20 days have passed, did they fix SBMM?

Below is simple outline of the patent. It does not mean they have implemented all or to what extent or even the values of the algorithm. But this was the general idea. We have no visibility of true extent.

https://arxiv.org/pdf/1702.06820.pdf

"Measure player's disengagement by their churn risk"
- Means how likely they are to stop playing based upon certain time period. They indicate example of a week, but reality is that if you stop playing for an hour, you are put into an easy lobby to create engagement for further games. Churn risk = State of engagement. There is a numerical value assigned to a player dynamically.

"Where each player is a node, and an edge between two players is their sum churn risk if paired"
- Means put low engagement players without each other in lobbies. This can be manipulated by smurf accounts as new accounts are locked into easy lobbies until certain rank is reached, just means stuck with season 1 Legends and 15 min tutorial. I lost my PC account and my new PC account after over a week have not seen one player over rank 20.

Example of churn risk

Win, Lose and Draw
Last 3 outcomes as example
WWW = 4.6 to 4.7% - Means they won 3 times and low risk of disengagement
LLL = 5.1% - Means they lost 3 times and at higher risk of disengagement

"The optimization objective can be tuned for various interests e.g. in game time or even spending"
- Spending? Psychological wise does this imply a player who enjoys the lobbies will spend more on the store than someone who plays the game less?

"For example a low skill player beating a high skill player yields a large update in adjusting their skill"
- Elo system that increases your rating of skill depending on the skill gap with your opponent. This could mean anything from rank difference, KD, kills, death etc. No indication of "How" in the paper of what stats.

"This may suggest that when the player pool reaches certain size, the choices of opponents are enough to rescue those players
on the edge of churn."
- Lets say a Master/Apex Predator throws away 3-5 games in a row. Does that mean they will put into easy lobbies as well as losing 5 times in a row is considered high churn risk regardless of their badge.

"Moreover, we can even change the objective function to other core game metrics of interest, such as play time, retention, or spending. EOMM allows one to easily plug in different types of predictive models to achieve the optimization."
- Again mention spending? A relationship between EOMM and store spending.

All the useless mathematics in the paper says you can simulate engagement of players based on recent success.

Sounds very intelligent. But here is reality.

You can manipulate match making

- Wait until dropship counts 0 to leave or suicide. Play a game, start a no fill suicide game, Play normal. Alternate or there is a certain number of times you contiuously do that to lower your churn risk. Use no fill out of courtesy to other players when doing suicide.

- Stop playing for an hour or even switch to another account and back and forth to give impression of lower engagement.

- Premade and random squads are put into same queue. Choose a server where local time is non social hours like 3am. There is less chance of premades and sweaty players at the cost of higher latency and longer match making time.

- You are rewarded by doing hot drops, 1v3 pro wannabe, down quit etc. Several losses like that puts your churn risk higher by dying and not one kill.

Success of a game is not based on it's own merit alone. It is because sometimes we play certain games not because we enjoy it but there is no alternative. Lesser evil so to speak. If there was a BR game a person enjoys more they would move over regardless of how good the previous game was. Success in this case also based on comparison.

This paper is nothing new in what we already know. But wonder about whether if or maybe spending is associated into EOMM match making.

Just remember it takes a creative designer 1-2 days to create a skin. Lets say a skin costs $10. Some AAA RPGs takes a whole studio and 3 years but charge $50.

Gambling - A way to take more money than normal trading. Open 500 Apex loot boxes etc. They know simply letting people buy what they want directly won't yield as much. Once you hit 500, you'll obviously think well I've invested this much so what is a litte bit more. Or it is only $10, whats a little more so it continues without a ceiling limit as a complete total over X number of months.

Since creating a skin is so lucrative in money, why bother trying to creating anything else of value. Developers are not destroying gaming, it is players themselves who buy into it. Yea OK "Show them support", might as well support a landlord. Better to support indie developers or other studios.

Just play the game as it was intended. Free to play.