[QUOTE=Dubslow;293740]Well, you've succeeded in confusing me as to when I was previously confused.
:surrender::geek::truck::lol:[/QUOTE]
Quick off the mark there.
I'm sure Retina, Ernst or other could invent a Python sketch here:smile:

David
x

[QUOTE=aketilander;293719][COLOR=black][FONT=Verdana]
[/FONT][/COLOR][COLOR=black][FONT=Verdana]2012-03-21 1.107 Expected time to next prime is (50/0.024) 2083 days[/FONT][/COLOR]

or if we look at the whole period from 2008-10-15 -- 2012-03-21 the expected time to next prime is (1243/0.640) 1942 days.[/QUOTE]

I'll do my best to retain my sense of humour here (as UncWilly
advises me/us frequently).

The reason I estimate the time rather than the expo is
1) the actual expo could vary by ~15M
2) as you have diagnosed, the "expected time" is easy to estimate.

A glance at your/my table suggests that the Expected Time to
the next prime discovery is ~6 years.

The good news is that it has "bottomed out".

The bad is that maintaining this equates to a 20% in LL effort/year.

Now what the **** are GPUs and fast CPUs doing?
More in another thread.

David

[QUOTE=davieddy;294368]2) as you have diagnosed, the "expected time" is easy to estimate.

A glance at your/my table suggests that the Expected Time to
the next prime discovery is ~6 years.

The good news is that it has "bottomed out".[/QUOTE]

Well I think this gives a reasonable estimation but I would guess that the math guys could question it. I would say as long as the decrease is more or less linear the estimation would be valid, but lets say if the decrease would slow down we would underestimate the time and if the decrease would speed up we would overestimate the time. The variations during the last 4 years have not been very large so we could argue that its a rather good approximation to guess that the decrease is linear.

Yes I think at the moment it has "bottomed out" due to the speed jump of AVX.

I think your estimation (a very intelligent one) is the best we can have at the moment, but maybe one of the math guys can give an even better estimation?

[QUOTE=aketilander;294369]
I think your estimation (a very intelligent one) is the best we can have at the moment, but maybe one of the math guys can give an even better estimation?[/QUOTE]

I think you'll find Bob will simply tell you that my heuristics
have not yet been proven.
Little does he know:smile:

[url=http://www.youtube.com/watch?v=y9VLtelUUMk]Who one knows?[/url]

Senility is a bummer.
 

[QUOTE=davieddy;294373]
No [URL="http://www.youtube.com/watch?v=y9VLtelUUMk"]one knows?[/URL][/QUOTE]
I meant:
[url=http://www.youtube.com/watch?v=TBsWYYBnNK4&feature=related]Who knows where or when?[/url]

David

Don't get me started on the Spanish Inquisition, or I'll refer you to Ernst.

Flattery will get you anywhere
 

[QUOTE=aketilander;294369]
I think your estimation (a very intelligent one) is the best we can have at the moment, but maybe one of the math guys can give an even better estimation?[/QUOTE]

Try Paul.

[QUOTE=aketilander;293719][COLOR=black][FONT=Verdana]A summary of Davieddy's and others post above:[/FONT][/COLOR]

[FONT=Verdana][COLOR=black]2001-04 we expected 1.64 (found 2)[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2004-07 we expected 1.28 (found 4)[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2007-10 we expected 0.69 (found 3)[/COLOR][/FONT]

[FONT=Verdana][COLOR=black]2008-10-25 1.747 [/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2008-12-03 1.719 Expected time to next prime was (39/0.028) 1393 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2009-10-25 1.528 Expected time to next prime was (326/0.191) 1707 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-05-11 1.420 Expected time to next prime was (198/0.108) 1833 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-10-25 1.342 Expected time to next prime was (167/0.078) 2141 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-11-22 1.328 Expected time to next prime was (28/0.014) 2000 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-12-25 1.314 Expected time to next prime was (33/0.014) 2357 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-01-11 1.306 Expected time to next prime was (17/0.008) 2125 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-02-05 1.295 Expected time to next prime was (25/0.011) 2273 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-03-07 1.281 Expected time to next prime was (30/0.014) 2143 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-05-26 1.244 Expected time to next prime was (80/0.037) 2162 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-06-22 1.231 Expected time to next prime was (27/0.013) 2077 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-07-15 1.221 Expected time to next prime was (23/0.010) 2300 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-09-10 1.195 Expected time to next prime was (57/0.026) 2192 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-10-16 1.177 Expected time to next prime was (36/0.018) 2000 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-12-20 1.149 Expected time to next prime was (65/0.028) 2321 days[/COLOR][/FONT]
[COLOR=black][FONT=Verdana]2012-01-31 1.131 Expected time to next prime was (42/0.018) 2333 days[/FONT][/COLOR]
[COLOR=black][FONT=Verdana]2012-03-21 1.107 Expected time to next prime was (50/0.024) 2083 days[/FONT][/COLOR]
[COLOR=black][FONT=Verdana]2012-04-25 1.090 Expected time to next prime is (35/0.017) 2059 days[/FONT][/COLOR]

[/QUOTE]

(updated)

[QUOTE=Uncwilly;285197]petrw1 guessed 55,555,543 would drop on 5/5/2012
[YOUTUBE]angi1vwUkQc[/YOUTUBE][/QUOTE]Well we passed that.


davar55 on 5/24/2011 said that 77,000,000 (or there abouts) would be the next MP and the find date would be in 6/2012.
ET_ is next with 20/12/2012

We have passed (average gap + 2.75 std dev) today.

