Is that the new credits at the end ?

If so there`s not as much variation as I would have expected !

That being the case it could be said 5.5 had little effect, of course there are other cases that may be made.

The other strange thing is that users dropped dramatically yet points didn`t. Yet we all know lower credits per person have been the order of the day since the introduction of the new system. Seem`s strange !

I`ll take it as read user`s means just that and not host`s !

---

Not all Czech`s bounce but I`d like to try with Barbar ;-)

Make no mistake This IS the TEDDIES TEAM.

Is that the new credits at the end ?

If so there`s not as much variation as I would have expected !

That being the case it could be said 5.5 had little effect, of course there are other cases that may be made.

It's exactly what I would have expected, as Rosetta gives the median of the claimed credits to all. The sum will be equal, but for the single user it will be quite different. My Linux stock client gets a lot more credits per hour now, while I expect some Windoze "opti" gets less. But we both get nearly the same per science now.

Is that the new credits at the end ?

If so there`s not as much variation as I would have expected !

That being the case it could be said 5.5 had little effect, of course there are other cases that may be made.

Well, if you want my Opinion. Which is just an opinion. The last two days show a drop in credit/day, but it's not severe. Given that I'm experiencing a slight boost in credit/hour or credit/day I'd assume others are as well. That said, I've also heard that some have dropped rosetta. Perhaps the drop we might have seen by those leaving is being offset by the "extra" credit the other users are getting? Perhaps the users of third party software having their granted credit dropped is also a part of this? It all kind of mixes together. I wish there was some way to show "models/day" or something, but since those have varying run times, it wouldn't help figure this out.

there are problems of timing with my above theory, in that Some users stopped last Sunday or there abouts, so it's really tough to say one way or the other for me, as to exactly why it's the way it is.

which is why I said in the first post "I don't know what it means". lol

My credits per hour are slightly higher now than previously - probably for most/all official client users on Windows, and this might well be more profound for P4 and moreso for Linux users. I don't think the use of the optimised clients is that widespread to offset this - its obviously a larger proportion of the power users but there's a lot of production from team none and a lot of people just install and forget.

I don't know how to explain the jump in users though - did someone mention that Predictor had no work for a while?

---

My credits per hour are slightly higher now than previously - probably for most/all official client users on Windows, and this might well be more profound for P4 and moreso for Linux users. I don't think the use of the optimised clients is that widespread to offset this - its obviously a larger proportion of the power users but there's a lot of production from team none and a lot of people just install and forget.

I don't know how to explain the jump in users though - did someone mention that Predictor had no work for a while?

The jump in users is easier for me, and that's the reason I posted it. Remember that little jump is roughly (just guesstimating from chart) 200 users, but we KNOW that aprox. 200 users joined/rejoined on the day the new credit system was announced.

In 59-66 days we'll see the affect of those who have left.

My credits per hour are slightly higher now than previously - probably for most/all official client users on Windows, and this might well be more profound for P4 and moreso for Linux users. I don't think the use of the optimised clients is that widespread to offset this - its obviously a larger proportion of the power users but there's a lot of production from team none and a lot of people just install and forget.

I don't know how to explain the jump in users though - did someone mention that Predictor had no work for a while?

The jump in users is easier for me, and that's the reason I posted it. Remember that little jump is roughly (just guesstimating from chart) 200 users, but we KNOW that aprox. 200 users joined/rejoined on the day the new credit system was announced.

In 59-66 days we'll see the affect of those who have left.

~30 days, the charts are of 'have granted credit in past month'

---

Team mauisun.org

The new system is Lotto. My last 2 results:
CPU-time: 10,061.41 GC: 68.19
CPU-time: 9,951.78 GC: 105.76

So much more points for less work? ...

:-) At last, someone concerned about getting too many points... ;-)

In a more serious vein, there seem to be two items of concern:
1. Variation in the amount of points between problems that appear to be similar.
2. A seeming lack of correlation between CPU time and points.

