My HeadsUp for the title experiment.

[Daliman]

Well, I'm going to log the next 100 times i get HU for the title in a $200 SNG, logging my entering chipstack and my result W or L, then compare total chipstack equity to my win %. If don't like my results from that, then the NEXT 100, I'm lgging the same info, yet pushing EVERY SINGLE HAND,(not necessarily CALLING though...), then comparing the 2.

I think I have a good HU game, but the fact is that I place about 1% more total than I win (11.1% win, 12.3% place), and after about 2500 tourneys, I think i'm starting to approach statistical relevance here.

Anyways, I'll post my results nightly. I'm thinking the first 100 will take a month or so,(gonna be away on vacation for 2 weeks late Dec.)

[Unarmed]

THANK YOU.
I've always wanted to try this and am very interested in seeing the results.

[Irieguy]

I'll post an early prediction:

1. Your push strategy will be at least as good as your standard strategy.

2. Push strategy will not be more than 15% better than your standard strategy. (meaning 15% of your standard result + your standard result... not 11.1% + 15%.)

The only problem is that a power analysis of the data would show that you need more samples than you have the patience to collect in order for the results to be significant.

Irieguy

[Daliman]

[ QUOTE ]
I'll post an early prediction:

1. Your push strategy will be at least as good as your standard strategy.

2. Push strategy will not be more than 15% better than your standard strategy. (meaning 15% of your standard result + your standard result... not 11.1% + 15%.)

The only problem is that a power analysis of the data would show that you need more samples than you have the patience to collect in order for the results to be significant.

Irieguy

[/ QUOTE ]

#1 Possible. Especially when time is factored in, this may become my new permanent strategy.
#2 Also possible.

Yeah, i know this won't tell a hell of a lot, but it's something I'd like some basis of thought on. Although I also think my previous results are/were more due to starting at a lower chipstack on average. I'm guessing if I start ~T100 lower Heads up, that could directly account for the difference mathematically, (EASTY!)
MAN do I wish pokertracker could be tweaked for this kinda stuff.

[eastbay]

[ QUOTE ]
Well, I'm going to log the next 100 times i get HU for the title in a $200 SNG, logging my entering chipstack and my result W or L, then compare total chipstack equity to my win %. If don't like my results from that, then the NEXT 100, I'm lgging the same info, yet pushing EVERY SINGLE HAND,(not necessarily CALLING though...), then comparing the 2.

I think I have a good HU game, but the fact is that I place about 1% more total than I win (11.1% win, 12.3% place), and after about 2500 tourneys, I think i'm starting to approach statistical relevance here.

Anyways, I'll post my results nightly. I'm thinking the first 100 will take a month or so,(gonna be away on vacation for 2 weeks late Dec.)

[/ QUOTE ]

I would guess that your win/place split may have a lot more to do with your 3-way play than it does your HU play. Just a thought.

eastbay

[Daliman]

Night 1 results.

7 HU matches.
29921 total beginning chips
4-3 record.

[rachelwxm]

Daliman,

I have a program that does this, and HU is a simplified version of my analysis. And my emperical studies of several hundreds HHs shows that you need 300-400 HH to have some statistically meanningful results for ROI around 20-40%. Since ROI for 200 is generally believed to be smaller, I would guess you need 800+ games or 200+ HU unless you are extremely good at HU. /images/graemlins/smile.gif

Just some random thoughts. The results would be interesting.

[Daliman]

I used your program before, I'm pretty sure, and it doesn't seem like it did what I am doing,(unless a different program...). Didn't it measure your "luck" of hot/cold runs of hands?

But yes, I know it's not going to be too statistically significant.

[rachelwxm]

[ QUOTE ]
I used your program before, I'm pretty sure, and it doesn't seem like it did what I am doing,(unless a different program...). Didn't it measure your "luck" of hot/cold runs of hands?


[/ QUOTE ]

My program is not so easy to use and it's still sitting on my computer unless someone hacked it. Are you talking about a different thing? /images/graemlins/cool.gif

[Daliman]

I guess so, i DL'd a prog recently, thought mebbe it was one you posted.