:busy:

[QUOTE=Uncwilly;299025]
We have passed (average gap + 2.75 std dev) today.
[/QUOTE]
Have you visited Gamblers Anonymoius?

The probabililty of one or more prime before y years from now is
1 - e[SUP]-y/6[/SUP]

David

[QUOTE=davieddy;299027]
The probabililty of one or more prime before y years from now is
1 - e[SUP]-y/6[/SUP]

[/QUOTE]How was that derived? :-)

[QUOTE=cheesehead;299089]How was that derived? :-)[/QUOTE]

I don't know but his prediction puts it at about 391 years until it can be guaranteed. doh I doubt it ever really guarantees it but that's the first time it passes rounding of the 28 digits I have it showing.

[QUOTE=science_man_88;299096]I don't know but his prediction puts it at about 391 years until it can be guaranteed. doh I doubt it ever really guarantees it but that's the first time it passes rounding of the 28 digits I have it showing.[/QUOTE]

Hint: What t would you need for guaranteed success? That is, can you set the probability function equal to 1 and solve for t?

[QUOTE=cheesehead;299089]How was that derived? :-)[/QUOTE]
Poisson with an expected time of 6 years.

David

[QUOTE=davieddy;299131]with an expected time of 6 years.
[/QUOTE]
I think he meant how was that derived :wink:

[QUOTE=Dubslow;299132]I think he meant how was that derived :wink:[/QUOTE]
I think Richard will understand my sketchy answer, but since I like you, here it is (Hope Bob isn't reading this:)

Probability of finding a prime in dy years is dy/6.

Let P(y) be the probability of no prime before y years.

P(y + dy) = P(y)*{1 - dy/6)

dP/dy = -P/6

P = e[SUP]-y/6[/SUP]

David

Hint: think "exponential decay".

... where [i]6[/i] is a perfect number. :-)

[QUOTE=cheesehead;299136]... where [I]6[/I] is a perfect number. :-)[/QUOTE]Where 6 years is the expected time as monitored by me,
with the able assistance of Ake...

[QUOTE=davieddy;299135]

Probability of finding a prime in dy years is dy/6.[/QUOTE]

I meant how did you derive this. I am certainly well versed in Ordinary Differential Equations -- I just took the Partial Differential Equations final on Monday.

[QUOTE=Dubslow;299138]I meant how did you derive this. I am certainly well versed in Ordinary Differential Equations -- I just took the Partial Differential Equations final on Monday.[/QUOTE]
Then you probably know what a "wave" is:smile:

BTW
 

William may have something helpful to say about all this.

Trust me. I'm a gynocologist.

David

[QUOTE=Dubslow;299138]I meant how did you derive this.[/QUOTE]

Which part of "derive" do you understand?

[QUOTE=science_man_88;299096]I don't know but his prediction puts it at about 391 years until it can be guaranteed. doh I doubt it ever really guarantees it but that's the first time it passes rounding of the 28 digits I have it showing.[/QUOTE]
Only two things are certain in this world:
Death and Taxes

D

BTW that was a big U turn of an edit you did!

[QUOTE=davieddy;299223]Only two things are certain in this world:
Death and Taxes[/QUOTE]
The version I learned was [i]Only two things are certain at this site: DEC and VAXes.[/i]

[QUOTE=xilman;299243]The version I learned was [I]Only two things are certain at this site: DEC and VAXes.[/I][/QUOTE]
Ant and Dec?
Shake and Vac?