For (1.), this is my impression of the process: The calculations required to develop a trajectory--a decoy-- are enormously complex. These are not closed-form direct calculations as in finding the answer with a direct formula. Rather, there are numerous random choices ("Monte Carlo" methods). In addition, the "landscape" over which the search is being made for a "deep valley" of low energy is apparently extremely irregular, with huge thin walls, spikes, wide valleys at high elevation, and perhaps small deep holes on otherwise high terrain. Then calculations are trying to find the deepest point on a very large landscape by hunting around, much like one in total darkness, testing the nearby terrain to see of it gets lower. This is done
by randomly poking around. But just because the terrain gets higher doen't mean that there isn't a beautiful deep valley just over the rise. The calculations are also guided by heuristics, or scientific guesses, to allow the search to continue uphill for a while. Other methods allow the search to hop over a relaively long distance in a random direction just to see if there are lower spots not too far away. It's conceivable that in this terrain, the very lowest point--the global minimum--may be an extremely deep golf hole on a green at the top of a mountain. (Forget about getting your ball back!) With the terrain being so ragged and onpredictable, with different starting points for the search, and with the moves or jumps being random, it seems reasonable that the CPU time to get to some localized deep point--a local minimum--could vary of an extremely wide range.

For (2.), one has to consider the probability that the mix of instructions actually executed during these extremely complex random guided searches does not mirror the mix of instructions used in some benchmark, including the "benchmark" consisting of the average across a number of other computers with different instructio sets. For example, if your computer's processor happens to encounter a whole series of operations that are carried out within the processor's caches, it will literally fly through that series. The avarage computer may not have had that advantage, so it slogs through the operations using operations against the main processor RAM, spending more cycles in the process. As a result, your computer gets done faster, and gets the points sooner, resulting in more points per hour. Asking forgivenes in advance for the analogy, consider that our computers are like trucks pulling a weighted sled around a track. Some people have more horsepower, and can consistently get a sled around faster than others. Sometimes, for some sleds, options like XM radio may help a truck to drag them even faster, while sometimes XM won't help at all. What counts in the end is getting the sled across the finish line.

I apologize in advance if I have mangled any concepts beyond recognition; I'd rather seek forgiveness rather than ask permission... :-)

Happy crunching!

All that rhetoric aside, let's look at this example.

2 hr 45 minutes to do 11 decoys, and 2 hr 47 minutes to do 11 decoys.

1 WU gets over 50% more credit than the other.

Now, I could understand this variation if we were only talking about one decoy, but this is 11 per WU! So for this PC, it took roughly 15 minutes to do each decoy, regardless of the type!

Why the big difference in score then?

See Post moved to "Report Unusual Credits thread.

There is a nice upward trend in active users, active teams, new users per day, and new teams per day... nice to see! Look at boincstats here. :-)

There is a nice upward trend in active users, active teams, new users per day, and new teams per day... nice to see! Look at boincstats here. :-)

...yes, Rosetta active teams are rising (by about 1% or 20 teams in the last couple days) while Einstein and SETI active teams are more or less flat. I'd like to hear Occam's take on that (haven't seen him on the boards for some time, perhaps he stopped crunching ;-).

---

Team betterhumans.com - discuss and celebrate the future - hoelder1in.org

August was always going to be weird. A lot of people either upped their Rosetta quota or suspended other projects during CASP. When CASP finished, it has to take some time before LTD's even out and other projects catch up what they lost.

For the record, since the 23rd all of my granteds have been greater then the "claimed", both machines standard software, under Windows on Intel P-IV chips.

The slower machine that crunches 20hr wu's the mark up is pretty constant, claim ~140 granted ~170, the other, faster, machine is set to crunch 12hr wu's, these vary a little more, typically claiming ~75 and being granted between 89 - 109. I don't know if that is of interest to anyone, but there it is anyway.

I also suggested one of my wu had been granted excessively, but that was before the switch, and was a known issue. It is in the other thread, here.

---

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

There is a nice upward trend in active users, active teams, new users per day, and new teams per day... nice to see! Look at boincstats here. :-)

...yes, Rosetta active teams are rising (by about 1% or 20 teams in the last couple days) while Einstein and SETI active teams are more or less flat. I'd like to hear Occam's take on that (haven't seen him on the boards for some time, perhaps he stopped crunching ;-).

