PDA

View Full Version : Statistical Proof/Debunking of Peak Hours Advantage- RE:'s Needed


Messy_Jesse
08-08-2005, 02:49 AM
Ok. Enough with the speculation over whether peak hours are better for ROI, hourly rate, etc. I want proof. Here's how you are gonna help me. If enough people will post their ROI, ITM, and number of games played for each period of the day, we can compile the list and get a statistical validation or invalidation of whether or not it is actually better to play during peak hours.

So, I'll start, and I'd appreciate it if others would also post, for the limit they play at THE MOST, their ROI, ITM, and # games played for the hours used with Aleo's spreadhseet. I'll keep a running total, and I thinkt that the results will be very interesting and helpful to everyone. So, here goes...

(straight from the stats page of Aleo's spreadsheet...)

Time #T Top 3 ROI
Night 164 39.63% 28.05%
Morning 9 66.67% 81.82%
Forenoon 20 20.00% -54.55%
Afternoon 100 26.00% -20.00%
Dinner 159 38.36% 18.93%
Evening 165 33.94% 20.11%

EDIT: If, for whatever reason youd rather not post your stats but would like to contribute, if you PM them to me I'll erase the PM after I anonymously record the stats. You dont have to be a winning player or have a 29% ROI at the 200s to help out with this.

jwesty5
08-08-2005, 02:51 AM
EST?
CDT?
GMT?

Afternoon somewhere is midnight somewhere else.

Nottom
08-08-2005, 02:54 AM
EST!

Night 270 33.7% 12.1%
Morning 1 100.0% 358.7%
Forenoon 6 33.3% -23.5%
Afternoon 100 35.0% 2.8%
Dinner 93 28.0% -5.3%
Evening 116 32.8% 8.4%

Obviously playing in the morning is the bestest.

HighestCard
08-08-2005, 03:02 AM
CST

Night 26 42.3% 44.8%
Morning...too early for me...
Forenoon 16 50.0% 34.2%
Afternoon 32 50.0% 10.7%
Dinner 46 43.5% 38.7%
Evening 112 41.1% 16.7%

I only wish I had more...

Messy_Jesse
08-08-2005, 12:26 PM
bump.

Slim Pickens
08-08-2005, 02:49 PM
about 1k 11's, 22's, and 33's, all Pacific time.

night: 26, 30.8%, -26.8%
morning: 2, 50%, -9.1%
forenoon: 84, 27.4%, -2.9%
afternoon: 289, 39.1%, 15.2%
dinner: 416, 38.9%, 17.0%
evening: 277, 40.4%, 22.5%

I shouldn't quit my day job.

Freudian
08-08-2005, 03:42 PM
Forenoon is death

Night 563 42.9% 24.1%
Morning 275 36.4% 13.0%
Forenoon 116 32.2% -11.9%
Afternoon 248 39.7% 16.3%
Dinner 680 40.2% 22.2%
Evening 823 38.6% 18.4%

08-08-2005, 03:53 PM
mostly 22s, Pacific time

Night 40 42.5% 5.6%
Morning 0 #DIV/0! #DIV/0!
Forenoon 85 40.0% 19.6%
Afternoon 261 39.1% 11.7%
Dinner 340 39.4% 17.3%
Evening 529 42.0% 23.7%


Anyone want to take odds that Freudian ends up being 80%+ of our group Morning sample?

08-08-2005, 04:05 PM
Time #T Top 3 ROI
Morning 50 40% 19%
Forenoon 1 0% -100%
Afternoon 8 38% 7%
Dinner 50 34% -5%
Evening 115 38% 18%
Night 406 35% 10%

Slim Pickens
08-08-2005, 04:42 PM
Make sure you post the time zone you're using. You can change the "Lookup" tab to bin the results any way you want. It might even be necessary just based on the three-hour difference between Eastern and Pacific, and then there's the internationals too. I'll experiment with some modifications to the sheet to standardize this...

samr
08-08-2005, 04:49 PM
EST

Time #T Top 3 ROI
Forenoon 295 41.4% 13.8%
Afternoon 333 42.0% 19.9%
Dinner 128 43.8% 19.5%
Evening 6 33.3% -30.6%

Interesting results.

shejk
08-08-2005, 05:22 PM
There is something very weird with the results my spreadsheet presents:

Time #T Top 3 ROI

Night 145 37.2% 22.9%
Morning 38 34.2% #DIVISION/0!
Forenoon 22 40.9% #DIVISION/0!
Afternoon 122 27.0% -63.0%
Dinner 168 43.5% 547.3%
Evening 143 41.3% 288.2%

I'm guessing my crazy rois has something to do with me having played at four different limits, anyone know how I can get my true numbers?

btw, this is GMT+1 (CET).

Newt_Buggs
08-08-2005, 05:52 PM
If I had my stats on this computer then I would contribute my 1,500 $55s, oh well.