[QUOTE=aketilander;293719][COLOR=black][FONT=Verdana]A summary of Davieddy's and others post above:[/FONT][/COLOR]

[FONT=Verdana][COLOR=black]2001-04 we expected 1.64 (found 2)[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2004-07 we expected 1.28 (found 4)[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2007-10 we expected 0.69 (found 3)[/COLOR][/FONT]

[FONT=Verdana][COLOR=black]2008-10-25 1.747 [/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2008-12-03 1.719 Expected time to next prime was (39/0.028) 1393 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2009-10-25 1.528 Expected time to next prime was (326/0.191) 1707 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-05-11 1.420 Expected time to next prime was (198/0.108) 1833 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-10-25 1.342 Expected time to next prime was (167/0.078) 2141 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-11-22 1.328 Expected time to next prime was (28/0.014) 2000 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2010-12-25 1.314 Expected time to next prime was (33/0.014) 2357 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-01-11 1.306 Expected time to next prime was (17/0.008) 2125 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-02-05 1.295 Expected time to next prime was (25/0.011) 2273 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-03-07 1.281 Expected time to next prime was (30/0.014) 2143 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-05-26 1.244 Expected time to next prime was (80/0.037) 2162 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-06-22 1.231 Expected time to next prime was (27/0.013) 2077 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-07-15 1.221 Expected time to next prime was (23/0.010) 2300 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-09-10 1.195 Expected time to next prime was (57/0.026) 2192 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-10-16 1.177 Expected time to next prime was (36/0.018) 2000 days[/COLOR][/FONT]
[FONT=Verdana][COLOR=black]2011-12-20 1.149 Expected time to next prime was (65/0.028) 2321 days[/COLOR][/FONT]
[COLOR=black][FONT=Verdana]2012-01-31 1.131 Expected time to next prime was (42/0.018) 2333 days[/FONT][/COLOR]
[COLOR=black][FONT=Verdana]2012-03-21 1.107 Expected time to next prime is (50/0.024) 2083 days[/FONT][/COLOR]

or if we look at the whole period from 2008-10-15 -- 2012-03-21 the expected time to next prime is (1243/0.640) 1942 days.[/QUOTE]
Feel another update [url=http://www.youtube.com/watch?v=JeRa3RtBiIU]coming on strong[/url]

2012-05-29 1.073

With luck you can work it out by yourself by now.

D

Expected new Mprimes < 79M
 

[QUOTE=davieddy;300665]
2012-05-29 1.073[/QUOTE]
2012-07-10 1.052

42days/0.021 = 2000 days expected to the next discovery.

I am inceasingly concerned that multiple cores, much faster
CPUs (LL tests) and GPUs have not reduced this expected time.

This concern is only amplified by the fact that there is
a concerted, short-sighted effort to shorten the LL tail,
while NEW LL tests and the progress of the wavefront* are neglected.

LL/TF 53M-63M is in a [B]MESS.[/B]
(No names. No pack drill)

David

* In the words of a WW1 song:
"There was a front; but only God knows where".

[QUOTE=davieddy;304417]short-sighted effort to shorten the LL tail,[/QUOTE]What if the next prime to be known is in that tail? The work still has to be done, no?

[QUOTE=Uncwilly;304448]What if the next prime to be known is in that tail? The work still has to be done, no?[/QUOTE]
Yep.

[QUOTE=Uncwilly;304448]What if the next prime to be known is in that tail?
[/QUOTE]
That remembered me when I was young and stupid and I put a clothes pin on the tail of the cat.. She was turning and turning and turning, like a tornado, and getting angrier and angrier and noisier and noisier... Someone you know?

I created a "countup" since last GIMPS prime on timeanddate.com.

I found these data:
[URL="http://www.mersenneforum.org/showpost.php?p=175938&postcount=16"]http://www.mersenneforum.org/showpost.php?p=175938&postcount=16[/URL]
[QUOTE]I have found it but strangely it shows that it was found on 2009-04-12 02:27 - that can't be secret for so long?[/QUOTE]

[URL="http://www.mersenne.org/"]http://www.mersenne.org/[/URL]
[QUOTE]On April 12th, the 47th known Mersenne prime, 242,643,801-1, a 12,837,064 digit number was found by Odd Magnar Strindmo from Melhus, Norway! [/QUOTE]

So I created the counter at location Melhus, Norway and starting April 12th 2009 at 4:27 am CEST (Central European Summer Time) which is 2:27 am UTC which primenet reported.


Counter:

[URL="http://www.timeanddate.com/countdown/to?iso=20090412T0427&p0=2571&msg=Time+since+last+GIMPS+prime"]http://www.timeanddate.com/countdown/to?iso=20090412T0427&p0=2571&msg=Time+since+last+GIMPS+prime[/URL]


Sadly its high, we passed 1,000 days long ago and 10[sup]8[/sup] seconds 28 days ago.

There have historically been large gaps between Mersenne primes (all comparisons refer to their logarithms):
[LIST][*]M#13 is four times as large as M#12[*]M#15 is twice as large as M#14[*]M#21 is twice as large as M#20[*]M#28 is twice as large as M#27, which in turn is twice as large as M#26[*]M#32 is 3.5 times as large as M#31[*]M#36 is twice as large as M#35[*]M#39 is twice as large as M#38[/LIST]
It's entirely possible that M#48 is going to be in the, say, 100M range, or even higher. My sixth sense tells me that discovering it is truly going to test our patience.

