Day of the week patterns

[bicyclekick]

Wow, there is a pretty big discrepancy between my winrates based on day of the week. I know people have said this before a milion times, but now I have data over a pretty decent sample (about 100k hands) and can see it myself

Sunday and monday are the same - pretend at a base of 1, tuesday and wednesday at 2 times the sunday/monday rate , thurs and friday 2.5x, and saturday at 3x.

man, I should just do all my homework/hanging out with gf/partying and stuff on sunday/monday and play poker in the last part of the week.

[Sponger15SB]

well i've got 100,000 hands and mine is all over the map.

I'm not on my PT computer, but lets just say none of my numbers make any sense... its something like my biggest winning day is thursday and my worst is saturday or something.

[Zetack]

[ QUOTE ]
Wow, there is a pretty big discrepancy between my winrates based on day of the week. I know people have said this before a milion times, but now I have data over a pretty decent sample (about 100k hands) and can see it myself

Sunday and monday are the same - pretend at a base of 1, tuesday and wednesday at 2 times the sunday/monday rate , thurs and friday 2.5x, and saturday at 3x.

man, I should just do all my homework/hanging out with gf/partying and stuff on sunday/monday and play poker in the last part of the week.

[/ QUOTE ]


Where's sample size man when you need him? If you played the same number of hands on every day you'd have 14k on each day...I'd question whether 14k hands is anywhere enough to say that's your winnrate for that day. If your results aren't evenly divided then the problem becomes worse for many of your days.

Oh, and have you filtered for variables such as time of day, game levels etc etc?

--Zetack

[bicyclekick]

I thought about all of that and agree sample is still small. Still I think there is some merit to it.

[dogmeat]

Howdy Kick,

I've got a few interesting quirks on my daily patterns also. Thursday morning from midnight to 2am is a minus 19.16 for me - yikes, but Sunday at the same time is +36.86 . Also, Sunday from 6pm to 8pm is a minus 11 and 8pm to 10pm is a minus 9.58, which give me an overall loser on sundays of 1.50BB per hour. Wednesday shoes a minus .52, but Thursday, even without that disaster two hours is still +4.45

I can discount my Thursday negative number because it is based on some very small numbers (I rarely play past midnight now).

So, do I suck on Sunday and Wednesday? Maybe. Today is Sunday and I'm up $96. Maybe I'll take the rest of the day off. /images/graemlins/laugh.gif

Next week between 6pm and 10pm on Sunday I think I'll play a couple tournaments - problem solved.

Sample size is Party and skins only - just about 10K hands per day, and yes, this is really small and kind of silly, but still interesting. When you adjust for the actual two-hour periods that I do play, there is just 1100 hands per daily group.

When I played a lot of 7-card stud on Stars, I loved Sunday evening because I could run over the game. Maybe I'm trying to do the same on the Party hold'em games?

Dogmeat /images/graemlins/spade.gif

[Tosh]

[ QUOTE ]
I thought about all of that and agree sample is still small. Still I think there is some merit to it.

[/ QUOTE ]

Does a win rate claim have any merit at 14k hands? No. Therefore I don't see how you can claim any merit here either.

[SinCityGuy]

Based on posts I’ve seen over the past couple of years, there is some anecdotal evidence suggesting that win rates are higher on Friday and Saturday, and a bit lower on Sunday and Monday. My guess would be that there are more recreational players in the pool on Friday and Saturday, while Sunday and Monday might have a higher concentration of skilled players. There are several possible reasons for this. Here’s one. The recreational players who get paid on Friday are more likely to play after getting their paycheck, while they might be tapped out by Sunday and Monday.

I don’t have any empirical evidence to back all of this up, but that’s my opinion. I always play Tuesday thru Saturday, and I take Sunday and Monday off.

[Cosimo]

[ QUOTE ]
Where's sample size man when you need him?

[/ QUOTE ]

Sample Size Man can go to hell!

Over 7023 hands, my BB/100h ranges for 2-hour periods ranges from -32 to +27 BB/100h. Woot, changed the time zone and now my highest result is +44BB/100h.

So, uh, maybe Sample Size Man does have a point. I'd guess I'd need around 300k hands for this chart to start settling out, especially since I have only played in 26 of the 84 different chunks.

(PokerTracker, Ring Game statistics, Session Notes tab, hit the More Detail button)

[RED_RAIN]

[ QUOTE ]
The recreational players who get paid on Friday are more likely to play after getting their paycheck, while they might be tapped out by Sunday and Monday.

[/ QUOTE ]

You can also consider some people get paid twice a week so around the 1st and 15th.

I was just recently talking to someone who was playing $2/$4 with $50 total and I told him it was suicidal. He then went on to state he doesn't get paid to the 15th. A player who is playing with 12.5 BBs is probably going to lose most of it early on as you state the Friday/Saturday theory. So perhaps there could be a theory with some merit, but would need some good stats to back it up.