|
This is probably the first post on this topic. It's will be impossible
to cover all of this in one blog, but I want to begin an important
discussion.
There is a lot of discussion on 2+2 about the presence and effectiveness of poker bots. Here is some general background:
1. Bots have been around for years and they are winning at a good rate (2 bb/100+) at the midstakes (1/2 to 10/20).
2. Their technology and speed will only get better.
3. If done correctly, it will be impossible for pokersites to detect them.
4. They could effectively kill the games.
All of this adds up to being pretty scary. Some top calliber players
have stated that the problem is very serious and with some depressing
implications. I'll try to summarize some of the ideas that have been
floated around.
1. Bots have been around for years and they are good.
This shouldn't come as a big surprise, but some of the most primitive
bots would all play the same style, grind out a small profit (and make
tons off rakeback). They were easy to detect and highly exploitable once
you could identify them. They would mostly play at the micros. I just
recently learned that this new wave of bots has been playing the
midstakes and killing them.
Here is a graph of the compilation of ~70 accounts playing 6-max from 2/4 to 10/20:
Source:
http://forumserver.twoplustwo.com/56/medium-stakes-pl-nl/important-news-about-botting-party-poker-msnl-cashgames-1122426/
And here is another alias of 104 accounts winning 4.12 bb/100:
Source: http://forumserver.twoplustwo.com/showpost.php?p=29129344&postcount=292
2. Technology is getting better.
From what I understand, overall speed of computer processors and memory
is doubling every two years. Thus, computers will only continue getting
faster. Current bots are already able to beat the games and they will
only get better as programmers keep working on them.
3. Future bots will be undetectable.
Bots in the past were easy to find - they would make decisions
quickly and play during specific times of the day. Their stats were also
identical and they wouldn't sit at the same table with one another.
All of this can be accounted for. For one, the botting software can be run on a separate computer which uses screen scraping
to make decisions. You can then have a human sitting and executing the
decisions that the computer dictates. There is no way a site can guard
against that.
I don't think the current bots are set up this way, but it wouldn't take much to establish such a setup.
4. Bots could kill the games.
The poker economy is based around recreational players. If recreational
players read about poker sites being infested with bots that are out
there making money, they might be deterred from playing. As bots become
better and better, they will continue to become more and more
unexploitable which will reduce everyone's winrate. At some point, the
regs will not find it profitable to play anymore and then there will
truly be "no money in poker - everyone is solid."
Right now, we are not there yet, but it's an ominous possibility. Bots
are a reality, so the discussion should now center around what can be
done to minimize their effectiveness.
|