[QUOTE=ixfd64;304493]There have historically been large gaps between Mersenne primes (all comparisons refer to their logarithms):[/QUOTE]:no:
That is nice, but based upon the GIMPS era primes, it is not unlikely that 48, 49, and 50 lie below 60,000,000. Currently it looks like most of the new LL work is being done between 45 - 55,000,000.

[QUOTE=LaurV;304468]That remembered me when I was young and stupid and I put a clothes pin on the tail of the cat.. She was turning and turning and turning, like a tornado, and getting angrier and angrier and noisier and noisier... Someone you know?[/QUOTE]Yep.
Now leave off her tail and put a bowl of milk in front of her.

[QUOTE=Uncwilly;304504]but based upon the GIMPS era primes...[/QUOTE]

Irrelevant. Our run of good luck has no impact on the probabilities for the next gap size.

[QUOTE=Prime95;304508]Irrelevant. Our run of good luck has no impact on the probabilities for the next gap size.[/QUOTE]... if certain conjectures are correct.

[QUOTE=Prime95;304508]Irrelevant. Our run of good luck has no impact on the probabilities for the next gap size.[/QUOTE]

[QUOTE=cheesehead;304520]... if certain conjectures are correct.[/QUOTE]
Best summed up in one word: Poisson.

Expected new primes <60M is currently ~0.3

Probability of no more primes <60M is e[SUP]-0.3[/SUP] = 0.74

David

PS Bear this in mind when reading Crandall's obligatory page
[url=http://primes.utm.edu/notes/faq/NextMersenne.html]"Where is the next Mersenne Prime?"[/url]

[QUOTE=davieddy;304531]Bear this in mind when reading Crandall's obligatory page
[url=http://primes.utm.edu/notes/faq/NextMersenne.html]"Where is the next Mersenne Prime?"[/url][/QUOTE]

I think you meant Caldwell not Crandall.

[QUOTE=Prime95;304534]I think you meant Caldwell not Crandall.[/QUOTE]
Close but no cigar:smile:

D

[QUOTE=LaurV;304468]That remembered me when I was young and stupid and I put a clothes pin on the tail of the cat.. She was turning and turning and turning, like a tornado, and getting angrier and angrier and noisier and noisier... Someone you know?[/QUOTE]

[QUOTE=davieddy;304506]Yep.
Now leave off her tail and put a bowl of milk in front of her.[/QUOTE]
To pursue this analogy further:
a bowl of full creamed milk may tempt her more than a trough of skimmed.

What is the range of n for each interval?

[QUOTE=henryzz;304548]What is the range of n for each interval?[/QUOTE]Having taught physics, I can guess what you are on about.
log ("exponent") is ~ 1.48

Hope that helps.

D

PS It takes years to appreciate the distinction between a log and an exponent.

[QUOTE=Prime95;304534]I think you meant Caldwell not Crandall.[/QUOTE]
Of course not.
It was Chalsall

[QUOTE=davieddy;304572]

PS It takes years to appreciate the distinction between a log and an exponent.[/QUOTE]Not if you've seen a log-rolling contest.

[YOUTUBE]eusMzC7Rx7M[/YOUTUBE]

[QUOTE=cheesehead;304739]Not if you've seen a log-rolling contest.[/QUOTE]
No. But I've sawn a few.
There is a cheese rolliing contest (Gloucestershire?) that seems pretty dangerous to me.
They send some round(ish) ones down a steep hill, and some cretins attempt to follow them.

They occasionally sustain minor damage.

D

[QUOTE=davieddy;304748]
There is a cheese rolliing contest (Gloucestershire?) that seems pretty dangerous to me.
They send some round(ish) ones down a steep hill, and some cretins attempt to follow them.

They occasionally sustain minor damage.

D[/QUOTE]

[url]http://www.cracked.com/article_15209_the-10-most-insane-sports-in-world.html[/url]

[QUOTE=davieddy;304748]No. But I've sawn a few.
There is a cheese rolliing contest (Gloucestershire?) that seems pretty dangerous to me.
They send some round(ish) ones down a steep hill, and some cretins attempt to follow them.

They occasionally sustain minor damage.

D[/QUOTE]
I assume the ambiguity as to what's damaged is intentional. Either the cretins or the cheese wheels.

Need to Google cretins. I've heard the word and know it's an insult, but other than that, no idea.

[QUOTE=davieddy;304748]No. But I've sawn a few.
There is a cheese rolliing contest (Gloucestershire?) that seems pretty dangerous to me.
They send some round(ish) ones down a steep hill, and some cretins attempt to follow them.

They occasionally sustain minor damage.