Slim Pickens
08-08-2005, 06:40 PM
There were problems with earlier versions of the spreadsheet, specifically with the time and day splits. I fixed those in the later versions, and they seem to work fine now. If you look on the "readMe" tab, it should say "last modified by [username]." If [username] isn't me, dfscot or astarck (I don't know if he changed this) then you need to get a newer version of the spreadsheet from http://mowrmowr.com/poker. I use 2.4 (in it's folder). Astarck added a daily summary tab that takes a little more computer power in version 2.5. Anyway, try getting a modern version, cut-and-paste your stats (check the User Guide for some important details) and see if that fixes your funky splits.

SlimP

shejk
08-08-2005, 06:53 PM
Oops, I don't even have a readme-tab.

Thanks a bunch, I'll just dl a new one and try again.

Slim Pickens
08-08-2005, 07:51 PM
OK. I figured out how to modify the spreadsheet lookup tab to change your time zone. It's a little bit complicated so back up the Christ out of you spreadsheet before you do this.

1) Back up your file. That's not just a figure of speech.

2) You will have to insert extra rows into the lookup table at the top of the "Lookup" tab. That's now the cells in rows 1 through 6, columns A, B, and C. The rows must be inserted using Insert>Rows command rather than just tacking them onto to bottom because there are a lot of cells in other tabs that will not automatically update to include your new row if it's just added to the bottom. Click on cell A2 and slecet Insert>Rows. A blank row should appear in the middle of the table.

3) Put new time values in the table. Column A is the "begin" time, column B is the "end" time, and column C is the letter corresponding to that time period. For historical reasons, the letters mean this:

N: night = 12AM-4AM
M: morning = 4AM-8AM
F: forenoon = 8AM-12PM
A: afternoon = 12PM-4PM
D: dinner = 4PM-8PM
E: evening = 8PM-12AM

Here's the tricky part. For this experiment, you need to adjust these times to correspond with some standard time zone, say Eastern. As an example, my spreadsheet entries are all in Pacific time. I need to correct all the time periods by making the borders three hours earlier, so my adjusted periods need to mean this:

N: night = 9PM-1AM
M: morning = 1AM-5AM
F: forenoon = 5AM-9AM
A: afternoon = 9AM-1PM
D: dinner = 1PM-5PM
E: evening = 5PM-9PM

The problem is that the chart won't wrap around for that 9PM-1AM block, so it has to be entered in two separate lines. That's why you needed the extra line. In addition, the lines must be in chronological order starting at 12AM, so my final lookup table has to look like this:

Column A / B / C (rows 1 through 7)
12:00 AM / 1:00 AM / N
1:00 AM / 5:00 AM / M
5:00 AM / 9:00 AM / F
9:00 AM / 1:00 PM / A
1:00 PM / 5:00 PM / D
5:00 PM / 9:00 PM / E
9:00 PM / 12:00 AM / N

I hope this works for anyone who's interested. PM or post if it doesn't make sense.

Slim Pickens
08-08-2005, 07:54 PM
All times Eastern now.

night: 198, 37.9%, 20.5%
morning: 7, 42.9%, 1.0%
forenoon: 4, 25.0%, -54.5%
afternoon: 132, 34.1%, 11.6%
dinner: 340, 39.4%, 16.0%
evening: 413, 39.0%, 15.1%

Myst
08-08-2005, 08:08 PM
Haha, Jess, if you wanted my extensive data, you just had to ask :P

Total Tournaments: 3051

Time #T Top 3 ROI

Night 482 36.7% 7.4%
Morning 317 35.6% 6.7%
Forenoon 131 38.2% 0.8%
Aftrnoon 594 35.2% 3.2%
Dinner 700 36.9% 10.1%
Evening 815 37.7% 13.0%

My highest ROI is at Dinner and Evening, but then again, my sample size is alot higher for both.

Messy_Jesse
08-09-2005, 10:14 PM
So far, there are only about 6000 SNGS. I really think that a few more thousand would help out a lot. I know that there are more than 7 people on this forum, and Id appreciate it if youd help contribute your stats. Especially the players who play for a living and could contribute a few thousand. PM your stats if you want, or whatever, but please help out.

-Jess

EDIT: This post has 550 views, and 12 replies. COME ON!!!!

08-10-2005, 12:18 AM
These are for the 11s -

72 37.50% 21.21%
0 #DIV/0! #DIV/0!
15 46.67% 63.64%
22 59.09% 90.08%
0 #DIV/0! #DIV/0!
46 30.43% -11.07%

sorry about the missing ones

MuckedTheNuts
08-10-2005, 01:33 AM
All EST

Night 100 34.0% -2.3%
Morning 0 #DIV/0! #DIV/0!
Forenoon 76 38.2% 28.3%
Afternoon 123 30.9% -1.9%
Dinner 108 40.7% 28.1%
Evening 36 50.0% 44.5%

axeshigh
08-10-2005, 03:35 AM
If you have any way to enter them automatically in the spreadsheet (I can't) I guess I can send you my roughly 800 sngs worth of data.

Jay36489
08-10-2005, 03:43 AM
What does this prove? Some people play better against the weak tight daytime players, and some play better against the crazy drunken late night players. Its more a matter of what style you play. There is no true better or worse.

Newt_Buggs
08-10-2005, 03:59 AM
if everyone has a lower ROI during 6am est then its a pretty good indication that there are less fish dumping money into party making the games as easy as during peak hours.