Enhancement Converter "Adjustments"

[Zepp]

I've noticed that the average number of converters I need for targeted conversions at level 50 has gone from around 25 to around 45. Is this simply the RNG going haywire, or has there been an adjustment to the Enhancement Converter probabilities that has messed things up?

 

Anyone else experiencing similar shifts?

[MTeague]

I suspect it's just RNG is R. 

But neither have I done a serious mathematical analysis.

[Yomo Kimyata]

I have not noticed any difference since the patch.  What are you converting?  It sounds like a combination of out of set and in set?

[KauaiJim]

The RNG in CoH is very much like this.  It really could just be the RNG.  Here's one post where we were talking about it:

 

Speaking of purples... - General Discussion - Homecoming (homecomingservers.com)

 

 

[Sakai]

Very much random. Sometimes you get lucky and sometimes you don't. And sometimes you REALLY DON'T GET LUCKY 😞

[Zepp]

1 hour ago, Yomo Kimyata said:

I have not noticed any difference since the patch.  What are you converting?  It sounds like a combination of out of set and in set?

It's a combination. I've noticed that Movement, Sniper, and Slow enhancements are coming up more often and Defense, Resistance, and Heal are coming up less often. I've also noticed that the number of in-set conversions to get a proc has appeared to increase. My guess was randomness, just wanted to know if there had been any adjustments to make the RNG less random resulting in weighted results.

[Outrider_01]

RNGesus hates us, have converted one set like melee that turns into another melee and back into the first set.  So same 2 sets over and over, every IO to the point I got like 4 sets of each.  After 50 wasted converts and trying to prove a different result is not coincidence I just destroy it.

[Heraclea]

My stopping point on converter gambling is 'anything that any character might be able to use' and I haven't noticed any issues.  OTOH, rolling for the one missing piece in the set is like drawing to an inside straight, and you should expect long odds. 

[Snarky]

i either get really lucky.  take one random purple and convert it to Hecatomb!  and then one conversion to the exact one I need!  Or....I sit and convert Steadfast Protection back and forth between the two not plus global defense as i watch my profits drain and want to hurl the computer off the desk...

[KauaiJim]

What Snarky said.  When I play the converter game it's usually either strike gold or strike out.

 

Edited to add:   And oh yeah, I have had it bounce back and forth between a couple that I didn't need until my converter supply was depleted.

Edited April 24, 2021 by KauaiJim

[Zepp]

Well, it looks like my luck turned around after posting this. Yesterday I converted 11 to exactly what I want with just under 300 converters. Today I converted 9 in 240. The irony is that those numbers come from my toon who is working on lvl 30 conversions which should be in the 38-42 range. So I went from extremely unlucky to really lucky! I hope this post doesn't change my luck again... oops...

[Yoru-hime]

It turns out that Converter RNG is directly linked to forum post count. Therefore any and every post about Converter luck changes that poster's luck, thereby making them look quite foolish.

 

Our devs are tricksy and ebil. They've gotta be to stay ahead of the marketeers. 😄

[Yomo Kimyata]

37 minutes ago, Yoru-hime said:

It turns out that Converter RNG is directly linked to forum post count. Therefore any and every post about Converter luck changes that poster's luck, thereby making them look quite foolish.

 

Our devs are tricksy and ebil. They've gotta be to stay ahead of the marketeers. 😄

 

oh $!&@.

[Riverdusk]

I think my current worst record is 37 attempts (111 converters used) to convert something in-set before I finally got the one I was going for.  Something that should be a 1 in 5 chance.  I've also often gotten it very first try though.

 

 

[MTeague]

3 hours ago, Riverdusk said:

I think my current worst record is 37 attempts (111 converters used) to convert something in-set before I finally got the one I was going for.  Something that should be a 1 in 5 chance.  I've also often gotten it very first try though.

 

 

I'm picturing you sitting there fuming in Vendetta Mode, grumbling in Yosemite Sam voice "I'll get that blasted enhancement converted if it's the Last Thing I Ever Do."  Knowing your long past the point of doing it even remotely economically, tossing influence out the window to buy more converters, and just being in a contest of wills with the game to finally produce the stupid piece that you wanted lol.

 

Lord knows I've done exactly that more than a few times.

[Yomo Kimyata]

On 4/24/2021 at 3:51 PM, Zepp said:

It's a combination. I've noticed that Movement, Sniper, and Slow enhancements are coming up more often and Defense, Resistance, and Heal are coming up less often. I've also noticed that the number of in-set conversions to get a proc has appeared to increase. My guess was randomness, just wanted to know if there had been any adjustments to make the RNG less random resulting in weighted results.

