One hundred trials later...

There is (effectively) a streak-breaker. It's called Empyreal merits. Functionally it's the same thing.

And that "streak breaker" is used for other things, hence the objection to the random tables. Effectively, it is a double punishment for the "unlucky".

Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters

This is the real sticking point. Back when there wasn't anything to do with the merits, it didn't matter as much. Get unlucky, oh well you'll get it eventually with merits.

Now, though, the person who gets lucky not only gets their VR quickly, but then is free to spend their merits to unlock all those goodies. Meanwhile, the unlucky soul who's on his 60th trial is feverishly saving 30 emps to blow on a VR instead of costume parts.

~60 trials in with this defender, 1 VR. Far as I can tell, at 10% the odds of having less than 2 VRs after 60 trials are about 1.3%. I suppose there always has to be the one really unlucky person, but why does it have to be me?

Exactly. This is what I was referring to when I said that an early good roll is worth lots more than a late one. If the numbers are correct, buying a VR with Empyraens is a losing proposition that applies only to people who fail to roll correctly. Buying a single VR costs you the same amount in Empyraens as 1/2 of the most expensive PVP IO on the market. There is no way to interpret it other than taking a huge loss due to a random mechanic. This is, to the letter, exactly why random drops at the end of TFs needed to be converted to merit rewards, that you could choose to gamble with by spending the merits. Right now if you get a VR drop, you not only get the value of the VR, you get to keep the Empyraens.

I think you're continuing to miss the point of what the exact benefit of randomness in a reward system actually is. By definition a random result will sometimes generate a higher than average result, and sometimes a lower than average result. If you are going to automatically interpret every result that is lower than average as "taking a huge loss" and that this is always an undesirable result for a reward system to generate you're just axiomatically opposed to random systems.

The other way to interpret this situation is that the chance of acquiring less than average rewards is the price people pay to have a chance of acquiring greateer than average rewards, and even under the worst case scenario the deterministic merit rewards place an absolute floor on reward earning, which provides a safety net against the worst possible outcomes the random system can generate.

It's not "misunderstanding" to say that I don't agree with your interpretation. I can see the numbers myself. I think you are gifted with numbers but also filled with hubris if you think you are the only person with the ability to evaluate them and hold opinions about what they mean.

[PS I'm concerned that you add that chance is the "price people have to pay," but ignore the issue that it is the price people pay that is the issue many people have with the system. It's like we zipped right past each other on logic trains going two different directions.]

[PPS it is only now that I understand how you are reinterpreting my statement about "taking a huge loss" as the equivalent of getting low rolls. This is an incorrect characterization. The huge loss is for a character who has to spend Empyraens, because Empyraens buy more than just salvage. IF the odds are 10%, and it takes 30 trials to get enough Empyraens, it is undeniable that the end chance of not getting a Very Rare is only 4%. That means about 1 in 22 players will have to spend them. You say that is a good thing, the "price people pay." I say it is not. If the chance is some other number than 10%, the same applies, but to more or fewer people. This is a question of equitability, and it is not easily settled.]

The issue is that the safety net is full of holes.

Both the "lucky" and "unlucky" get the same safety net. After 8 Empyrean Merits, the "unlucky" person has a choice between more grinding, a costume part, or a needed rare component. After 30 Empyrean merits, the "unlucky" person has a choice between grinding more trials, saving more for special recipes, or a needed very rare component. The "lucky" person doesn't need to choose as they already have the rare or very rare component.

The "safety net" is also outrageous compared with the value of other things the Empyrean Merits can buy.

Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters

But that is *still* suggesting that the main problem with random rewards is different people get different rewards. Of course the "safety net" benefits everyone equally: if it specifically benefited the people who benefited from random rolls less to equalize their rewards then the randomness itself would be a sham.

Axiomatically the preference for random rewards is the preference to risk getting less in order to potentially get more. This risk only actually exists if people actually sometimes get less. Every single one of those people will have no actual option to eliminate the effects of that risk because if they can the risk doesn't exist.

Its actually a complete waste of time to try to prove somehow that with random rewards some people will get a lot less than others: that's an easy thing to stipulate: in a random system some people will *have to* get more and some will *have to* get less. If that specifically doesn't happen, that would actually place the randomness source itself or the system in total under suspicion for being bugged.


That's a completely separate subject, but while I have to keep mentioning (and it is often skipped or overlooked) that the actual numbers are subject to negotiation (as they usually are) the principle that the merits should be able to buy something else is actually not contradictory to the system. If they had nothing else to buy, then eventually those merits would be nearly valueless to the people that got higher drops through luck. They would then, in effect, be neutralizing (partially) the advantage of being lucky. Or to put it your way, the safety net would not be benefiting everyone equally. That's counter to the normal intent of having a risk-based reward system.

You seem to be thinking the two systems should be judged independently, as if people got to select one or the other. They are not intended to work that way. The way they are intended to work is that every trial delivers some reward X. X is separated into a fixed (depending on trial activity) reward of merits, and a lottery ticket for the component drop. We've debated the relative value of those two rewards previously, but they are not intended to be replacements for each other. You're supposed to get both, and the idea is rather than have a reward X that varies from a very low value to a very high value, X varies between a moderate value and a high value due to the merits increasing the floor of the worst case reward to a much higher level than the component drop alone.


But to loop back around and say that the players getting lucky component drops will always be ahead of the ones that don't, ergo the system is broken, is just another restatement of the same preference that you don't want random rewards. The preference that you don't want random rewards is basically tantamount to the statement that you want everyone to have no more, and no less rewards than anyone else. Its the same objection, not a deeper or more obvious one.

It is if you assert its the only interpretation, or more generally the only reasonable one. You should be careful slinging that hubris brush when you make statements like that, since I'm not the one stating that there is only one reasonable point of view to the question of whether random reward systems are fair.

I'm not talking about "random systems" in abstract. I'm talking about this one. Random systems are not all alike. There was an attempt by some posters earlier to link dislike of random drops at the end of trials to random rolls during combat, but the two have nothing in common and the comparison is meaningless. Same thing if we expand to talk about reward systems with a random component. Purple IOs are random rewards, but a different kind, and their mechanics are only tangentially related to whether random rewards should be the primary determination of advancement speed through this particular leg of this particular game.

What IS undeniable--or I should think it is--is that a "deterministic" system where there is a 96% chance that the random reward kicks in prior to the termininating element that makes the system "deterministic" is actually more random than it is deterministic. Again that is IF the chances for a V Rare drop are really 10%. Basically the odds are so high that "random" is almost "deterministic" in itself for all but a handful of players. You may think that is equitable, but I disagree.

[EDIT: Several edits for clarity. Sorry for confusion if you started a reply.]