Please disregard my reply....

[UMTerp]

I've been doing very well heads-up. I'd like to think it's because of my superior play, but I think I'm just getting pretty lucky with unbridled aggression. Has anyone been able to maintain a win rate similar to this long-term? I'd like to know if I'm due for a little downswing:

373 Stars Turbo SNGs (About 45% $16's, 45% $27's, and 10% $60's):

Place Times Percent
1 65 17.43%
2 48 12.87%
3 49 13.14%
4 36 9.65%
5 55 14.75%
6 37 9.92%
7 36 9.65%
8 31 8.31%
9 16 4.29%
TOTAL 373 100.00%

Also, is the discrepancy between 4th and 5th place finishes normal, or am I doing something weird??

[Ian J]

Hi Dali,

In the experiment, let's say you have 63o in the BB. You each have T5k and blinds are 250/500. Your opponent comes in for 1300. Are you pushing here? Or is it only when you just open push or push when he limps the SB?

[UMTerp]

You also need to set up rigid calling standards for when he pushes for the experiment to have much merit.

[Daliman]

[ QUOTE ]
Hi Dali,

In the experiment, let's say you have 63o in the BB. You each have T5k and blinds are 250/500. Your opponent comes in for 1300. Are you pushing here? Or is it only when you just open push or push when he limps the SB?

[/ QUOTE ]

Only as first in, unless i have a warranted push/call hand.

[Daliman]

Not really. More of an experiment for my own purposes than specific scientific merit.

Anyways, day 2 results.

5 HU matches.
17184 total chips
2-3 record

Cumulative stats;
12 HU matches, 6-6 record, 47105 starting chips (60k is equal start average.)

Also, as an aside, I'm 6/6/7/17 the last 2 nights, for a decent ~4700 profit. Should be an easy rate to keep up, only $133.19 profit each, and what, a 60% or so ROI?

/images/graemlins/cool.gif

[1C5]

This is interesting, thanks for posting the results.

[rachelwxm]

Daliman, you are doing great! One question, what is the typical blinds size when you start HU?

[Daliman]

GUESS TIME!

40% 300-600
25% 250-500
20% 200-400
8% 150-300
6% 100-200
1% 50-100

2 times ever at 15-30.

[Zelcious]

How many do you play at once? 2 ? 4 ?

[AA suited]

[ QUOTE ]
GUESS TIME!

40% 300-600
25% 250-500
20% 200-400
8% 150-300
6% 100-200
1% 50-100

2 times ever at 15-30.

[/ QUOTE ]

what does this mean?

[The Yugoslavian]

I think it means that Daliman is not exactly sure what the blinds usually are when he's heads up. Thus, this would just be an approximate guess off the top of his head.

At least, that makes sense to me, he could certainly be talking about how much limit poker he thinks Sklansky plays, /images/graemlins/wink.gif.

[Daliman]

Yes, sorry, I should have quoted. It's in reference to my approximation of blind levels when i get HU.

Also, I 4 table.

[Daliman]

Horrendous night tonight. I don't remember the last time I went this card dead around the bubble and in the $$$. -2200 tonight

Anyways, the sordid details
4 HU matches
8495 total chips (.......)
0-4.

Cumulative totals.
16 HU matches
55600 total chips(80k is par)
6-10 record.

The decent luck hath ended. Bah

[eastbay]

That makes both of us. Had a "run" of 6 bubbles in a row tonight, every last one with final chips in as a fav.

eastbay

[Daliman]

Now, I don't know if the STEPS tourneys are directly involved, but with clearer waters than usual, I was able to book a $4300+ win tonight.

/images/graemlins/grin.gif

Anyways, on to the numbers, looking a LOT better now. I happened to go 7-1 HU tonight, added with an 0-1 from the 4 SNG's I played Saturday night, putting my HU record back in the black, which I'm beginning to appreciate more coonsidering my avg chip count at start of HU.

