TF/SF Merit rewards: An observation.

Actually, it was perfectly within the context of his post. reread it.

He has been contradicting his previous statements.

But not of the thread and my previous and after posts. You just took a question I posed to the thread as some type of proof. If that's the best you can do, then you need lots of practice.

I've already disclaimed that I run speed TFs/SFs along with normal instances. I even proposed a solution that would keep merits the SAME for speed runners while providing bonuses to teams that did not avoid content. Keep up.

Yes but you are not accounting for the other groups of players Pum, it's not just S and N.

Lets take the ITF for example if i may(based on my personal experience and observations):

You can run that in what under 25 min?

Group on freedom that runs TF's all day do it in 30-40min

Pug speedish ITF on both Justice and Freedom 35min-1hr

Pug normal ITF or Masters runs 1hr 30-2 hours

lowbie ITF run for xp 1hr-4 hours

You have a few groups running TF's

Player S like you stated

Player N also from your statement

Player F (speedy but not so fast as S but still multiple TF's done in a night)

Payer X (slow like N but again all they do is TF's)

Plus I am sure there are more groups. Some people split TF's into several days even.

So depending on the distribution of these groups across the different server populations is what determines the avg. Now for S's to significantly scew the average there would have to be ALOT of them. I have not had the opertunity to run with any tf's on virtue yet so I can only speek from experience on freedom and Justice which are usually well populated. Based on that the "S" group is likely within the minority. I truely believe that the avg time is heavily based on slower speed runs and kill most runs.

Again the fact that the avg time (merit wise) for and ITF across servers is 90 min and one of the faster speed times is 20 min (a vast difference) sugests "S" playes are outnumbered by a combination of other players enough to keep that avg time much higher than the speed times.

I realize you where pointing out other sugestions on how to award merits my post was directed to the "omg people who speed through TF's are ruining my rewards" posts. Again I believe it is more likely that this is a true average and ploted data would probably be a bell curve with your speed and long tf's at the ends, which while not fair to all is probably fair to the vast majority of the popuation.

I've pointed out that the ITF is probably a well balanced TF/SF, since there aren't many points that are skippable, yet the the variation in times is still quite significant. As ITF is probably the most popular TF/SF, I would guess that the N types figure more heavily than S type players by sheer numbers.

However, not every TF is as popular, and the S type players get represented more thoroughly through those TFs. RSF, KHTF and Eden are extreme examples of this.

Note that even figuring the difficulty level of the RSF, it is only awarding 25 merits to STF's 37ish merits. Eden's skimpy reward also shows that the S type players are dominating the times to disproportionately represent themselves in the overall population. The takeaway is not just that the S players are doing it faster, they are getting "counted" more than once because they're finished sooner and then they pile another TF on top of that.

That multiple participation effect by faster players means that the slower players and casual players are underrepresented in the sample, which means their efforts are counted less in calculating the TF average/median time. It suggests they are getting a double penalization of running fewer TFs (deliberately, by inexperience or by lifestyle) and then getting fewer rewards when they do finish the TF. I've suggested a (slight) bonus in a previous post to work around this issue.

I used Player N and Player S to illustrate the extremes. There are many players who fall between both groups. I don't think that it matters which server the populations reside on, from the developer's point of view.

The abilities of the population are likely to be distributed on a standard curve, but the data set gets skewed because the faster folks get to be counted more than once. Just remember that what the devs have chosen to measure isn't like the height of a person, which gets counted once. They are measuring time to completion, and then the participant can choose to run it again, and get counted again.

Isn't that exactly the point? The speed runners have defined a reward level for KHTF and Eden where it is no longer desirable for the speed runners, and it is cold comfort for the the non-speed runner to scrape through Eden in 90-180 minutes only to find that their merit reward won't even sustain a random roll.

The devs can't significantly bump up the the low rewards (although over time, the TFs may increase in average time if no speed runners work those TFs) or the rewards will be out of line for the task. They need to rework the TF a bit to narrow the gap and/or bump up rewards for non-speed runners in a way that doesn't reward them for purposely going slow.

Sure you can. The value you place on it is debateable, but it is always greater than 0, which by default will favor a longer duration taskforce as it will drive the value closer to the 1 merit per 3 min value if we assume the task is being completed in the dev assigned time.

As the number distances itself from 0 it benefits longer duration tasks more and more. Even if we state that vill and hero tasks are formed in the same time it automatically puts villains at a disadvantage because their tasks are on average shorter.