Perhaps Occam is on holiday for August... ;-)

What would be more helpful in breaking down the numbers on BOINC stats is if BOINC would break the total vs. active members in all graphs. For instance, the pie graph that shows %team members vs. % nonteam members doesn't specify if those percentages are from the total or the active members. It would also help if a better representation of active team credits vs. active Non-team credits could be shown for a more accurate picture. That is to say, how many active teams has Rosetta gained and how many Non-team members were gained as well.

In short, BOINC stats does not do a very good job in representing the numbers at least in the short term.

---

There is a nice upward trend in active users, active teams, new users per day, and new teams per day... nice to see! Look at boincstats here. :-)

I don't think this has ANYTHING to do with the new credit system, and everything to do with Predictor going on hiatus, and the temperatures coming down, a lot of people don't run their systems in the summer. All the Predictor people needed a new Boinc project, (me included) or are coming back online for the winter, and if they stayed with disease based projects, this would be the logical choice. That project was getting over 3,000,000 credits a day, add the people coming back online, disperse that out over all the projects and I think you will see where the increase really came from. I believe if it were not for the timing you would be seeing a decrease.

From what I am seeing this project is now granting less credit than Predictor. My credit per hr is now down to under what I was getting on Predictor, and continues to fall, even after switching from 10hr runs to 4hr runs.

I came here from Predictor and am using cruncher3’s client now, but for the first week I was here before the new credit system I ran the stock client and was granted credit equal to or greater than Predictor. I then ran the optimized client and was granted credit of about twice my average on Predictor. Now it’s less with either. I tried the stock client when the credit system changed for a day and the results were even worse, then switched back to the optimized, it still continues to fall. If it continues to fall, and gets to the point where its 200 credits a day less than Predictor, I will be looking for a different project. 200 x 365 = 73,000 that’s almost more credit lost per year that the few people in here that fought to change it have combined. I don’t jump around to different projects; I crunch one at a time.

I run 4 highly tweaked 2600 Mobile Xp's all water-cooled, and highly overclocked. They benchmark out between FX 53's and FX 55's. Inexpensive equipment expensively tweaked to the hilt. Over in Predictor there was no P4's or dual Zeon’s, or X2 under a 4200 that could match my output per machine. Now I'm getting stomped by lowly P4's. Credits are important for those of us that tweak to an extreme, not to say that the benefit to science is not, but I do this to see just how much I can squeeze out of a rig to crush the more expensive hardware. I do this by credit measurement, if science benefits as a byproduct good. But I will not sit here and watch the overclocking art, be basified by an unfair credit system. More credit for more work performed? I don't see it. I am seeing the opposite.

So much for cross platform uniformity.

This project seems now to be one of the projects that grant the least credits per hr of computer time. Hope all the whiners are happy. And I don't care what someone else's spreadsheet shows; I'm going by actual knowledge of what I was getting there, and what I am getting here.

It looks like me and my 24/7 10.4 gigs of CPU power are out of here, if nothing changes in the next 2 weeks, (that’s what the powers that be asked for) or my credit per hr falls to an unacceptable level. I'd start looking how to tweak the credit system before the mass exodus happens.

If I can't find another DC disease based project that has not been screwed up by the very vocal minority, I'll be shutting them down, selling all the parts, and go and buy myself something nice. My wife would be very happy that the 2 new rigs that I am in the middle of building will not be built.

I'm just sick and tired of the fact that a very vocal minority can have such an impact on a project to such an extreme measure of unbearable damage. All the baiting, backstabbing, backdating, wanting of all relevant information of old credits, (just to keep pummeling this subject into submission), and Boinc cross platform uniformity bull crap, has damaged this project beyond repair.

This will be my one and only post in this message board. I will not reply to anyone’s response to any of the facts or feelings in this post, they are mine, and I don’t care what you think. I will not be baited, swayed, or in any other way respond to anyone else's responses to this post, so don't even try.

I was not going to post until the end of the 2 weeks or the credit per hr fell below Predictor, but the fact that some here seem to think that the new credit system is the reason for the rise in people had to be addressed. You are living in a dream world.

JMO

There is a nice upward trend in active users, active teams, new users per day, and new teams per day... nice to see! Look at boincstats here. :-)

... From what I am seeing this project is now granting less credit than Predictor. My credit per hr is now down to under what I was getting on Predictor, and continues to fall, even after switching from 10hr runs to 4hr runs.