9 HU Matches
50916 total chip(9160 in one of them! PRESS PRESS PRESS!)
7-2 record

Overall totals
25 matches
106516 total chips(125000 is par)
13-12 record.

Average stack starting HU is thus far 4261 a quarter of the way through.

[Daliman]

Decent night tonight, solid, if unspectacular.

7 HU Matches
30837 total starting chips
4-3 record

Overall totals
32 Matches
137353 total chips(160k is par)
17-15 record

[HC5831]

I've thought about the benefits of using the all in everyhand HU strategy once down to 2 for a while now. Although I have not tried it. I go in probably 75% of the time as it is now.

I have a relevant question here though. What is the minimum blinds where this strategy is close to optimal. If the blinds are 150/300, then going all in every hand is a bad idea because your opponent has a chance to catch a hand. I've set the bar at 250/500, although 200/400 might be the level. What are your thoughts or anyone else's?

HC

[J-Lo]

i play SNGs @ pokerstars, and i like to set an allin or fold standard when the smaller stack is about 8-10 xBB

[Daliman]

[ QUOTE ]
I've thought about the benefits of using the all in everyhand HU strategy once down to 2 for a while now. Although I have not tried it. I go in probably 75% of the time as it is now.

I have a relevant question here though. What is the minimum blinds where this strategy is close to optimal. If the blinds are 150/300, then going all in every hand is a bad idea because your opponent has a chance to catch a hand. I've set the bar at 250/500, although 200/400 might be the level. What are your thoughts or anyone else's?

HC

[/ QUOTE ]

I tend to agree with you. I barely push at all 150/300, and about half as much as 250-500 and above at 200-400.

[rachelwxm]

Great numbers, do you mind quoting how many total SNGs you play for those HUs? Just curious.

[Daliman]

[ QUOTE ]
Great numbers, do you mind quoting how many total SNGs you play for those HUs? Just curious.

[/ QUOTE ]

Sure, i'll incorporate them from now on, but total to now is 120.

[rachelwxm]

A rough calculation shows that this does +.55% ROI from your HU. What is your running ROI for those 220 games?

[Daliman]

[ QUOTE ]
A rough calculation shows that this does +.55% ROI from your HU. What is your running ROI for those 220 games?

[/ QUOTE ]

31.6% for 120 games. HIGHLY sustainable......

[Daliman]

Crappy night tonight. All 3 times I got HU for the title, I ended up getting my $$$ in 2-1 fave or better only to lose /images/graemlins/frown.gif .

3 HU matches
11270 chips
0-3 record
16 SNG's Played.

Cumulative stats.

35 HU matches
17-18 record
147623 Starting chips(175K is Par.)
136 SNG's played.

[xtravistx]

any updates?

[Daliman]

Crappy night

1 HU match
1410 chips
16 SNG's

Overall totals
36 matches
17-19 record
149033 total chips
152 total SNG's.

[bads33d]

Hey dailman, I did a couple 200s with you today, I was also in your step 4 yesterday (the guy who took everyone out, lol). Im DonButtons__

I had a ok night

16 200 SNGS
2-2 heads up
2 firsts
2 seconds
4 thirds

[Daliman]

After the $2700 loss I suffered Sat night, I didn't update my stats since I was so depressed, and I ended up starting toight being down 2000 after my first 11, and only having $115 left in my account, with my last 3 ok stack games going of the night, all in mid stages.

First, the ugly beat;
1: mor05gan ( $1505 )
Seat 2: signiasam ( $1745 )
Seat 3: JucyLones ( $1875 )
Seat 6: DarkStargasm ( $3445 )
Seat 8: ZeeOne ( $1430 )
Trny:7880066 Level:7
Blinds(200/400)
** Dealing down cards **
Dealt to DarkStargasm [ Kd As ]
signiasam folds.
JucyLones is all-In.
>You have options at Table 13293 Table!.
DarkStargasm is all-In.
ZeeOne folds.
mor05gan folds.
** Dealing Flop ** [ 6d, Kc, 6h ]
** Dealing Turn ** [ 6s ]
** Dealing River ** [ 6c ]
JucyLones shows [ Ac, Jc ] four of a kind, sixes.
DarkStargasm shows [ Kd, As ] four of a kind, sixes.
DarkStargasm wins 1570 chips from side pot #1 with four of a kind, sixes.
JucyLones wins 2175 chips from the main pot with four of a kind, sixes.
DarkStargasm wins 2175 chips from the main pot with four of a kind, sixes.

