Suggestions and Ideas

Discussion regarding the making of b-games.org a BOINC competition site.
Post Reply
Skillz
Site Admin
Reactions:
Posts: 1854
Joined: Sun Sep 15, 2019 3:03 pm

Suggestions and Ideas

Post by Skillz »

I would really like to hear suggestions and ideas from all teams who would like to participate.

My goal here is to not make this a "one man" show, but have multiple people from multiple teams all working together to make this work. This should alleviate the burden of one single person running this.

With that being said, all ideas and suggestions are welcome.
User avatar
Coleslaw
Reactions:
Posts: 22
Joined: Sat Dec 26, 2020 8:41 am

Re: Suggestions and Ideas

Post by Coleslaw »

So, I finally got to skim over a lot of the suggestions already made in other threads. However, some of them were made before some of these others. Therefore I will try to address the concerns in this thread per the topic.

I don't like the idea of having "Test Sprints" that award points for the event. To me, this is no different than an actual sprint. Teams will still run these to get extra points or a leg up. I feel this will do deliberate harm to a project rather than honest inconvenience. Any stress testing of a project should not be done without an award in sight as that is what brings out those that really break shit. Also, having the project sign off on any stress testing of the project would be respectful. I'm assuming this is intended. When I ran DC-Vault, I pretty much made this a mandatory practice and everyone was content doing it when adding new projects. We used the BOINCStats team challenges platform for this as it psychologically separated the test from the ongoing contest with DC-Vault. The stress test should also only be done once to prove they are capable. No other stress testing should occur unless the project has shown inability to keep up with modern growth of hardware. No "new" projects IMHO should be added mid year. So, a stress test on them should occur before ever being added and thus should not count as points for the event.

I really don't like the idea of user manually signing up. A simple Team sign up should suffice. I know on paper it sounds great to fix a lot of the issues everyone has been having with Formula-BOINC but one of the first question new users always ask when volunteering for an event is if they have to do anything other than run the project. They aren't just doing this to get signed up correctly but are also weighing their options on if the event is too much of a pain in the ass to set up. You may think it is no big deal, but it is already a pain in the ass to set up the client, the projects (which all are different), get joined to your team, enable stats exporting (again depends on project), and finally the ability to start stress testing and learning what is needed to run the project(s) properly before even getting to the competition side of things. A lot of users are happy to support their team as long as they don't have to do anything but set and forget. Even with a one time manual sign up, it is daunting with all the other things you must do these days. The reason so many still prefer Folding@home is because it is severely dumbed down for them. BOINC is just to advanced in the sense of options and configuring. It is also why a lot of the former heavily active users eventually settle for just running WCG when they "retire".

The number of sprints is an issue with Formula-BOINC and will most likely be here as well. The first year someone joins the event, it seems cool and fun. However after doing it a year and possibly two, many users start to feel sprint fatigue. Considering that there are already the PrimeGrid series challenges (which burns people out on PG) and the SETI.Germany events, having a gazilion sprints to juggle on top of it just becomes too much work. This of course leads teams to cut ties with some of the long standing events that are popular.

I know [H] has had to drop some events in order to weigh whether it was a conflict or not to the year goals. The reason I bring this up is that if there is a manual team sign up, it needs to be yearly. We heavily considered having [H] removed from Formula-BOINC as we did not think having our name low on the competition rankings to fairly reflect our capability. However, we also don't want to tell team members that they cannot partake. So, we kept it there. Other teams may not make the choice we did and I can tell you that even after years of having FB out there, many teams and individuals have still never even heard of it....... There was no marketing. Not even from the teams supporting it. I spent a lot of time and effort over the years promoting my team to make them appealing to new users and to be respected in the community. People do see these things and make assumptions. I think by having the team sign up each year will also weed out those that are not aware of the event. It also prevents people from having the argument like they used to ban Gridcoin. Gridcoin has an official policy to NOT participate in any challenges they have not been directly invited to. They have suffered more attacks and grief than any other team out there for pathetic and stupid reasons. They took on this policy so that their name isn't continued to be used as an attack weapon and to just try to co-exist. By making it a team registration EVERY year, means they are actively wanting to compete. No more people saying "they don't even care about" or "they don't even know they are participating" kind of excuses.

