Thursday 26/04/2012, 15:16

Thought I should wait, but then decided otherwise.
The DT lottery algorithm programmed into the UR mainframe is becoming a bit predictable. If you draw a frequency distribution curve with the already awarded prices, we can find most of them huddled around two magic numbers.
I have found two. 33 and 94. Meaning if you are placed around these two places in DT, you stand a better chances of winning a prize in DT, the former being better (less standard deviation in the case of latter).
Any thoughts on this?

Thursday 26/04/2012, 16:52

I think you deserve Fields Medal smiley

Friday 27/04/2012, 05:16

It's still not a huge sample size, but something to keep an eye on.

Friday 27/04/2012, 11:00

Exactly my point. Not a huge sample size, but even with this limited sample size, the emerging trend is clear.
The random number generating algorithm, if not re-programmed on a periodic basis, will generate random numbers clustered around predictable numbers.

Friday 27/04/2012, 16:29

You might be looking too much into it, only time will tell. smiley

Monday 30/04/2012, 16:04

Haha, mate, well said. The sample size is too low to make good deductions, but still, look at the results after my posting this. if it ain't 33+/- 8, it's 94+/- 8 smiley

Tuesday 01/05/2012, 00:24

Yeah, but if you rank the highest among the 5 winners, you still get the best card... today's DJ Korr Cr winner was in 8th place, while 120th place got Noodile Cr...

and, from experience, aiming for a particular spot in DT can be difficult, especially now with shorter timers and more players XD

so what would be the odds of winning a Cr anyway? around 1/360, or probably slightly better, considering the people who rank in more than one DT per day... so, if you play in tourney every day for a year, on average you should win at least one Cr! smiley

in any case, i'm not sure it's really worth it to sacrifice the extra prize money every time, when there's really no guarantee of winning a Cr anyway. and it's still random enough that top 25ers or even top 10s will get the grand prize, a lot of the time. that being said, it's interesting to see how this will develop...

Tuesday 01/05/2012, 12:39

@FofP: Dude, look closer. There is a Scarlett Cr going to the 33+6th place and Dwain Cr to the 94+5th place. Check the result well before posting just for posting's sake.

Tuesday 01/05/2012, 15:06

Merlin have you checked this for ELO prices aswell?smiley
Or is it only in DT?

Tuesday 01/05/2012, 17:41

Haha, mate, I found this by happenstance, but statistical analysis proves I have almost struck paydirt. If someone goes through the ELO draws with a fine comb like SPSS (I might do that if I can find the time), he might find something interesting.
DT draws are easier to track because it is new and I have all the draw results till now at hand; that's why I analysed this instead of ELO.


Reply to this subject