As it turns out though, I ended up winning this one still, along with BOTH the others, so went from -2k to +400 in about 30 minutes there. Before I won the 3 in a row, i had my deepest downturn ever hit of 6800, or about 32 buyins.

I decided to quit anyways though just because I was so spent from the earlier brutality that It seemed like the right thing to do.

Anyways, the 2 night numbers

6 HU matches
31834 total chips
4-2 record

Overall stats
42 HU matches
21-21 record
180867 Starting chips(210k is par)
188 SNG's

Also, a quick note on something I have noticed; I have yet to lose when I start with the chip lead HU. Good to know.

[Daliman]

Great night after losing 2450 on UB NL cash games, i make that much+ a little more in just 1 SNG's

5 HU Matches
18903 total chips
3-2 record
11 SNG's

Overall results
47 HU matches
198960 total chips(235000 is par)
24-23 record.
199 SNG's

[lacky]

If I could combine your reaching the money with my heads up play we'd be rich. Most of the profit I make comes from heads-up results.

My numbers at $55's are

total 1155
total in money 436
total 1st 165
total 2nd 134
total 3rd 137

It's the only making the money 38% that bugs me. I'm studying my middle game trying to improve that number.

[Daliman]

my ITM is only about 39% this month.

[DonButtons]

[ QUOTE ]
Great night after losing 2450 on UB NL cash games, i make that much+ a little more in just 1 SNG's

5 HU Matches
18903 total chips
3-2 record
11 SNG's

Overall results
47 HU matches
198960 total chips(235000 is par)
24-23 record.
199 SNG's

[/ QUOTE ]

One of those wins that night came from sucking out on me /images/graemlins/frown.gif. I think your aq beat me ak, I forgot.

I made a questionable call with 55 on you that you commented on, hope you see my other post on that situation.

[bigredlemon]

I just tried it for 5 games. Win rate far below average. Not doing that again.

[HC5831]

I haven't seen a post in a while.. Are you still doing the expiriment? What are the final results?

HC

[Daliman]

I was on vacation, and had/have been playing the 1K steps, but am having a bad luck run in there, so will likely be resuming shortly.

[Daliman]

OK, getting back into the 200's. Had my first loss so far this experiment after going in with chip lead when first hand after getting heads up I get AKs in in BB for 500, SB limps, I have 6154 total, and make it 1500 to go, he thinks a bit, then pushes, having zero fold equity. I insta-call, he has 44 that holds, of course. gott love it. Very next hand, I have ~2300 left now, and i get 99; I push from the SB, he thinks a bit, and calls 1800+ more with J2d..... you know the rest. Gotta love poker....

Anywho, here's the #'s

7 HU matches
2-5 record
25729 total chips
38 games.

Overall stats.
54 HU matches
26-28 record
224689 total starting chips(270k is par)
238 Games

[Daliman]

Ok, i didn't update my results every single time I played recently, but I do have them all here now, so here's the sordid details.

P.S. My streak of winning the first 16 or so times I started with the chip lead is a thing of the past. I'm 8-6 my last 14 going in with a lead now, one of them where I lost going in with 8444 /images/graemlins/frown.gif (not much I could do; had him allin J8d vs 46d first hand, still lost, and he limp/called allin from even stacks with KTd and hit a T to beat my PF allin raise with A8o a bit later. What can ya do?

23 matches
110602 total chips
9-13 record
not sure how many games this was, but not alot,(been on a bit of a run.)

