- Joined
- Oct 12, 2015
- Messages
- 3,818
- Thread Author
- #1
Hey all, thought I'd give my thoughts on the recent situation regarding bans, and the steps we're taking to minimize them.
TL;DR
We've just published a small set of changes that we believe should significantly reduce the number of bans across certain activities, primarily for bots that were using Direct Input. We're continuing to work on further improvements to combat recent Jagex detection changes (details below).
Before we get into it, I think it's important to remind people that there has always been an arms race between Jagex and bot developers: they develop new detection technologies, we develop new technology to circumvent them. In recent months, Jagex has taken significant steps toward improving their macroing detection systems, primarily through behavioural analysis, impacting all bot platforms, not just RuneMate.
Things you can expect to see in the (very) near future
Some closing thoughts
(my personal opinion, for those interested)
To me, it seems that the changes Jagex has made recently take a more holistic approach to detection, building a model of how a character plays throughout its lifetime and flagging deviations from that model, rather than relying on the narrower, more targeted approaches they've used in the past.
I've seen marketing materials from other platforms over the last several weeks claiming to make sweeping changes in response to this shift in strategy. Some of these changes genuinely seem like a step in the right direction, but the majority, in my opinion, still miss the point. Many are fundamentally iterations on ideas that platforms have been experimenting with for well over a decade, and recent detection trends suggest those approaches alone are no longer sufficient.
The areas I've outlined above are the ones I believe will have the biggest long-term impact, and we'll be focusing our efforts there rather than relying on superficial "antiban" systems or one-size-fits-all solutions.
It's my strong belief that the biggest impact will come from better educating users on healthy playstyle patterns, alongside developers continuing to improve the behavioural realism of their bots. Our goal as a platform is to provide users and developers with the tools and education they need to make that happen.
TL;DR
We've just published a small set of changes that we believe should significantly reduce the number of bans across certain activities, primarily for bots that were using Direct Input. We're continuing to work on further improvements to combat recent Jagex detection changes (details below).
Before we get into it, I think it's important to remind people that there has always been an arms race between Jagex and bot developers: they develop new detection technologies, we develop new technology to circumvent them. In recent months, Jagex has taken significant steps toward improving their macroing detection systems, primarily through behavioural analysis, impacting all bot platforms, not just RuneMate.
Things you can expect to see in the (very) near future
- Mouse Improvements
We will be working in collaboration with our Premium Bot Authors to build a new mouse system that more closely resembles how an actual human would interact with the game, both in terms of speed and accuracy. This new mouse system should properly track moving targets, resulting in fewer misses and more human-like behaviour. This is less about the path the mouse takes to get from A to B, and more about mimicking how a person tracks a target across their screen. - PlayerSense Improvements
The PlayerSense API was a great first look at building gameplay patterns tied directly to player accounts, but it has become dated. There is still a lot of room for improvement in this system, with our goal being to have bots resemble the way you play across all activities. - Scheduling
Given the recent detection changes, it's more important than ever for bots to resemble humans as closely as possible. That means doing things like taking a break from an activity to do a farm run, or switching activities when you get bored. We intend to introduce a scriptable scheduling system that will enable users to switch activities more naturally. - Tighter control over the use of Direct Input
Direct Input was initially introduced to enable a certain class of bots that required incredibly fast and accurate inputs, particularly for high-intensity PVM activities. We released this feature flagged as experimental, warning users that it could potentially carry increased risk. While Direct Input was originally introduced for a narrow set of use cases, its adoption expanded well beyond those scenarios. Based on what we've observed, we believe this increased exposure contributed meaningfully to recent detection rates.
Since its release, we've made many improvements to the client that mean Direct Input is no longer necessary in the vast majority of cases. Moving forward, we intend to more tightly restrict its use, disallowing it for ordinary activities that simply do not require it.
Some closing thoughts
(my personal opinion, for those interested)
To me, it seems that the changes Jagex has made recently take a more holistic approach to detection, building a model of how a character plays throughout its lifetime and flagging deviations from that model, rather than relying on the narrower, more targeted approaches they've used in the past.
I've seen marketing materials from other platforms over the last several weeks claiming to make sweeping changes in response to this shift in strategy. Some of these changes genuinely seem like a step in the right direction, but the majority, in my opinion, still miss the point. Many are fundamentally iterations on ideas that platforms have been experimenting with for well over a decade, and recent detection trends suggest those approaches alone are no longer sufficient.
The areas I've outlined above are the ones I believe will have the biggest long-term impact, and we'll be focusing our efforts there rather than relying on superficial "antiban" systems or one-size-fits-all solutions.
It's my strong belief that the biggest impact will come from better educating users on healthy playstyle patterns, alongside developers continuing to improve the behavioural realism of their bots. Our goal as a platform is to provide users and developers with the tools and education they need to make that happen.
