who just communicate using common language in a game discussion forum it's pretty vaguely defined.

Edited May 20 by Vanamonde
tyop