Overall stats.
77 HU matches
35-42 record
335291 total chips(385000 is par)
306 total games.


I will try to finish this off by the end of this week. Shouldn't take more than 80-90 games more(famous last words?)

[schwah]

i think those numbers are definately possible long term. in my experience, stars turbos have quite a few more weak/tights you can take advantage of when the blind crunch hits than party does. this lets you get a significantly higher number of wins.

i haven't played enough for the statistics to mean much, and my results are a good bit better than they'll look long term im sure. but anyways heres my results for the first 120, mostly $60s with a few $114s and a very few $27s:

1st 31 25.8%
2nd 14 11.7%
3rd 7 5.8%
4th 17 14.2%
5th 14 11.7%
6th 12 10.0%
7th 9 7.5%
8th 8 6.7%
9th 8 6.7%

43.3% ITM
46.1% ROI

again i know that the sample size isn't big enough to mean more than i PROBABLY am winning more than my fair share.

[Daliman]

Haven't felt much like playing as of late, so I have played only sparingly. Bit of a bad run again, but evened it out some tonight. Anyways, here's my most recent stats.

7 HU matches
4-3 record
31449 total chips

Overall stats.
84 HU Matches
39-45 record
366740 Total starting chips( 420000 is par)
339 total games.

I WILL finish this this week.

[Mr_J]

You're seriously doing this? Think of the $$$ you're losing in the process /images/graemlins/wink.gif

[jack spade23]

exactly what j said

[Daliman]

Yeah, It hurts alot.

[Mr_J]

Out of interest, what sort of 1st vs 2nd % do you usually get?

[Daliman]

at the beginning of this, it was 11.1% to 12.2%

[HC5831]

Daliman,

So is your experiment finished now? What are your conclusions? I'm very interested in this.

HC

[Daliman]

[ QUOTE ]
Daliman,

So is your experiment finished now? What are your conclusions? I'm very interested in this.

HC

[/ QUOTE ]

Well, as I was approaching the end of this was when I hit my 43 buyin drop in the 200's, so I lost interest in it, along with most everything else. I'll look through my records to see if I can find where i finished off, but if i can't, I'll fill it in with my current info.

[raptor517]

dalidaddy, since i have been doing so terribly bad anyway, i might take it upon myself to complete the heads up for the title experiment. i should be able to get pretty decent results in a matter of a couple weeks. its probably not giving up much ev at all, just getting it in every single time.. ill look into it some more. holla

[1C5]

Yeah that would be very cool to see.

So you would simply push every hand but not call with every hand heads up right?

[1C5]

Also, as a side benefit, hourly rate would go up a bit as the games would be over quicker right?

[Apathy]

[ QUOTE ]
Also, as a side benefit, hourly rate would go up a bit as the games would be over quicker right?

[/ QUOTE ]

No, playing less then optimally HU to finish the games quicker will not increase hourly rate.

It is interesting to see if you can do better by pushing every hand though, good luck raptor.

[1C5]

[ QUOTE ]
[ QUOTE ]
Also, as a side benefit, hourly rate would go up a bit as the games would be over quicker right?

[/ QUOTE ]

No, playing less then optimally HU to finish the games quicker will not increase hourly rate.

It is interesting to see if you can do better by pushing every hand though, good luck raptor.

[/ QUOTE ]

But do we know that is actually less than optimal?

[Apathy]

no we don't (that wasn't my point)...hence the experiment /images/graemlins/tongue.gif.

[1C5]

So I should have said hourly rate will go up if results are the same or very close to the same.

[raptor517]

well, i really have no idea what will happen during this test, and im probably going to be too lazy to make notes on it every single time, but ill play a lot and push every hand and see how it goes. with any luck, ill have a 20% roi again in the 109s and be on my way to being a bajillionaire. bwahahahha

[johnnybeef]

[ QUOTE ]
well, i really have no idea what will happen during this test, and im probably going to be too lazy to make notes on it every single time, but ill play a lot and push every hand and see how it goes. with any luck, ill have a 20% roi again in the 109s and be on my way to being a bajillionaire. bwahahahha

[/ QUOTE ]

holla?

