Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

What about your hourly rates for these times? Also, what do you use to track your data? Anyone else have these kind of stats they can post so we can compare and see how much may be due to player specific factors?

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

[ QUOTE ]
What about your hourly rates for these times? Also, what do you use to track your data? Anyone else have these kind of stats they can post so we can compare and see how much may be due to player specific factors?

[/ QUOTE ]

I use SnG Tracker. Honestly, I'm embarassed to post my hourly rates.

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

Enough of this already.

Let's test the significance of hypothesis that results differ by time of day.

For simplicity I'll take that the std on RoI of a tourney is 165%. Std actually depends on RoI, but I will neglect that.

Ilya, you needed to give an extra significant digit for all results, but I'll work with what you gave.

NOON-4PM: 647 tournaments, (4.0+-6.5)% ROI, (36.0+-1.9)% ITM
4PM-8PM: 1097 tournaments, (11.0+-5.0)% ROI, (39.0+-1.5)% ITM
8PM-MIDNIGHT: 1107 tournaments, (17.0+-5.0)% ROI, (40.0+-1.5)% ITM
MIDNIGHT-4AM: 753 tournaments, (13.0+-6.0)% ROI, (39.0+-1.8)% ITM

Average: 12.0% RoI, 38.8% ITM

Chi^2: 2.61/3 DoF (degrees of freedom) RoI; 2.89/3 DoF ITM.

The numbers you observed are perfectly in line with random deviation from the average.

What makes the significance of your claim worse is observer's bias. If you saw results that did not show the trend you expected, you would be more likely to think that your samples are small, and less likely to report the results.

Nobody needs stats to know that RoI is better in the evening, but your stats alone are way short of making any observational claims.

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

Hi Izverg,

How much larger would my sample size have to have been for these results to have statistical significance?

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

[ QUOTE ]
Hi Izverg,

How much larger would my sample size have to have been for these results to have statistical significance?

[/ QUOTE ]

That would depend on how much deviation per time of day there really is. If the deviation is as large as you found it from your limited samples, you'd only need ~5 times more data to start making significant claims. If the deviation is smaller, you'd need more.

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

[ QUOTE ]
[ QUOTE ]
Hi Izverg,

How much larger would my sample size have to have been for these results to have statistical significance?

[/ QUOTE ]

That would depend on how much deviation per time of day there really is. If the deviation is as large as you found it from your limited samples, you'd only need ~5 times more data to start making significant claims. If the deviation is smaller, you'd need more.

[/ QUOTE ]

I see, thanks.
Now I guess I should go read a book so I can understand the math you used to reach your conclusions.

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

Well I think surely that's more important than ROI isn't it really? And I ask because having achieved my first 1000 tournaments, I have found that a decent ROI doesn't appear to translate to a decent hourly rate, a thing I don't really understand, so i'd like to see how it matches up for a successful player like yourself... c'mon

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

[ QUOTE ]
Well I think surely that's more important than ROI isn't it really? And I ask because having achieved my first 1000 tournaments, I have found that a decent ROI doesn't appear to translate to a decent hourly rate, a thing I don't really understand, so i'd like to see how it matches up for a successful player like yourself... c'mon

[/ QUOTE ]

Well here's the thing, I like to 4-table because I feel that I make better decisions and I enjoy playing much more when I have time to observe players, make copious notes, make read-based plays, etc....it makes me feel like a human being rather than a money-making robot...also I like to chat...also I am much less susceptible to tilt when I play fewer tables...Also I find that I can 4-table continuously almost indefinitely, whereas if I try to 8-table continuously, I get flustered and burnt out after 2-3 hours and have to take a break. Also I only have 1 monitor, and the little tables strain my already-bad eyes. Soooo....my hourly rate is like $22/hr at the $22s.

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

[ QUOTE ]
$22/hr at the $22s.

[/ QUOTE ]

Better than working in a factory.

Re: My stats over the last 3677 tournaments by 4-hour time blocks
 

[ QUOTE ]
[ QUOTE ]
$22/hr at the $22s.

[/ QUOTE ]

Better than working in a factory.

[/ QUOTE ]

Not if the factory is unionized and comes with service based hourly rates, a big fat pension benefit, post-ret health benefits, and a health plan and insurance benefits. The value of those things can be massive but not transparent. Some of those guys make a pretty sweet haul.