lots of good info on the bot profiles available. look for bot selling website with forum. look around in there.
This post is to explain the playing style of the default profile that comes with the Holdem bot, answer common questions about it, and provide advice for using it profitably.
Text Version Access
At the top of the Working Profiles board (you must be a registered forum member to access that board) are a group of “sticky” posts which stay in that top section. One of these is the thread for the Doodle profile. Post 1 in this thread, that is the very first post, contains the text version of the Doodle profile. You can copy it from there and paste it into notepad and save it for loading into the bot.
Post 1 of the Doodle thread always contains the latest version and older versions are not kept because the profile always improves with the newer versions.
The profile that comes with the bot download is the same profile, but in .ppl format - which means the user cannot access it and it does not quote codelines that are executed in the logs. Therefore if you want to post hand histories from the bot’s session log for suggested improvements, please use the text profile version so I can see what codelines were executed in the actions taken. Also, the text profile may be a newer version than the .ppl version, although these periods where they are different versions are usually short.
MTT Version
Also in that batch of sticky posts is a free “MTT Kit” that I created. This is a short text profile that is designed to manage your stack well in tournaments as it grows and shrinks. It can be combined with any cash game profile, such as the Doodle, by blending the individual betting round codes as instructed in the first post on that thread. It can also be used as an entire text profile at the same time that you have a .ppl profile loaded, which will have the same net effect as blending it on a text profile (except that only the codelines from the kit will be quoted in the logs).
I always blend the MTT Kit with the Doodle and post the blended text profile at the end of the MTT Kit thread when I post the latest Doodle profile. You can access this text profile by going to the end of the MTT Kit thread and start looking backwards until you find my most recent post of the “Mtt Kit + Doodle”.
History
When we first added post-flop PPL variables I made a post in the forum titled “flop doodle” just to demonstrate some of the creative possible uses with the new variables, such as check-raising and slow-playing big hands from early position. A few people started using it and reporting good results and posting suggestions for improvements. I also I started making “doodle” posts for the Turn and River. It just kind of grew from there and eventually I combined them all with a good preflop code I was working on and the rest is history (you can read the entire Doodle thread if you want to read that history).
At last count there were over 200 pages on the Doodle thread. Most of the improvements are the result of our members posting hand histories from the bot’s session log showing where play can be improved. So it really is a true community project, although I decide on what suggestions get implemented. These have gotten to be few and far between these days, so the new versions are a lot slower in coming now.
Playing Style
The Doodle profile plays a solid, classic tight-aggressive style (also known as TAG) for No Limit Holdem. This means that it plays tight and folds most hands preflop. When it does play, it is usually the aggressor in the hand, raising to open and betting when given the chance. This gives it two ways to win. It can make a hand or the opponents can all fold to the bets. Most serious professional online players play a TAG style. It does, of course, back down when an opponent gets stubborn about sticking around when the bot does not have a strong hand.
Some of our members who have run it for more than 100,000 hands have posted stats from Poker Tracker or Holdem Manager and they are classic TAG stats. This is simply not debatable because the stats don’t lie.
Preflop Raising
The number one complaint that I get from newbies using this profile is they see it making min-raises preflop in some situations and don’t like it. The reason they don’t like it is because they watch too much TV-Poker and nobody ever raises the minimum on television. Also, somewhere sometime somebody wrote that “the correct” preflop raise size is 3.5 times the big blind or some such nonsense. So they have just enough knowledge to be dangerous and figure they have to change this profile to make pot-sized preflop raises, or a 300% raise, or whatever ridiculous notion they got stuck in their mind as what they “always” should do (very few things in poker are actually an “always”).
That notion about never min-raising preflop is based on a limited knowledge about poker in general, and more importantly a lack in understanding of the formula that this profile is designed to use. I can guarantee that I have been playing longer, and more successfully, than most of the people who worry about the Doodle’s preflop min-raising.
The fact is that this profile does also make many pot-sized preflop raises. When it min-raises it does so because I think it is best move from a risk-reward standpoint in that particular situation. This especially takes into account the flop-continuation betting code that is being used.
The preflop betting pattern is not predictable because sometimes it also min-raises with AA and KK. Your opponents only need to see that once before they start giving your min-raises some respect.
Flop-Continuation Betting
The other common complaint I get is that the standard flop continuation bet is pot-sized and some people don’t like that. They want to make it something like 70% of the pot. This actually makes some sense, because if they are making large raises preflop then they have a real problem with the bet-size on a flop-continuation bet. The pot has gotten too big and is now out of their control. So it is consistent with a philosophy of making larger preflop raises to make smaller flop-continuation bets.