[citanul]

no, not holla,

bwaahahahah

[The Student]

uh, oh - is raptor sick or something?

[iMsoLucky0]

Who here would enjoy his posts much more if they ended in a "Bwahahaha" ? I know I for sure would. Much easier (and funnier) to read than "Holla."

[pokerswami]

I just read this thread for the first time, and something seems amiss.

Daliman's experiment was to play and log 100 HU matches with his regular strategies. He would then evaluate the results, and if he was dissatisfied he would switch to the all-in strategy for the next 100 HU matches.

You seem to be saying that you're going ahead with the all-in strategy before first logging the results of 100 HU matches with your regular strategies.

What am I missing?

[Daliman]

[ QUOTE ]
I just read this thread for the first time, and something seems amiss.

Daliman's experiment was to play and log 100 HU matches with his regular strategies. He would then evaluate the results, and if he was dissatisfied he would switch to the all-in strategy for the next 100 HU matches.

You seem to be saying that you're going ahead with the all-in strategy before first logging the results of 100 HU matches with your regular strategies.

What am I missing?

[/ QUOTE ]

Well, a lot of people misunderstood my initial intention as I'd just push every hand, but that was a fallback if I wasn;t happy with my HU results after this. From my last results post I think i was winning about 48% of the time with about 41% of the chips, which is pretty solid. I thought I might be getting in equalish chips and just sucking HU, but that didnlt appear to be the case. I will pick up from where I left off tomorrow, promise, and will then give a mea culpa on the whole thing.

Pushing every hand cannot hurt rap's results. /images/graemlins/cool.gif

[HC5831]

utt ooo. Are you saying the results you have posted are from your "optimal" play and not pushing every hand? I'm only interested in the results of pusing every hand HU. I couldn't care less what your HU % is at the 200 level. /images/graemlins/cool.gif

HC

[citanul]

read the freaking post if you want to know what the results are about.

citanul

[Daliman]

[ QUOTE ]
utt ooo. Are you saying the results you have posted are from your "optimal" play and not pushing every hand? I'm only interested in the results of pusing every hand HU. I couldn't care less what your HU % is at the 200 level. /images/graemlins/cool.gif

HC

[/ QUOTE ]

Bah on you, sir.

The push system has already proven to be approximately 47% effective at a minimum, but blinds make a big difference. I think the sweet spot is about 200-400.

[raptor517]

well well well, maybe raptor will have to just push every hand. i dont think too many things right now could hurt his results. oh wait, thats me, damn u dalipunk

[HC5831]

[ QUOTE ]
Bah on you, sir.

The push system has already proven to be approximately 47% effective at a minimum, but blinds make a big difference. I think the sweet spot is about 200-400.

[/ QUOTE ]

Daliman,

I would expect the level of the blinds to be key. 150-300 is definately too early to push every hand. 200-400 seems borderline to me. 250-500+ definately. I'm basing these thoughts on 1k starting chips. This would be a good strategy against a player who you know is better than you. In addition, an allin every hand strategy is good for players whose normal play HU is less optimal than the pushing strategy. I'm just trying to figure out how successful it is.

I thought your posts were for the push system, but I saw some posts to the contrary and got confused. I'm still not 100% positive, but I'll just assume your results are the post system. That being the case, I am very interested in the experiment. I was just saying that I don't care how you normally do HU without the experiment. You have my respect as a poker player on the forum, no offense intended. Sometimes us former Marines need things spelled out for us.

HC

[mosch]

Well, you inspired me to find and look at my numbers and to discover that I no longer have first place skew.

my last 819 single table sngs at stars: (about 25% $16, 50% 27, 25% 60)
1 - 126
2 - 126
3 - 122

I'm glad I looked because I used to have more firsts than seconds, and I also had more 4ths than 3rds... neither of those is true any longer, and I'm wondering if I've accidentally started playing to Not Lose on the bubble, and hadn't noticed.