Preface: I am writing these in simple, product backlog item (PBI) format in the event that is helpful to the team managing intake and grooming activities. General format is as follows:

 Story: As a filter, I must X, so I may Y.
 Rationale: Paragraph of player rationale for the suggestion.

Questions or request for clarification from SEGA or PSO2 welcome.


OVERVIEW:
There are a few, simple criteria that SEGA can put in place for all free account creation that can make automated filtering/suspension of suspicious activity accounts that are RMT spam bots much easier. In the following, I am outlining individual filters that, working as one, should quickly cull most of the spam bots; not to mention doing so without adding to your ticket/support/enforcement overhead.

STORY 1.
As a filter, I must factor several variables to determine targeting confidence, so I can reliably detect and suspend spam bots.
*

RATIONALE:
It is possible to use a series of simple rules to quickly target and suspend spam botting accounts. By combining a flood detection, pathing, naming, text string, and chat channel ruleset, it is possible to positively identify target characters and accounts for suspension.

RMT Bots are very unlikely to file support tickets asking for reinstatement. On the other hand, players are highly likely to do so. Either way, all the tickets of people reporting spam bots (duplicate, triplicate, etc) can be mitigated or potentially eliminated.

This would also lay a foundation for further refinement to profile and process community enforcement as a strategic support alternative given the nature of the MMO environment. A preliminary filter set is outlined in following stories.

STORY 2.
As a filter, I must be able to determine a pattern of flooding chat, so I can reliably detect spam bots.

RATIONALE:
Much like the "bad word" filter, it is possible to filter chat characters, lines, or send requests, as well as combinations thereof by crafting a simple baseline mask of common RMT spam bot announcements. There are even items from which derivative parsing might be quickly conceptualized (i.e., Ex2SM) should there not be ready tooling within the organization for such. The basic pattern of chat for RMT spam bots is established and slow to change due to certain necessities of "doing business" in text; this can be used to your advantage.

STORY 3.
As a filter, I must be able to determine a pattern of movement and pathing, so I can reliably detect spam bots.

RATIONALE:
Because of the unique structure of the PSO2 game environment, RMT spam bots have a pattern of pathing and behaviors that can be identified by similarity, thus used to further refine a detection listener for RMT spam bots. These could be detailed here, but it would be counterproductive. Kindly ping me via PM if you would like me to polish and deliver them separately.

STORY 4.
As a filter, I must be able to determine a pattern of naming convention, so I can reliably detect spam bots.

RATIONALE:
There are licenses available for tools to filter and protect against "non-name" data patterns in character creation. This could be used both to delay account creation of RMT spam bots as well as quickly identify those that make it through the system (refinements to the filter, ongoing). These could be detailed here, but it would be counterproductive. Kindly ping me via PM if you would like me to polish and deliver them separately.

STORY 5.
As a filter, I must be able to determine a pattern of text and channel delivery, so I can reliably detect spam bots.

RATIONALE:
As with the preceding, there are patterns of chat behaviors that may be codified for use in filtering as part of an RMT spam bot detection/suspension listener. These could be detailed here, but it would be counterproductive. Kindly ping me via PM if you would like me to polish and deliver them separately.

SUMMARY
Taken together, the above items would form a robust listener appliance/tool that could be used at scale to quickly detect and respond to the behaviors common to RMT spam bots. This post outlines a basic set of requirements for developing a listener applicance/tool to automate both RMT spam bot detection AND potentially automate support ticket to listener handoff for 'instant response' processing in near real-time.

This post, presented by a new player who happens to be a retired product and program manager in the I.T. software development industry, is provided both as an attempt to be helpful to you and to improve the overall player experience of a game I have quickly grown to delight in and enjoy.

I MIGHT be open to more than casual suggestions should there be interest. I see a lot of opportunity in NGS and any work in this arena now could/should be easily translatable across the board/brand/products.

Regardless, thank you for listening and considering.

EDIT HISTORY:

  1. Typos.
  2. Format for ease of reading.