D[/QUOTE]
Oh my, oh my, something is wrong with you... what's happening? you forget the [URL="http://www.youtube.com/watch?v=KOyQBSMeIhM"]youtube link[/URL].. :razz:

My previous pessimistic predictions:

M48 77,000,000
M49 205,000,000 (an expectable big gap)

Add:

M50 280,000,000

Time frames: 12/31/2013, 12/31/2016, 12/31/2017

[QUOTE=davar55;307174]My previous pessimistic predictions:

M48 77,000,000
M49 205,000,000 (an expectable big gap)

Add:

M50 280,000,000

Time frames: 12/31/2013, 12/31/2016, 12/31/2017[/QUOTE]

Actually, on 12/25/2010 you predicted 52,100,000 would be the next at the end of 2011 (12/31/2011 as noted in my spreadsheet).
On 5/24/2011 you said that 77,000,000 would occur by June 2012 (6/1/2012).
And "For M49: (big gap to) ~205,000,000 by 2/2015 reached analytically"

[QUOTE=davieddy;304417]2012-07-10 1.052

42days/0.021 = 2000 days expected to the next discovery.
[/QUOTE]2012-11-08 0.990

Expected time to next prime = 121/0.062 days = 1952 days.

I am reluctant to conclude that prospects are improving, because of the current focus on the tail of the LL wave: a distinctly short-term policy.

David

[QUOTE=davieddy;317542]2012-11-08 0.990

Expected time to next prime = 121/0.062 days = 1952 days.

I am reluctant to conclude that prospects are improving, because of the current focus on the tail of the LL wave: a distinctly short-term policy.

David[/QUOTE]
I've never taken a statistics class, so please be kind, but that 1952 days marks the point where there's a precisely 50/50 chance that a prime will already have been found, correct?

[QUOTE=jasong;317820]but that 1952 days marks the point where there's a precisely 50/50 chance that a prime will already have been found, correct?[/QUOTE]

No. We often think in terms of a Normal Distribution, where the "middle" is the same regardless of whether we have in mind the mean, the median, the mode, or some other measure or centrality. But when you get away from symmetric probability distributions, these things no longer coincide.

The probability model used here is the exponential distribution if we think of it as continuous, or the geometric distibution if we think of it as chunked into daily increments. David gave the mean (what we usually mean when we say average). You are asking for the median.

[QUOTE=jasong;317820]I've never taken a statistics class, so please be kind, but that 1952 days marks the point where there's a precisely 50/50 chance that a prime will already have been found, correct?[/QUOTE]
You are not a million miles out:

I am saying that the probability of no new Mprimes before d days is

e[SUP]-d/1952[/SUP].

So probability of no prime for 1952 days is 1/e = 0.37

e[SUP]-d/1952[/SUP] = 1/2 when d = 1952 * 0.69 = 1353

David

Brief update. The following are our next closest predictions:

[CODE]ET [COLOR="Green"]12/20/2012[/COLOR]
ewmayer [COLOR="green"]45,000,000[/COLOR] <-originally posted in '06 before M47 was found.
wreck 49,000,000 [COLOR="green"]1/1/2013[/COLOR]
Dubslow 60,000,001 [COLOR="green"]1/1/2013[/COLOR]
Oddball 78,000,000 [COLOR="green"]1/1/2013[/COLOR][/CODE]

We are now 3.5 sigma beyond the average gap interval. (11/28/12). (Further along if the gap between in '08 is ignored.

My Virgin Ivy's very first LL will be PRIME!!!
 

Drum roll for assignment until I actually ask for one .... soon.

[QUOTE=petrw1;320665]Drum roll for assignment until I actually ask for one .... soon.[/QUOTE]

Almost forgot....

49532123 - January 12, 2013

[QUOTE=petrw1;322036]49532123 - January 12, 2013[/QUOTE]
Previously you predicted 55,555,543 would be found prime on 5/5/2012.

[QUOTE=Uncwilly;322037]Previously you predicted 55,555,543 would be found prime on 5/5/2012.[/QUOTE]

Even Kreskin misses on rare ocassions :smile:

And JB already LL'd it as Non-prime.

Did anyone compute a composite of the 'guesses' herein?
An average of the guesses, weighted or not?
I'm curious about which guesses come nearest these averages.

[QUOTE=petrw1;322036]Almost forgot....

49532123 - January 12, 2013[/QUOTE]

The date has come and gone, but the exponent hasn't...keeping fingers crossed...

Maybe when people post their guesses, they should post how they came up with the guess. This would be different from crank postings because people wouldn't be claiming how awesome their method is, just how they came up with their guesses. If the guess(es) are correct, then we can re-visit their method.

[QUOTE=davar55;324711]Did anyone compute a composite of the 'guesses' herein?
An average of the guesses, weighted or not?[/QUOTE]
I can do that, I have all the data stored away. I will get to it a little later.

Edit:
62,714,889 on7/21/2010 (for all semi-serious guesses, including those carried over from the previous Guess Mxx threads)
64,987,482 on 12/27/2011 (for only those guesses since our last discovery)

[QUOTE=NBtarheel_33;324731]The date has come and gone, but the exponent hasn't...keeping fingers crossed...[/QUOTE]

You are correct....I tried to be nice my rearranging my worktodo to complete lower exponents first....

