Regarding:

«I do not believe that a cost-benefit analysis makes sense in this context. It is not plausible to quantify or estimate the potential future socio-scientific benefits from the projects we work on. The data generated from our simulations here is stored and has the potential to be used long after the current research teams who manage these projects have moved on.»

I believe you are missing a point: let's simplify and take for granted that the results will never be lost. The results will most likely provide huge benefits, that's for sure. Let's now consider two alternative realities:

1. We crunch now
2. We crunch later (let's say, in a couple of decades or otherwise when the climate emergency is better handled)

In both cases we will most likely have huge long-term benefits. To evaluate what is the best decision, you must estimate the marginal benefits and costs of one decision versus the other. The decision to crunch now versus later results not in those most likely huge benefits of crunching per se, but in the difference between the benefits of crunching now versus later. Let's call this difference: "marginal benefits". Then, we also have the "marginal costs" (the difference between the costs of running today vs later). Everyone that is crunching WCGrid today, if acting rationally, must be thinking that the marginal benefits are greater than the marginal costs. The marginal benefits can be more or less significant, but the same goes to the marginal costs: the money we use for electricity today (which could be used for a number of good things, including fighting Climate Change or other types of injustices), as well as the pollution we create now while running WCGrid - on top our Climate Crisis - has a direct negative impact on all of us. While there could be cascading marginal benefits, there are *for sure* cascading marginal costs. To me, it is nowhere near clear that the marginal benefits are greater than the marginal costs. As I've suggested earlier, if it is difficult to estimate the long-term marginal benefits of crunching now vs later, it is much easier to estimate the short-medium term marginal costs of crunching now. If we estimated these costs, maybe we could eventually be able to reach conclusions such as:

1. Crunching WCGrid costs X lives (or to be less dramatic, using our efforts in another program - the very best alternative program(s) - saves X lives); knowing that X is <far lower> than the number of the estimated people that could benefit from - say - cancer treatment or cure (or whatever the target of each WCGrid project) earlier, we can say with a great amount of confidence that we <are in a good track if we keep crunching on WCGrid>.

OR

2. Crunching WCGrid costs Y lives (or to be less dramatic, using our efforts in another program - the very best alternative program(s) - saves Y lives); knowing that Y is <far greater> than the number of the estimated people that could benefit from - say - cancer treatment or cure (or whatever the target of each WCGrid project) earlier, we can say with a reasonable amount of confidence that we should <use our efforts elsewhere, at least until the conditions change>.

As I wrote before, I do not believe that a cost-benefit analysis is appropriate nor practicable. It is not possible to quantify the marginal benefits because of a great degree of uncertainty in what will be produced. As far as people acting "rationally", you are assuming that people are conducting some complicated cost-benefit analysis when, 1) behavioral economics would suggest that people do not typically behave that way and 2) how would you ever quantify the pleasure people derive from being a part of a community of crunchers/the joy from participating in something bigger than themselves (i.e. the sociality aspect), or from WCG participation being an educational device?

I think a more constructive use of time would be to stop eating meat instead of worrying about how many lives are lost due to WCG's infinitesimally small (if even empirically possible to consider) contribution to global warming. Again, it is not possible to quantify/estimate how many lives will be "saved" from the research we conduct here, because scientific investigation is rarely so cut-and-dry, action-impact, intervention-amelioration oriented. At least for me, this is about contributing to the corpus of scientific knowledge in a way that is dynamic: a handful of simulation results from my computer join other results who can be interpreted later and perhaps yield new simulations that, perhaps ad infinitum, eventually lead to "lives saved". Therefore, a cost-benefit analysis is simply not practical.

wcgridmember, I agree with you, that the environmental impact and cost of WCG should be minimized. To do that, it must also be estimated. What I do not agree with, is that the WCG staff should be blamed for not doing this. WCG is a platform, were everyone can participate with any hardware one likes. WCG has no control over this. So every single contributor is responsible to crunch with a minimal environmental and cost impact. And you know? There are repeated disussions about exactly this topic on this forum and elsewhere. Most serious participants I know really care about crunching efficiently and put considerable effort into doing this. So when do you start to care, think about how you can participate most effectively, and stop to blame other people?

I don't think you're going to get anywhere by posting on the forums.

There's been disagreement back and forth, so let's face it that neither side will back down.

Instead, what you should do is get an official response from WCG; https://www.worldcommunitygrid.org/viewContactUs.do

Hi again:

I understand that you would like to better understand the costs, which totally makes sense. That said, we cannot call it a cost-benefit analysis, then, only a "cost analysis." For those who want to track how much electricity they are consuming through WCG participation, electricity use monitors like "Kill a Watt" can be purchased on Amazon for roughly $20.

I do think it is incorrect to categorize WCG as being unwilling to estimate the costs, as you describe. Their rationale, from my understanding, is that users' spare computing time already exists and this is a noble way to utilize it. Measuring "marginal costs" is difficult for WCG to do on a systems-level because, for one, it is not possible for them to know how people would be using their computer otherwise. I totally support you, and anyone else conducting grid computing, to measure their wattage usage differential: i.e., wattage used while crunching WCG minus idle computing wattage, with some kind of factor for additional machine wear based on that unit of increased wattage, which is extremely hard to estimate. I just don't think it is necessary. Whatever "marginal cost" created by my small tablet crunching 8 threads at a time is facially small compared to, for instance, the similar wattage my cat's water fountain consumes—and that fountain hasn't contributed to scientific knowledge nor facilitated my own interest in this research approach and the projects using it.