What I am here to tell you is that I tried it both ways and looked at the numbers nine ways from Sunday. And I have come to the conclusion that you are better off in the long run making smaller preflop raises (with marginal raising hands), and larger sized flop-continuation bets. This is for two reasons. The first is that people still fold to min-raises preflop without decent hands, and that is what they hold most of the time. So if you raise from the cutoff when everybody folds to you with JT it is likely to be just you and the big blind on the flop, and many times the big blind will fold to the preflop min-raise anyway.
On the flop the pot will be small if he does call. So a pot-sized bet will be a strong-looking bet and he has to fold if he didn’t flop a decent hand, which is most of the time. But the pot is small enough to make that strong bet not too expensive.
Here is the second reason: If you raised the pot preflop with that JT instead, a pot-sized bet becomes too expense and testing has shown that you will not do as well in the long run. If instead you make a 70% of pot-sized bet, well that just looks weaker and will inspire some opponents to play back at you with nothing, or "float" you with a call for their weak pair.
It makes sense either way but the philosophy should be consistent. If your preflop steal-raises are pot-sized then yes a 70% of pot bet, or even smaller like 60%, is necessary. However if your preflop steal-raises are just min-raises then you can afford to make the strong pot-sized continuation bet. I have tried it both ways in my poker career and I have arrived at the conclusion that min-sized steal raises followed by pot-sized continuation bets just work better overall (for one thing you can widen your stealing range). And since the Doodle is my profile, that is how it plays.
Best Games
The Doodle plays more aggressive as the table gets short-handed. Therefore it should be fine at 6-max tables if that is your thing. You will clear bonuses faster and rack up rakeback a lot quicker in 6-max.
However, that is the logical equivalent of paying more in taxes so your refund will be bigger. You only get a % of it back, so paying more and getting more back is always net worse than paying less in the first place. Traditionally a TAG player does better at full 9-max tables because the cost of sitting there is less. The blinds come around less frequently and there are more players to split of the cost of the rake.
That does not mean you should sit out when the table gets short! I do not recommend that. The text version of the profile is set to never sit out. There is quite a bit of short-handed and heads-up code in the profile and it usually holds it’s own when the table gets short. I have even seen the Doodle break aggressive players when playing heads up. Yes your variance will be higher and the cost to play increases. Don’t let that scare you. Overall it is better to never sit out because the tables will usually only fill back up if the existing players are playing.
Stakes are another matter. The lower stakes games are of course easier to beat than higher stakes games. Table selection tactics will of course find more profitable games then blindly joining any table. Playing at night when people are drinking will of course yield better results than the daytime when the room is full of regulars. A TAG player generally wants a loose, passive table so keep that in mind. That means lots of people calling before the flop and not a lot of preflop and flop raises from your opponents. Play at the right times, in the best games, and this profile should do fine as high as 25NL.
Results
There are unsolicited graphs posted in the personal journals section confirming that this profile is beating the lower stakes games handily. Here are links to two of the graph images:
169,000+ hands at NL2 (includes rakeback on the blue line):
http://a.imageshack.us/img248/7704/165kgraph.png
89,000+ hands at NL5 (does not include bonus cleared or rakeback earned):
http://img251.imageshack.us/img251/6580/11thgraph.gif
Please note that a graph with less than 50,000 hands is of questionable value and it really takes 100,000 hands to get a reliable depiction of performance. This is because there are occasional 12,000+ hand downswings on any winning profile. So graphs with less than 10,000 hands are pretty much a joke.
You can use the Doodle at NL10 to clear any deposit bonus and expect to get at least 2/3 of it as profit, assuming you have the time to clear it. True, you can get all of it as profit (plus some profit from playing to boot) by playing NL5 or lower but it will take a lot longer. And time is money.
At Bodog this profile will beat NL25, especially at night USA time when the sports bettors are in the poker room. It will also perform well in any environment where a typical good TAG player has an edge.
When the jackpot gets over $400K at UB this profile does well in the NL50 and even some of the NL100 jackpot games.
Modifying the Profile
You are of course welcome to take the text version of the Doodle and tweak it to your liking. Heck you can do whatever you want with it, mess with the option settings, add new codeblocks, make it raise max with 72 of clubs, or print it out and wrap fish with it. I think it makes an excellent base to start from.
Perhaps the easiest way to do this is to watch the profile play and take good notes. Give the PPL Guide a read. Then start adding 1-liner code commands to the top of the betting rounds which will change specific situations that you want played differently in the future. Before you know it you will have your own profile. Who knows, you might just create a killer NL50 profile that can make your house payment. Good luck with it.