I like that it was suggested to make a quarterly point system for the Marathon to help address when projects come and go mid year. Personally, I don't think any new projects need to be added mid year but when projects die, they definitely shouldn't be weighed as heavy as ongoing projects. This was one of my suggestions to FB years ago and had gone unheard or unconsidered for years.

I definitely prefer BOINC-Games over b-games. I think it stands out and is more clear to the event. I don't think we have to mimic Formula-BOINC altogether though as it will certainly feel like a knock-off. But if that is how people want to go, then I am fine with it. I would encourage you to possibly integrate crypto options for supporting the event. Possibly even accept Gridcoin to show that there is no animosity there. You could even accept Obyte and Neummanium if you wanted to support the other crypto teams. I know that some feel that crypto perverts the community but quite frankly there aren't many teams that aren't claiming crypto in some way within its members. So, I don't see how they can get around it.

I may have more suggestions and feedback later. But to be clear, these are my thoughts and not those of the entire [H] team.
User avatar
Coleslaw
Reactions:
Posts: 22
Joined: Sat Dec 26, 2020 8:41 am

Re: Suggestions and Ideas

Post by Coleslaw »

Going further on the "test sprints". If you would want to show some kind of acknowledgement for those that diverted resources to test projects that were not getting "points" for the event, then you could make consolation prizes. Kind of like the "miss congeniality" award. These teams are the ones that sacrificed some potential scoring positions to help improve the event. You could even have a separate scoring system for them to show how they contributed more or less to others in the side event for testing. This doesn't even have to be tied to points. It could be for various things including feedback. You could set up a "bounty" list worth certain points. Compete the assessments for each bounty and show you were trying to contribute in some way...
Skillz
Site Admin
Reactions:
Posts: 1854
Joined: Sun Sep 15, 2019 3:03 pm

Re: Suggestions and Ideas

Post by Skillz »

Coleslaw wrote: Sun May 16, 2021 11:04 am So, I finally got to skim over a lot of the suggestions already made in other threads. However, some of them were made before some of these others. Therefore I will try to address the concerns in this thread per the topic.

I don't like the idea of having "Test Sprints" that award points for the event. To me, this is no different than an actual sprint. Teams will still run these to get extra points or a leg up. I feel this will do deliberate harm to a project rather than honest inconvenience. Any stress testing of a project should not be done without an award in sight as that is what brings out those that really break shit. Also, having the project sign off on any stress testing of the project would be respectful. I'm assuming this is intended. When I ran DC-Vault, I pretty much made this a mandatory practice and everyone was content doing it when adding new projects. We used the BOINCStats team challenges platform for this as it psychologically separated the test from the ongoing contest with DC-Vault. The stress test should also only be done once to prove they are capable. No other stress testing should occur unless the project has shown inability to keep up with modern growth of hardware. No "new" projects IMHO should be added mid year. So, a stress test on them should occur before ever being added and thus should not count as points for the event.
Well the point of running the test sprints is to test the server and since teams will be diverting resources outside of the competition it seems only fair to reward them with points. The amount of points earned has yet to be determined, but it will not be a lot. 5 points or less and you'll get the points no matter what place you end up. IE: First place will get 5 points, 3rd place will get 5 points, last place will get 5 points.