Shivians are optional to assist in driving your time below the dev assigned value (ideally anyway, some tasks may be skewed by their use previous to merit assignment). However, the sunk cost of team formation is not optional. It will invariably occur.

That said, if you use a shivian on a task then your total time spent on the task does increase because the shivian is expended and took time to get. Interestingly enough, if player A uses a shivian that was already in his possession and did not add to the teams formation time it would negatively impact Player A's total time spent on the task, while potentially lowering the time spent of all the other players.
The time spent to form the team that ran the mission prior to the TF would then be the task formation time. Until there is auto team assignment put in place (not saying I want it) team formation will always have a value greater than 0*

*An exception goes to planned Tasks that are organized outside of the game. Such as when they are planned on the forums. However, the value is still likely to be greater than 0 unless everyone involved logged in at the exact same moment.

The impact the sunk cost has on actual merits/min to time spent on the Task can appear pretty minor (as shown in my last post), but over time it can really add up.

No those are already accounted for in the formula because it is primarily based on median completion times.

Challenge is already accounted for. See Arcanaville's description on minimum thresholds for a greater understanding. While it may not be the most comprehensive system that could be employed it does account for challenge.
I seriously doubt the devs would implement anything like Dr Q these days. I'm sure their datamining shows that they are so infrequently played that removing them would have no statistical significance on the game. If new ones are introduced they will all be sub ~2 hr completions.

I can understand why you might have interpreted what I was saying as a request for longer duration SF's, but I actually wasn't. Quite the opposite actually. I want them to slice and dice the super long hero tasks like they are doing with Posi. Many of them need serious revision to bring them up to acceptable content standards. I'd also like to see 3-5 more SF's added in the 25+ game.

For clarification: I'm not saying that formation time should be factored into the dev formula for merit reward, just that end users should be aware of it when calculating things like earned merits/time spent.

No, it seems I missed you asking that originally. Their datamining for merit rewards post-I13 probably still includes all those pre-I13 speed runs, which is effectively keeping the merit count down, but there are a few problems with throwing the pre-I13 runs out and starting fresh:

1. The merit rewards for particular content are low (Eden, KHTF, Tarikoss, as examples) so there's very little incentive to do them "normally." Even after I13, I've seen very few teams do "kill-most" runs of any of those three pieces of content. This leads to number two:

2. Most teams post-I13 are still completing them relatively quickly. Even if that weren't the case, and merits were upped to the point where they'd actually be worth the time to run again, people would go back to running those pieces of content en masse, which is what merits were designed to prevent in the first place.

Yeah, you're probably right about the respec trials there - at the time I put the spreadsheet together I was simply thinking of the content which gave the most merits, and then figured out average times we were running it in. Never really crossed my mind that dropping the RSF and doing a respec instead would end up coming out better, which is odd especially considering almost everyone I know takes the Hami instead of merits on the RSF.

EDIT: Updated the spreadsheet based on Flea's feedback, and unless my math is wrong, that works out to about 1/3 merits more per hour for heroes, which means it would take sixty hours of running content for the heroes to get 20 merits (or one random roll) ahead of the villains. Just remember that the tab the spreadsheet opens to is the original numbers from I13, which aren't correct anymore - the latest merit numbers are on the third tab of the spreadsheet.

@macskull, @Not Mac | XBL: macskull | Steam: macskull | Skype: macskull
"One day we all may see each other elsewhere. In Tyria, in Azeroth. We may pass each other and never know it. And that's sad. But if nothing else, we'll still have Rhode Island."

I agree, but not to the same degree as the reward suffers from (whatever the proper term is that you don't get as large a reward). It may be worth the speeders time to run it once a day but (and here I am making assumptions based upon nothing tangible) it would also encourage more non-speed runs as the reward is no longer 'anaemic' as someone posted earlier. It is to be hoped that this would create a slightly better balance between the runs which are completed in a scarily fast time and those which are more sedate.

The speeders would probably still be disproportionally represented, just not AS disproportionally represented. No?

Thanks for the clarification. This makes much more sense to me.

I'm honestly not sure, as I've only run the KHTF three or four times, the Eden trial maybe two times, and Tarikoss zero times, since I13 dropped. Still, it might be an interesting experiment, and maybe the devs could actually provide such data for us.

@macskull, @Not Mac | XBL: macskull | Steam: macskull | Skype: macskull
"One day we all may see each other elsewhere. In Tyria, in Azeroth. We may pass each other and never know it. And that's sad. But if nothing else, we'll still have Rhode Island."