Should you decide to stay, it's my experience that you will most likely get more Granted Credit per unit CPU Time by increasing the run lengths--not decreasing them. If you're willing to try the experiment, that would be great.
Edit: misspelling of 'my'

Stevea, I would like to see those same statistics in 30 days time, but not just users who have gotten credit, but hosts which have gotten credit. After 30+ days the statistics will reflect what has been really going on here more accurately. Remember one user can have many hosts and after 30 days they will dissapear off of those statistics but the total amount of users will always increase. I'm sure the project team know there is a drop of work coming in and are concerned about this. If they are not concerned, they should be. But that is up to them entirely.
As from tomorrow, I will be at WCG, I'm sure some of you will be glad to hear.

---

See you around, Whl.

I'm here for the duration because I believe the ratings here are the best of any project. At least Rosetta@home doesn't grade us on connection speed, hard drive space, and installed memory.

Hi Vester, I'm sure I will see you around mate. :-) You know I respect you a lot and will always wish you well whatever you do. If there had'nt been all the talk about punishing people on Ralph, I might have still been here myself. As I've said a few times here before, it is not all about credits for me, there is a lot more to it than that. After all I am moving now with most of my team mates to a system that operates a quorum of 3.

Good luck Vester mate.

---

Ted3:

The new system is Lotto. My last 2 results:
CPU-time: 10,061.41 GC: 68.19
CPU-time: 9,951.78 GC: 105.76
... "Even out over time" ...

They do seem to even out over time. Although there's a few WUs with values that should be reported.

You've got two systems:
The lesser system has credit/seconds that prior to Aug 23, was hovering around 0.003 credits/second and since Aug 23 is just under 0.004 credits/second. There's two results that I'll post to the unusual credit list.

graph these 80 points, and see for yourself:
0.003785
0.003686
0.004013
0.003821
0.003726
0.00384
0.003547
0.003677
0.003671
0.003814
0.003813
0.003281
0.003475
0.003804
0.003571
0.003581
0.00371
0.003584
0.003785
0.004005
0.003045
0.003086
0.003051
0.003022
0.003501
0.003497
0.003005
0.003197
0.003241
0.002927
0.002899
0.002922
0.003029
0.003049
0.002858
0.002983
0.003059
0.008548
0.003168
0.003189
0.003406
0.003089
0.002956
0.003046
0.003393
0.003009
0.003053
0.003102
0.002817
0.002925
0.003161
0.002973
0.003123
0.003018
0.002932
0.003013
0.002925
0.002908
0.002745
0.002927
0.002908
0.002966
0.003082
0.00279
0.002837
0.00312
0.003037
0.002743
0.00325
0.003372
0.00284
0.003272
0.003166
0.00279
0.003339
0.002269
0.003129
0.003158
0.003327
0.003319

---------
The faster machine hovered around 0.006 credits/second until Aug 23rd, when it moved to hovering around the 0.0075 range. 6 WUs on the faster machine have out of spec credits, and there's a fair number of errors (both of which I'll post to the correct thread.)

I'm just eyeballing these charts, and it seems pretty easy to pick out the ones that deserve reporting. The majority of the points on the chart are pretty constant.

---

I'm confused. Someone please explain this graph:

http://www.boincstats.com/charts/chart_uk_rosetta_project_new_credits.gif



Personally, my chart shows a drop by about 50%, and my team's chart shows the same. I have looked at many of the top 25 teams out there, who have contributed 45% of the total credits, up to 7% each, and they are *way* down with the new credit system.

So how is it that the total credits have not dropped off? Increase in membership? Can it really be that much?

---

Reno, NV
Team: SETI.USA

Here is a list of WUs that errored out. With the last decoy being bad, I can see it not having any credit granted, but for one of mine having gone 51k seconds, I'm suprised that I don't have any decoys that validated and got credit. The other two systems had a 3 hour run time, and should have created a few decoys before hitting the one that errored out.

Are these errored results being uploaded to the server and the models/decoys that didn't error out being validated? (and if not, why not?)

---