If anything, teams will just throw 1 or 2 computers/hosts on the project just to accumulate some points.
Coleslaw wrote: Sun May 16, 2021 11:04 amI really don't like the idea of user manually signing up. A simple Team sign up should suffice. I know on paper it sounds great to fix a lot of the issues everyone has been having with Formula-BOINC but one of the first question new users always ask when volunteering for an event is if they have to do anything other than run the project. They aren't just doing this to get signed up correctly but are also weighing their options on if the event is too much of a pain in the ass to set up. You may think it is no big deal, but it is already a pain in the ass to set up the client, the projects (which all are different), get joined to your team, enable stats exporting (again depends on project), and finally the ability to start stress testing and learning what is needed to run the project(s) properly before even getting to the competition side of things. A lot of users are happy to support their team as long as they don't have to do anything but set and forget. Even with a one time manual sign up, it is daunting with all the other things you must do these days. The reason so many still prefer Folding@home is because it is severely dumbed down for them. BOINC is just to advanced in the sense of options and configuring. It is also why a lot of the former heavily active users eventually settle for just running WCG when they "retire".
Making this process as easy as possible will be the number one priority. This is the only way to sure-fire way to prevent participants from switching teams mid-race; specifically during a Sprint. If they want to crunch on another team to help them, then they'll have to create a new account, download new tasks and crunch them. No more sitting on millions of points worth of a bunker and then spread it around to different teams that will benefit their team.
Coleslaw wrote: Sun May 16, 2021 11:04 amThe number of sprints is an issue with Formula-BOINC and will most likely be here as well. The first year someone joins the event, it seems cool and fun. However after doing it a year and possibly two, many users start to feel sprint fatigue. Considering that there are already the PrimeGrid series challenges (which burns people out on PG) and the SETI.Germany events, having a gazilion sprints to juggle on top of it just becomes too much work. This of course leads teams to cut ties with some of the long standing events that are popular.
I agree 100%. I believe the schedule will be no more than 12 sprints per year. (Not including the test sprints) So basically 1 per month.
Coleslaw wrote: Sun May 16, 2021 11:04 amI know [H] has had to drop some events in order to weigh whether it was a conflict or not to the year goals. The reason I bring this up is that if there is a manual team sign up, it needs to be yearly. We heavily considered having [H] removed from Formula-BOINC as we did not think having our name low on the competition rankings to fairly reflect our capability. However, we also don't want to tell team members that they cannot partake. So, we kept it there. Other teams may not make the choice we did and I can tell you that even after years of having FB out there, many teams and individuals have still never even heard of it....... There was no marketing. Not even from the teams supporting it. I spent a lot of time and effort over the years promoting my team to make them appealing to new users and to be respected in the community. People do see these things and make assumptions. I think by having the team sign up each year will also weed out those that are not aware of the event. It also prevents people from having the argument like they used to ban Gridcoin. Gridcoin has an official policy to NOT participate in any challenges they have not been directly invited to. They have suffered more attacks and grief than any other team out there for pathetic and stupid reasons. They took on this policy so that their name isn't continued to be used as an attack weapon and to just try to co-exist. By making it a team registration EVERY year, means they are actively wanting to compete. No more people saying "they don't even care about" or "they don't even know they are participating" kind of excuses.
To make things easier, teams/members will only have to sign up ONCE in their lifetime. However, each year they will need to actively enter into the contest. Not 100% sure how to do that, but it will most likely just be a link sent to their email they can click to enter and/or some form on their profile they'll have to select to enter the new year.
Coleslaw wrote: Sun May 16, 2021 11:04 amI like that it was suggested to make a quarterly point system for the Marathon to help address when projects come and go mid year. Personally, I don't think any new projects need to be added mid year but when projects die, they definitely shouldn't be weighed as heavy as ongoing projects. This was one of my suggestions to FB years ago and had gone unheard or unconsidered for years.
I am not sure how I am going to go about on the points structure. The reason for adding projects to replace old projects is simply so the pool of eligible sprint candidate projects will remain close to the same. Plus, if a project drops out after the 2nd quarter then the next 2 quarters will be less options to gain points. Adding a project in that slot will alleviate that problem. However, if the start of the year there are 20 projects in the marathon, then there will never be more than 20 marathon projects throughout the year. Only the following year will a new project be added to the current list.
Coleslaw wrote: Sun May 16, 2021 11:04 amI definitely prefer BOINC-Games over b-games. I think it stands out and is more clear to the event. I don't think we have to mimic Formula-BOINC altogether though as it will certainly feel like a knock-off. But if that is how people want to go, then I am fine with it. I would encourage you to possibly integrate crypto options for supporting the event. Possibly even accept Gridcoin to show that there is no animosity there. You could even accept Obyte and Neummanium if you wanted to support the other crypto teams. I know that some feel that crypto perverts the community but quite frankly there aren't many teams that aren't claiming crypto in some way within its members. So, I don't see how they can get around it.
Can you elaborate on this? You want me to give crypto for participation?
Coleslaw wrote: Sun May 16, 2021 11:04 amI may have more suggestions and feedback later. But to be clear, these are my thoughts and not those of the entire [H] team.
Please keep them coming.
User avatar
Coleslaw
Reactions:
Posts: 22
Joined: Sat Dec 26, 2020 8:41 am

