Your 1 per 5 games would set the percentage at 2%, and that's well over one million players.
Reducing the percentage to your new 1% of the playerbase in now just well over five hundred thousand players.
We can round down and just call it 500k for the sake of this exercise.
Now, a full time job is considered to be 40 hours of work round here. That's 5 days of 8 hours. This does exclude lunch time, but paid breaks are the norm, and bathroom breaks are required by law to be paid. But there's also vacation days, and federal holidays. Let's exclude those for shits and giggles.
So, we have around 7 hours a day (35/week/person) of a mind-numbingly boring and repetitive task of monitoring the quality of decisions and checking for abuse of power.
How often do we want someone to give a once over for all these mid-game mutings and enhanced reports? 6 times a year? Maybe. They'd have a lot of opportunities to abuse their power in that time span. And you could go over ALL of their decisions, but that would take forever, so you take a random sampling of the times they've exercised their authority and hope that the threat of being caught abusing the power keeps everyone in check. Even a random sampling will give us too big of a number. Let's cut it down a bit.
So, 6 review times a year for 500k people is 3m check-ins per year. Let's have the supervisors pull up just 1 (one) replay of a game where the PMods gave an enhanced report for feeding/trolling and just 1 (one) game's full chatlog with timestamps for when they applied a mute.
That will be ....25 minutes average gametime viewed, and 3 minutes reading the chatlog and noting when/where/why the mute was applied. Now, if the enhanced report of trolling is correct and the mute was applied at what is considered the perfect time, you still have to do a bit of paperwork. Let's say it takes just 6 minutes to log. And 1 minute to blink your eyes and load up the next case.
So that's now 35 minutes x 3 million check-ins per year. That's 105 million minutes or 1,750,000 hours. At 35 hr/week and 52 weeks a year (also rounded down) that is 961 full time employees.
Federal minimum wage is $7.25/hour, and the highest State minimum wage is $11.50 in Washington State. (California is $10.5, so that's towards the upper end, if they go near headquarters) That's ~$12.6 million to ~$20 million per year. But, many states require health insurance for full time employees. There's also "the employer's contribution to social security" and the hardware costs to give them enough computers to do the job and desks and chairs and electricity and heating/cooling and plumbing and parking space, and the bandwidth for all the data they're sending/receiving, and code monkeys to design and maintain all the software, and you'll need some servers to hold the data.
An MIT article has an estimate of 2.7 times the base salary generally covers the costs associated with having an employee. So, we can just sorta crunch that together as $34 million to $54 million a year in recurrent costs.
The office space required for 961 employees is around 150 sq ft per person, that's 144k sq ft. If they put that near West LA, where Riot Headquarters is, Avison Young puts the cost of office space in West Los Angelas as $3.63 per sq ft. per month. That's only $6.2 million per year! Practically free!
So, we're at ~$60 million per year now. That doesn't cover the startup costs. But it's getting late, and I'm running out of steam.
Now, what did we gain from this expense?
You set the % at 1, so one game out of 10 on average will have a PMod in them. The games without a PMod would probably as a side effect be slightly worse than now.
These PMods can forcibly mute others, but If someone is bothering a person, it's well within everyone's capabilities to mute them individually. So we gained absolutely nothing on this front.
These PMods can also create enhanced trolling reports with more weight. You seem to want these reports to lead directly to punishment. You said that you want an employee to review every single one, but the 3 million check-ins per year from the hypothetical situation above costs $60 million dollars per year. That means that every report that is checked by an employee costs $20. There is no way to give 500k people the ability to make Riot pay $20 every time they think someone is trolling.
That is if every single report is validated as true, without a second thought about edge cases, or correcting an invalid report. So that is completely unreasonable.
So, instead, we have these PMods going around giving out punishments, with such minor oversight that it's laughable. That would be abused. The time and money spent investigating and correcting that abuse would be more.
This suggestion is unrealistic. It provides no benefits. It would be horrible.
