New Limit Autorates Ready for Some Testing!

[Tiburon]

[Tiburon]

[digital scar]

Hey Tib nice work on this stuff.

As TW said, some of what you are talking about in your thread seems a bit over my head so I just nod and smile when I read it.

I will try out you new rules and post whatever data you are looking for with them, the only problem is I don't know what data you would like. Let me know and I'll use your rules for awhile and then get you the info you want.

[Tiburon]

For the data I'm requesting,

in PT, go into your preference tab (in ring games). In the "Use hands from the following limits" section, check off all your full-ring limit hands from 0.50/1 up to as high as 10/20. Then, check the box saying "Use Player Filter." Underneath that, it will give you a chance to select what icon you want to use stats for.

Go icon by icon, recording (manually, sorry) the number of players listed, their total number of hands played, the total BB/100, and the percentage of players that are winning players.

Do this for each of the 14 icons and post your results here in a format like those used in the AutoRate thread stickied above or in the thread showing results from excession's NL auto rate rules being used in limit.

Thanks all in advance!!!

[Tiburon]

The idea behind expanding the passive definition was to eliminate the weak/tight. We want to take advantage of the passive players, not be trapped by a weak/tight player.

Here are my stats based on those rules--losing strategies in red.

Category............Players.................Hands.................BB/100.............% Winners
TP (Rock)...........235......................15,986..................-2.85...................41.77
sLP (Mouse)......130.......................9,109....................-9.19..................23.97
LP (Fishy)..........137.......................4,946....................-13.69................28.86

TN......................403.......................43,532..................+1.11..................53.58
sLN....................361.......................42,558..................+2.28..................53.28
LN......................201.......................15,414..................+2.12..................53.60

TA......................40.........................11,834..................+1.98..................58.00
sLA...................149........................20,485..................+1.97..................54.24
LA......................44..........................3,132...................+0.72..................53.33

UA.....................246........................41,154..................+3.49..................61.20
xL......................261........................16,160.................-11.70..................33.11
SM.....................176........................12,189.................-21.68...................8.63
BWT....................86..........................3,241..................-18.90.................18.48
CS.....................487.........................46,239.................-2.59...................42.81

There are still a ton of neutrals, more than I really wanted, but it does better split off the passives.

The remaining neutrals are better players, but still better than their aggressive counterparts. Ideally, we need to see that the BB/100 and winner rates increase as aggression increases. Striker might end up being right about those aggression numbers. I'm gonna try something different.

[Tiburon]

Latest permutation of the autorates:

Passives:
AF-Total < 1.25
AF-Turn < 1.00
AF- River < 1.00

I'm considering removing the turn and river from the passive aggression rules, but I'm not sure yet...

Aggressives:
AF-Total > 1.50
AF- Turn > 1.25
AF-River > 1.25

I'm not certain I like them, but they do provide the best spread of the rules. Aggressives win more than neutrals, neutrals win more than passives, and passives win more than the crappy groups (SM, CS, xL, etc.)

The rules can be downloaded here:



and the Excel spreadsheet I used to compare the changes (number of players & BB/100) in the rules can be downloaded from here:




By no means does this end the discussion. I'd like to continue the discussion of these rules, and I still want all the data I can get. The link says final, but with me, nothing's final and nothing is beyond tinkering. I'd like to think that these rules will help limit players better read their opponents. If anyone sees a way that they can be improved, please speak up. I'd prefer that you shared your information with the group, of course, and posted any data you can produce.

I'd like to thank excession for all his hard work on the original NL autorate rules, APerfect10 for helping him get those done, and everybody on the limit forum for their contributions to this board, this forum, and these rules.

The rules are public domain. All I ask for is that if you're going to post them someplace else, you give appropriate credit to BTP, excession, AP10, and myself/

[Acehawk74]

Tib.... being a math guy, im very very intrigued by what youre doing here. FANTASTIC work! Ill help in any possible way that I can!

[Tiburon]

I did some more data-mining and got this data from the new auto-rates:

Category.............Players............BB/100
TP..........................163.................-3.43
sLP........................90...................-11.31
LP..........................112.................-14.07TN..........................449..................0.33
sLN........................440..................1.20
LN..........................197..................-0.45
TA..........................57....................1.82
sLA........................146..................1.83
LA..........................66....................2.76
UA.........................256...................2.81
xL..........................263.................-10.92
SM.........................178.................-21.46
BWT......................87...................-18.00
CS.........................495..................-2.36

I broke down the UA group:

T-UA....................99.....................2.32
sL-UA..................75.....................5.45
L-UA....................82....................-1.84

Just some FYI....watch your UA's VP$IP. A L-UA usually = a maniac that will lose in the long run.

[digital scar]

Hey Tib. looks like your still kicking ass with your limit research. I ended up having to go out of town for the weekend and won't have any access to my database. I'm sure alot of others probably have more hands at highter limits then I do and it looks like you might be squared away with this stuff anyway.

Ok everybody have a good 4th.

[striker2550]

[edmunds5]

Hi, I would like to contibute my data. How do I get it from PT? Thanks.

[edmunds5]

ignore last post....i scrolled up.

[Tiburon]

Striker--

Right click the link for the rules, and save them as a text file. I've also re-uploaded them just in case:

[striker2550]

Thanks Tibby. Yeah I know how but the link just doesnt work for me. When I click it, a window opens then does nothing.

I went ahead and put the rules in manually anyway.

Ive only used them for 1 session so I dont have much of an opinion yet. But Im not to sure about the Tight Neutral/Unknown catagory. I think there are a few decent players there that just go dead on the river(below 1.25 river aggression) but otherwise are aggressive.

[Ricardooon]

You will be pleased to know I have given your rules to the missus for testing! Now you are scared...

Thanks for the good work, keep it up.

Rich