Re: Suggestions and Ideas

Post by Coleslaw »

No, I am not expecting you or anyone else to reward anyone with crypto. I'm stating to embrace both sides of the community and not ostracize one over the other.

If testing the server is set points regardless of contribution, I don't see why all the teams would really care outside of a few hosts. We already have teams that milk the system with FB. Why wouldn't they milk the freebie points here?

As far as jumping teams, we already saw people willing to make new accounts to team hop for various reasons. So, they will indeed do it if it is important to them. IIRC, you even did it once or more. But you are right, it would solve the one issue of people loading up millions of points and sitting on them ready to dump between teams. So, I see this as limited the headache a bit.

I hope you do find a quick and simple way for the user to sign up each year. That will make a huge difference. It will also change dynamics for large teams that have a lot of set and forget types out there. Only those wanting to when the challenge would count. Or would they merely need to register the site and then teams could sign them up?

Are the dates going to be fixed or will they be random for the sprints? Or is that still being debated on? 12 is still a lot of challenges unless all they partake in is BOINC-Games. I personally would rather see only say 6 sprints but have them twice as long with 2 projects back to back. That way it is only 6 weeks a year but you get 2 different projects in. AND bunkering the second one could factor in to how much you put into the first one. Right now the majority of people like keeping challenges small like under 2 weeks. But 6 days straight every 2 months isn't so bad. You could even guarantee it not to overlap Pentathlon (May) by doing it which helps with the fatigue issue.

As far as replacing dead projects with new.... I personally like projects to have been in action a while before getting their feet wet with challenges. If it is less than a year old, I don't fully trust it nor the admins to fully comprehend. I would rather to never had an infant project be considered even if proven tough. If half the marathon projects suddenly dry up, I'm fine with the lack of additional points to be had. That rewards those that got in early on and spread their resources out more but doesn't give them an insane amount of benefit for locking in position the whole year.
Skillz
Site Admin
Reactions:
Posts: 1854
Joined: Sun Sep 15, 2019 3:03 pm

Re: Suggestions and Ideas

Post by Skillz »

