Looking more and ore like another grave dug. 

People also need to understand that seasonal games will basically always have falling numbers throughout the season, as people drop off(to come back the next season) and new players are unwilling to join before the new season begins.

For example this is what the numbers in a random season of Path of exile(new seasons require a new character where everyone starts equally) look like:

Now if you zoom out the chart to a period of years it looks like this:

Path of exile is a much lower scale game but it’s very good for showing player behaviour patterns. The game went through basically years of constant long term growth but after the start of each season the numbers would fall constantly. Week-on-week numbers would be down 90% of the time over a period of years, but overall it would be growing. 

The same can be seen in something like Diablo IV if you want an example from a large AAA studio:

Big peaks at the start of the season then constant declines.

What I’m trying to say is that the current decline in numbers is no where near as important as what the numbers look like at the beginning of season 2 and season 3. Once we see around 3 seasons launch we can probably get a good idea of where Marathons long-term playerbase is. 

Now, don’t get me wrong, I’m not saying I expect Marathon to exhibit immediate long-term growth. There’s definitely issues with retaining players, especially more “casual” players if we want to use simplistic labels. The onboarding is not great and it can take a long while to really get a feel for the game. If it can get a peak of 50k concurrent for the start of season 2 then I think it’s in a decent spot. If the season comes out and the player numbers do not show a significant peak, then yes, that’s probably going to be a massive problem. At that point we might be in territory where active development stops after, say, one year of support and the game servers are kept up rotating existing seasonal content and whatever content was left to go when active development ceased.

Also, one last thing people need to understand is that concurrent numbers are only useful to a certain extent. You can’t read into them too much as they don’t provide the kind of granularity that some people try to interpret from them and number changes. The biggest problem is that a lot of factors affect them, mainly player behaviour and timezone splits. To highlight this you can look at two games: Delta Force and Helldivers 2. Yesterday on Steam Delta force had a peak concurrent user count of 140,410 and HD2 had a peak of 41,809. Here’s the thing though, more people played Helldivers 2 on Steam, than played Delta force during that time period. HD2 was 20th for daily active players on Steam and Delta Force was 32nd. Now this is an extreme example because it basically involves the full extremes of factors effecting CCU’s. Delta force is played in basically one territory: Asia(primarily China which is a single timezone) and it has players that are plying it for many many hours every day whereas HD2 has a lot more people dropping in for some quick play and leaving. The more individual players play the higher the concurrent numbers will be. The numbers for Marathon were likely being kept high by how engaged a certain part of the playerbase is, the people that have already played 100+ Hours. If those people pause playing until the new season then it has a huge effect on concurrent numbers over the course of a day. 

Doubt this will still be going by the end of the year. They've already lost 70% of there players.

No clue, but the budget shouldn't be much less than that.

It took years to develop and Bungie devs are not cheap at all. Money pile up over time easily.

That's why some smaller studios can keep a game with low volume of players for a good amount of time, to add content to it or for a community to form around it until it gets bigger.

Bungie can't. Sony will be taking extended substancial losses until it turns around, if it ever does.

So, unless there are plans that Sony sees potential and are worthy letting it live for a while, they'll do it, if there is not, or if those plans fail when they are implemented, the game will get canned.

forbes is a content farm. most of what paul tassi writes is just cribbed from any random reddit post he saw that day just like every other gaming related content farm.

Actually are live service games where these numbers are most useful. Given that you only get Steams numbers, which can be misleading for multi-platform titles. I agree though, that there can be artifacts of player behaviour. As always: data needs proper interpretation.

Reach was good, and I did play it. But there were cracks forming. And it certainly wasn't the event that Halo 3 was.

I'm guessing that their successes like HD2 are making more money than their failures.

Like it or not, a successful live-service game makes orders of magnitude more money than even a single-player game like Hogwarts Legacy. Roblox along is a double-digit percentage of the entire game industry.

All this "expert takes" on marathons failure is all based on refreshing steam chart players numbers every minute.

Ok lets see -

Grounded 2 - less than 3k players on steam ATM.
Sea of thieves - less than 7k players on steam ATM.
Halo infinite - less than 3k players on steam ATM.

The Finals - 14k players on steam ATM.

Marathon - 22k players on steam ATM

All way below Marathon numbers so I guess those are all gigantic failures based on the obsession with steam chart numbers logic? but let me guess somehow steam numbers don't matter for these other live service games only marathon? what's even funnier is half those games I listed is FREE TO PLAY and still has less player numbers than marathon that is $40.

That's a really bad comparison since Marathon is just a month old and Grounded 2 is the only one among the others that isn't years old. Exactly a week ago it peaked at 36k but now might dip below 20k in the next few days so if the player base doesn't stabilize soon then things will be looking very rough for it in a few months, let along a year from now. Dropping below 30% of its all time peak in just a few weeks is not a good start.