So, officially that assignment mentioned did NOT complete...

But on the other hand I did proclaim it would be Ivy's first completion that was the next prime....but alas...not to be...so maybe I had the exponent correct....Feb 12.

[QUOTE=davar55;324711]I'm curious about which guesses come nearest these averages.[/QUOTE]I missed that the first time around.
[QUOTE=Uncwilly;324745]62,714,889 on7/21/2010 (for all semi-serious guesses, including those carried over from the previous Guess Mxx threads)
64,987,482 on 12/27/2011 (for only those guesses since our last discovery)[/QUOTE]fatphil has come closest to both with 64,000,000 arriving on 6/30/2014, predicted on 9/12/2006

And the averages of those guesses that have not been surpassed by events:
66,135,910 on 11/7/2013

[QUOTE=davieddy;317542]2012-11-08 0.990

Expected time to next prime = 121/0.062 days = 1952 days.

I am reluctant to conclude that prospects are improving, because of the current focus on the tail of the LL wave: a distinctly short-term policy.

David[/QUOTE]2013-01-18 expected new primes <79.3M is 0.948
Expected time to next prime = 71/0.042 days = 1690 days.
[B]I am reluctant to conclude that prospects are improving, because of the current focus on the tail of the LL wave: a distinctly short-term policy.
[/B][QUOTE=jasong;324734]Maybe when people post their guesses, they should post how they came up with the guess. This would be different from crank postings because people wouldn't be claiming how awesome their method is, just how they came up with their guesses. If the guess(es) are correct, then we can re-visit their method.[/QUOTE]
OK: you're on:smile:
My guess is August 2017.
Exponent 79M

The reason is that I would bet evens (50/50) that these guesses will
be closer to the actual than everyone guessing differently and expect to make a profit.

David

[QUOTE=davieddy;325174]OK: you're on:smile:
My guess is August 2017.
Exponent 79M[/QUOTE]Let's see, you [URL="http://mersenneforum.org/showpost.php?p=176327&postcount=8"]earlier guessed that the number would be 60,000,000[/URL] and now you are changing it.
Also, the current best estimate for the P90-years for the range to 79.3M to go to zero is October 2019 (and dropping). If we expect 0.950 new primes in that range, then the likelihood of the next one showing up is, ~50% after 50% of the time it will take to complete the range. Half way from now to October 2019 is June 2016. So you are guessing long using your own basic logic. Also, the half-way point from the current lowest first LL to 79.3M is 61.8M. So, again your guess is too high.

[QUOTE=davieddy;325174]
The reason is that I would bet evens (50/50) that these guesses will
be closer to the actual than everyone guessing differently and expect to make a profit.[/QUOTE]
Correction:
To refute this, simply observe that the chance of the actual next Mprime exceeding the "expected" is 1/e while that of coming before it is 1 - 1/e.

[QUOTE=Uncwilly;325196]Let's see, you [URL="http://mersenneforum.org/showpost.php?p=176327&postcount=8"]earlier guessed that the number would be 60,000,000[/URL] and now you are changing it.
[/QUOTE]

[url=http://www.youtube.com/watch?v=GSxFiM4[/u8]Ain't it funny how time slips away[/url]

If we find the next prime in a yrar or so, my 5 year old "guess" won't look too bad will it?

David

[QUOTE=jasong;324734]Maybe when people post their guesses, they should post how they came up with the guess. This would be different from crank postings because people wouldn't be claiming how awesome their method is, just how they came up with their guesses. If the guess(es) are correct, then we can re-visit their method.[/QUOTE]
Expected new primes <72M is currently 0.693.

The probability of no new primes <72M is e[SUP]-0.693[/SUP] = 0.5
(Poisson distribution).

The probability of no new primes before the expected number has dropped by 0.693 is 0.5. (Same reason).
At the current rate (as monitored by me for years), this will take about 3 years.

I hope these observations will inhibit the dafter guesses which abound in this thread.

D

I am sure the American government knows the 48th Mp but they don't want to tell it to you...

[QUOTE=davieddy;325548]I hope these observations will inhibit the dafter guesses which abound in this thread.[/QUOTE]

Does not every untested candidate have the same chance of being a MP as every other untested candidate?

[QUOTE=chalsall;325619]Does not every untested candidate have the same chance of being a MP as every other untested candidate?[/QUOTE]
Nope

[QUOTE=chalsall;325619]Does not every untested candidate have the same chance of being a MP as every other untested candidate?[/QUOTE]No, some are vastly superior. See below.
[QUOTE=LaurV;325618]I am sure the American government knows the 48th Mp but they don't want to tell it to you...[/QUOTE]American government does not have MP's, they use Senators and Representatives.:davieddy:

[QUOTE=davieddy;325621]Nope[/QUOTE]

Please expand.