Coleslaw wrote: Sun May 16, 2021 4:15 pm No, I am not expecting you or anyone else to reward anyone with crypto. I'm stating to embrace both sides of the community and not ostracize one over the other.
Yeah, I am not against crypto. Not sure how it falls into this. If people want to earn crypto while doing BOINC. More power to them.
Coleslaw wrote: Sun May 16, 2021 4:15 pmIf testing the server is set points regardless of contribution, I don't see why all the teams would really care outside of a few hosts. We already have teams that milk the system with FB. Why wouldn't they milk the freebie points here?
Some of us will want to put everything we have on the new project just to see if it'll handle the load. Those who don't, then I couldn't care less. Even if they just throw 1 host on it, that'll be fine. You spoke of advertising/marketing of formula-boinc, well the test sprints will be a good way for new projects to get the word out as well. Although I need to come up with a different name besides "test sprints" as I don't really like the name.
Coleslaw wrote: Sun May 16, 2021 4:15 pm As far as jumping teams, we already saw people willing to make new accounts to team hop for various reasons. So, they will indeed do it if it is important to them. IIRC, you even did it once or more. But you are right, it would solve the one issue of people loading up millions of points and sitting on them ready to dump between teams. So, I see this as limited the headache a bit.
That's the only thing we want to remove. If a member starts a sprint on said team, they need to stay on that team till the end. If they create a new account to join said team. Then that's fine also. They'd have to download all new WUs and crunch them.
Coleslaw wrote: Sun May 16, 2021 4:15 pm I hope you do find a quick and simple way for the user to sign up each year. That will make a huge difference. It will also change dynamics for large teams that have a lot of set and forget types out there. Only those wanting to when the challenge would count. Or would they merely need to register the site and then teams could sign them up?
Having a "set it and forget it" attitude is what most people who run BOINC (or any DC project) do anyway. The only thing they'd have to do outside their norm is register with boincgames and opt into the competition each year.
Coleslaw wrote: Sun May 16, 2021 4:15 pmAre the dates going to be fixed or will they be random for the sprints? Or is that still being debated on? 12 is still a lot of challenges unless all they partake in is BOINC-Games. I personally would rather see only say 6 sprints but have them twice as long with 2 projects back to back. That way it is only 6 weeks a year but you get 2 different projects in. AND bunkering the second one could factor in to how much you put into the first one. Right now the majority of people like keeping challenges small like under 2 weeks. But 6 days straight every 2 months isn't so bad. You could even guarantee it not to overlap Pentathlon (May) by doing it which helps with the fatigue issue.
The dates will be fixed and known at the start of each year. Dates will not be "random" in a sense that a computer isn't going to generate a random date for the project, but the project themselves will be selected at random. They'll be announced when the Sprint begins (or hours before, not a full day) and users who sign up will have the ability to be notified via text message or email (possibly some other instant notification) of what the project will be and the exact time it will start.
Coleslaw wrote: Sun May 16, 2021 4:15 pmAs far as replacing dead projects with new.... I personally like projects to have been in action a while before getting their feet wet with challenges. If it is less than a year old, I don't fully trust it nor the admins to fully comprehend. I would rather to never had an infant project be considered even if proven tough. If half the marathon projects suddenly dry up, I'm fine with the lack of additional points to be had. That rewards those that got in early on and spread their resources out more but doesn't give them an insane amount of benefit for locking in position the whole year.
This will keep things interesting. Adding new projects, which to be honest, probably wouldn't happen all that often anyway. It's not like BOINC is flooded with new projects each year and lots of projects don't really die every year either. You'd be looking at 1 or 2 projects being added in each year, if that.

You spoke of teams like Gridcoin, who have been part of competitions yet their entire user base (or nearly all of it) doesn't even have a clue they're in the competition. Well, this can be said about all teams. Most teams have lots of members who contribute to the team and have no idea the competition even exists. You want marketing? Well, now the teams will have to reach out to their members (or new members) to get them to join so they can be competitive. Then they'll know the competition exists.
User avatar
Coleslaw
Reactions:
Posts: 22
Joined: Sat Dec 26, 2020 8:41 am

Re: Suggestions and Ideas

Post by Coleslaw »

The crypto thing falls into this because it seems to be a major issue in every BOINC competition out there. It is important for there to be an open stance on it from the very beginning.

I agree some will put everything they have into it which also opens their team up to fall behind. If you are ok with it then so be it. But just be prepared for the complaints to roll in from those thinking it is unfair later.

Fluctuation of projects are erratic. We had a few years there where it seemed a lot of "weak" projects were sprouting up every month and we have had cycles were long term projects all ended around the same time as well. Maybe not within the last couple but certainly have happened. Probably just how the grant funding goes. What I have seen is more "personal" projects springing up rather than University backed projects.

I agree that making teams talk about competitions adds free marketing but I think those happen already. Most of the teams either don't have forums or their teammates don't check in. So, I don't think this will change like you imply. You really need to have ambassadors to go around poking the giants. Instigators and protagonists provoking the teams into action (in a positive way) as well as going to the project forums and chatting up the events. (Remember when [H] did this to XS?) Formula-BOINC has already done a lot of damage with their terrible management of things over the years and have already pissed some of the community off. But what doesn't right? I only mention this from my experience from when I took over the reigns of DC-Vault. It was one of the oldest year round challenge sites in the BOINC community. And even still, every time I went to a project to post about adding the project or removing the project from DC-Vault, the response was WTF is DC-Vault? Formula-BOINC only started getting recognition because it was taking projects down. Not because it was fun. Some teams just never see mention of these events because honestly, they are lucky if their teammates check in at all. At [H] we still have a lot of people that set up BOINC but don't even come to the forums anymore. We can send out mass emails but it isn't like people are really responding to those either. I cannot say if it is going to spam folders or what. But as more people at projects read over it, it really does get them asking questions. Once BOINC-Games is in action, I suggest every team participating in its design pick a team to partner up with in the sense of trying to motivate them to come compete. Possibly a team more their size to be a competitor with.
Skillz
Site Admin
Reactions:
Posts: 1854
Joined: Sun Sep 15, 2019 3:03 pm

Re: Suggestions and Ideas

Post by Skillz »

Yes, I remember all too well when we got XS to start competing in FB. That was a fun year. 8-)

If your team members don't want to participate in boincgames, then why should you benefit from their output? This was the age old argument against Gridcoin. If each member has to sign up, then that argument is gone. The point of boincgames is for those who want to compete to compete against others who want to compete. Not compete against giant teams that just have 100s of members whom have no clue they are in a competition. This should level the playing field for all teams to be competitive. Because smaller teams will have a higher chance of getting 100% participation, while big teams may not get everyone on board.

People are gonna complain no matter what. As the old saying goes "ya can't please everyone."
User avatar
Coleslaw
Reactions:
Posts: 22
Joined: Sat Dec 26, 2020 8:41 am

Re: Suggestions and Ideas

Post by Coleslaw »

Right. You can never keep everyone happy. However, then we need to make sure that viewpoint is related from the beginning as well. It is the standpoint of the competition that ......... Then people coming in know what is expected and to leave the bitching behind. I'm fine with not including the sleeping giants. But some will bitch if it isn't made clear that the event doesn't support that view.

The only thing it will really affect [H] on is WCG and quite frankly the bulk of the points we put up during competitions is from active members anyways.

Have you guys made a list of teams to also reach out to for helping design the events?
Skillz
Site Admin
Reactions:
Posts: 1854
Joined: Sun Sep 15, 2019 3:03 pm

Re: Suggestions and Ideas

Post by Skillz »

There is nothing that says we don't support anything. I don't want to over complicate things with so many rules, statements, etc... If you want your team to compete, then your members will need to sign up individually. Your team founder will need to sign up as well to create your team and enter them into the competition.

The only thing I've done so far is hire a firm to build the site. Which is in progress right now. They told me ~2 months before they'll have a demo for me and it's been a month give or take a couple weeks. So I'm expecting a live demo in about a month to test out. I am sure it'll need to be revised a lot before it's the way I want it.

As for reaching out to teams, no. I have not reached out to any teams other than TAAT. I know it was posted on the FB web site and [H], but that's the only two places I know of. When the site is at a near complete state I will reach out to a few select teams ([H] included) to do a trial run on the setup. Iron out any bugs, or whatever might come up. My goal is to have the site live and function for the 2022 calendar year. So that means everything needs to be ready and working before December 2020.

At that point I will ask for help in reaching out to all the BOINC teams to get the word out about it.
User avatar
Coleslaw
Reactions:
Posts: 22
Joined: Sat Dec 26, 2020 8:41 am

Re: Suggestions and Ideas

Post by Coleslaw »

So that means everything needs to be ready and working before December 2020.
Gotta love typos...lol. Sounds like you have it pretty well decided on then. Just let us know when to break things and we will take a look. :D
Skillz
Site Admin
Reactions:
Posts: 1854
Joined: Sun Sep 15, 2019 3:03 pm

Re: Suggestions and Ideas

Post by Skillz »

HAha yeah, I meant 2021. :oops:
Post Reply