[QUOTE=chalsall;325624]Please expand.[/QUOTE]
1) MPs get thinner on the ground as they get higher (like all primes).
2) The amount of trial factoring affects it
3) I thought GPUto72 had something to do with this.

D

[QUOTE=davieddy;325629]1) MPs get thinner on the ground as they get higher (like all primes).
2) The amount of trial factoring affects it
3) I thought GPUto72 had something to do with this.

D[/QUOTE]

Flipping a coin...
:jokedrum:

[QUOTE=chalsall;325624]
[QUOTE=davieddy]
[I]Nope[/I]
[/QUOTE]
Please expand.[/QUOTE]

[code] N o p e [/code]

:smile:

[QUOTE=davieddy;325629]1) MPs get thinner on the ground as they get higher (like all primes).[/QUOTE]

Ah, stupid me. Of course...

That also explains why you're constantly lobbying to have low exponents be left to you personally.... :razz:

[QUOTE=ET_;325651]Flipping a coin...
:jokedrum:[/QUOTE]
Many words are spoken in jest.

Few of us cognoscenti around here doubt that GIMPS has had a lucky streak
How lucky?
7 "small" gaps in a row is like 7 "heads" in coin tossing.

D

[QUOTE=chalsall;325700]Ah, stupid me. Of course...

That also explains why you're constantly lobbying to have low exponents be left to you personally.... :razz:[/QUOTE]
Yep

[QUOTE=chalsall;325619]Does not every untested candidate have the same chance of being a MP as every other untested candidate?[/QUOTE]

[QUOTE=davieddy;325629]1) MPs get thinner on the ground as they get higher (like all primes).
2) The amount of trial factoring affects it[/QUOTE]
It os worth noting that the increasing probability of individual candidates is compensated for by the number of factors found. So the "expected number" is unaffected.

For further elucidation, I recommend William Blipp, or (if you are deperate) Ernst Mayer.

D

A prime has been reported and George is happy so far.
[SIZE="3"][COLOR="Red"][FONT="Arial Black"]All guessing is temporarily prohibited.[/FONT][/COLOR][/SIZE]

[QUOTE=Uncwilly;325892]A prime has been reported and George is happy so far.
[SIZE=3][COLOR=Red][FONT=Arial Black]All guessing is temporarily prohibited.[/FONT][/COLOR][/SIZE][/QUOTE]

This better not be a sick joke!

[QUOTE=c10ck3r;325901]This better not be a sick joke![/QUOTE]

qv: [url]http://mersenneforum.org/showthread.php?t=17704[/url]

[QUOTE=davieddy;325548]Expected new primes <72M is currently 0.693.

The probability of no new primes <72M is e[SUP]-0.693[/SUP] = 0.5
(Poisson distribution).

The probability of no new primes before the expected number has dropped by 0.693 is 0.5. (Same reason).
At the current rate (as monitored by me for years), this will take about 3 years.

I hope these observations will inhibit the dafter guesses which abound in this thread.

D[/QUOTE]
Ironic that you make this post predicting 3 years just 2 days before it appears another one has been found. Would this then be an 8th 'heads'?

I think I said after the last prime (or did I dream it?) that they were obviously starting to get 'closer together' again. Looks like I'm right. :grin:

[QUOTE=bcp19;326335]Ironic that you make this post predicting 3 years just 2 days before it appears another one has been found. Would this then be an 8th 'heads'?[/QUOTE]

Yes, this will be the 8th heads. The logic behind davieddy's estimates is valid. If we don't get substantial improvements in number of volunteers or speed of the hardware, then we are facing an expected ~4 years between finds. The scary thing is we could easily see a decade long drought if we run into a desolate area like between 216091 and 756839.

[QUOTE=Prime95;326373]Yes, this will be the 8th heads. The logic behind davieddy's estimates is valid.[/QUOTE]

Hmmm.
If this new find is or thm rumour, could you let us know so that we can play our traditional guessing game?

I'll start the ball rolling by pointing out that by comparing a "small" gap with a coin toss turning out to be "heads", I was referrring to 50/50 propositions. Assuming you took that on board, I would infer that this find is <56M.

David

[QUOTE=Prime95;326373]Yes, this will be the 8th heads. The logic behind davieddy's estimates is valid. If we don't get substantial improvements in number of volunteers or speed of the hardware, then we are facing an expected ~4 years between finds. The scary thing is we could easily see a decade long drought if we run into a desolate area like between 216091 and 756839.[/QUOTE]
Depending on the amount of slowdown, you could probably get new members by saying that the combination of Moore's Law(if it's still around) and the odds of the next prime indicate the odds of finding a new one soon are increasing.

[QUOTE=Prime95;326373]Yes, this will be the 8th heads. The logic behind davieddy's estimates is valid. If we don't get substantial improvements in number of volunteers or speed of the hardware, then we are facing an expected ~4 years between finds. The scary thing is we could easily see a decade long drought if we run into a desolate area like between 216091 and 756839.[/QUOTE]

Even worse, a gap like the one between 127 and 521, that one lasted 38 years.

[QUOTE=philmoore;326458]Even worse, a gap like the one between 127 and 521, that one lasted 38 years.[/QUOTE]
It lasted longer than that:smile:
As if sad anecdotes about the fate of Alan Turing were'nt numerous enough already,apparently he missed finding 521 because he only had 1024 bits od RAM at his disposal!

David

Wikipedia
[QUOTE]Mersenne primes take their name from the 17th-century French scholar Marin Mersenne, who compiled what was supposed to be a list of Mersenne primes with exponents up to 257. His list was largely incorrect, as Mersenne mistakenly included M67 and M257 (which are composite), and omitted M61, M89, and M107 (which are prime). Mersenne gave little indication how he came up with his list.[6]
[/QUOTE]
Sounds like he should be confined to Misc. Math Threads

[QUOTE=Flatlander;326481]Wikipedia
[QUOTE]Mersenne primes take their name from the 17th-century French scholar Marin Mersenne, who compiled what was supposed to be a list of Mersenne primes with exponents up to 257. His list was largely incorrect, as Mersenne mistakenly included M67 and M257 (which are composite), and omitted M61, M89, and M107 (which are prime). Mersenne gave little indication how he came up with his list.[6] [/QUOTE]

Sounds like he should be confined to Misc. Math Threads[/QUOTE]
On the bright side, [URL="http://primes.utm.edu/mersenne/LukeMirror/mersenne.htm"]he wasn't an Inquisitor[/URL], some people say. Indeed, he was probably quite a pleasant fellow.

Nobody expects...

Spam, spam, spam, spam, spam, spam! Glorious spam!

[QUOTE=Prime95;326373]If we don't get substantial improvements in number of volunteers or speed of the hardware, then we are facing an expected ~4 years between finds.[/QUOTE]

Even with sustained throughput jumping from 95 TFLOPS in November to 124+ TFLOPS as I write this? Whoever brought those Amazon computers online (hopefully someone who remains interested, has the ability to maintain Prime95 on them, and has permission to do so) really gave us a significant boost. Curtisc, I am sure, will upgrade his hardware as time goes on. We've barely scratched the surface of AVX. And let's not forget GPUs.

Have we ever considered putting together a basic package of information, getting started tips, and perhaps a Prime95 install CD, and sending it to various universities, schools, businesses, etc. to drum up potential newcomers to the project? Something concrete that a teacher (or students) could then present to the IT staff and say "hey, we really are interested in this GIMPS project, here's how we join, this is why it won't totally bollocks up our computers, and here's 10 reasons why humanity will suffer if we don't help find a new Mersenne prime." :smile: I would be happy to help develop some print materials - this might be an excellent idea with the coverage of GIMPS that will be occurring in the next few days and weeks.

The leaderboard on the home page says that we have 730,000 or so CPUs registered right now. Imagine if we could just double that number. How about raising it an order of magnitude?! And it's not that hard to do, when you consider that there are billions of potential candidate computers out there. But I believe that while we have done a good job of bringing users to GIMPS, the next wave of success will depend on our bringing GIMPS to users.

[QUOTE]The scary thing is we could easily see a decade long drought if we run into a desolate area like between 216091 and 756839.[/QUOTE]

Double the number of GIMPSters and you cut this to a five-year drought. Increase by a factor of ten and you cut this to a year's drought!

[QUOTE=Flatlander;326481]Sounds like he should be confined to Misc. Math Threads[/QUOTE]

:rofl:

That made my day!! I wonder if he used an EXACT THEOREM, or if it was divine inspiration? LOL!

[QUOTE=Batalov;326514]On the bright side, [URL="http://primes.utm.edu/mersenne/LukeMirror/mersenne.htm"]he wasn't an Inquisitor[/URL], some people say.[/QUOTE]

*Some* people say. The ones that got away. As opposed to those who got strapped to the comfy chair to listen to the voice of Ben Stein recite every single digit of 2^43,112,609-1. (Yes, little did Mersenne know that his favorite numerical torture implement was itself a Mersenne prime!)

[QUOTE]Indeed, he was probably quite a pleasant fellow.[/QUOTE]

Actually, from what we are learning in this thread, he was quite a crank, probably scrawling cmd-like messages and random claims of primality on the walls of the abbey. What if...he was Blazys-esque?!?

[QUOTE]Nobody expects...

Spam, spam, spam, spam, spam, spam! Glorious spam![/QUOTE]

THE SPANISH INQUISITION! Our weapons are trial factoring...trial factoring and LL testing...no wait, our two main weapons are trial factoring and LL testing...and double-checking...no wait, our three main weapons are...Cardinal Biggles, you'll have to say it!