 

Thanks in particular to the knowledge that @UberGuy shared in the markets forum, it would be difficult to easily change the system away from equally weighted random system we've got now.  I've done a few case study experiments on the RNG, and there is no reason for me to suspect that any vagaries in the RNG have been programmed in.

 

So far, my personal record for converting in-set to a specific IO is 56 rolls, or 168 converters.  I think my next closest is 42.

[Coyotedancer]

It all comes down to the RNG being a serious jerk sometimes. 😝

[Zepp]

My worst for in-set was 30, damn, I don't feel so bad anymore.

The set procs do seem to show up less than the others. In a six piece set I seem to have a 25% chance for the first enhancement, a 20% chance for the second or third, a 15% chance for the fourth or fifth, and a 5% chance to get the sixth. [Numbers are approximate based on 319 conversions].

[KauaiJim]

Well I'm not sure if I'd be bragging or exposing myself as a stubborn idiot if I tell y'all my worst.  Let's just say it was a wee bit more...

[Hew]

I have done several single inset converts _in a row_ with the standard 3 converters. 45 converters is nothing. That is 15 convert attempts. You do not get better than a 1 in 5 each roll no matter what, and statistics dictates that over time that does not change from 1 in 5. This is a common trap people subconsciously fall into, especially with lottery tickets (which is a massive industry capitalizing on this fallacy). I bought 5 tickets today, after buying 5 tickets yesterday, so my odds should be better (they are not, your odds still suck).

 

 

6 hours ago, Zepp said:

The set procs do seem to show up less than the others. In a six piece set I seem to have a 25% chance for the first enhancement, a 20% chance for the second or third, a 15% chance for the fourth or fifth, and a 5% chance to get the sixth. [Numbers are approximate based on 319 conversions].

This is misleading.

 

When attempting to get 6 pieces of a set, you have 100% chance to get your second io in the set. There is zero weighting being done.

 

On 4/24/2021 at 5:48 PM, Snarky said:

i either get really lucky.  take one random purple and convert it to Hecatomb!  and then one conversion to the exact one I need!  Or....I sit and convert Steadfast Protection back and forth between the two not plus global defense as i watch my profits drain and want to hurl the computer off the desk...

You arent lucky. This is highly common.

 

On 4/24/2021 at 4:54 PM, Heraclea said:

My stopping point on converter gambling is 'anything that any character might be able to use' and I haven't noticed any issues.  OTOH, rolling for the one missing piece in the set is like drawing to an inside straight, and you should expect long odds. 

This is the closest to reality.

 

Blowing a ton of converters to get a single piece in a set is HORRIBLE waste of resources. Recycle that into inf, use that inf to BUY the one you want, bypass rngeesus. You will make A LOT more inf, get a LOT more of what you want, by strategically converting, instead of trying to convert in set for a single piece and hoping you dont blow a ton of merits. With about 125 merits and 1 mil starting (fees for auction), I can easily kit out a full level 50 in a week or two of background market/convert putzing. No additional merits! 

[tidge]

I'll just add this bit of personal practice: There are certain sets where I will NEVER try to convert from uncommon to rare because the odds of getting another common are practically guaranteed (at a particular level). Those get a random roll on the uncommon table first. Hello Snipes.

[Yomo Kimyata]

43 minutes ago, Hew said:

 

 

Blowing a ton of converters to get a single piece in a set is HORRIBLE waste of resources. Recycle that into inf, use that inf to BUY the one you want, bypass rngeesus. 

 

Sometimes it's a great use of resources.  I've given this example in my universally acclaimed thread 

 

 

Let's say you need an Eradication proc, but they are trading at ! 5mm !  There are five other members of the set, and they are each trading at 2mm.  Converters are trading at 66,666.

 

Starting with any non-proc, you have a 1 in 5 chance of getting what you want in one roll or 20%.  You have a 4 in 5 chance of not getting what you want in the first roll, and a 1 in 5 chance of getting what you want in the second roll, or 36%.  After 3 rolls: 49%; 5 rolls: 67%; 10 rolls: 89%.  The average number it will take will be 5 rolls (this is a geometric distribution, so it has an (infinitely) long tail).  

 

If you are converting one, it may take one conversion, but it may take a billion.  If you are converting ten, it will probably take around fifty conversions.

 

Five rolls costs 15 converters which costs 1mm inf.  My shortcut rule, taking into account the bid-offer spread, is that if one member of a 6-element set costs over 2mm less than I can easily sell another member, and I can do it in bulk, I will do that trade all day and night.  In fact, I do.

Edited April 26, 2021 by Yomo Kimyata
I like editing.

[Heraclea]

32 minutes ago, Hew said:

This is the closest to reality.

 

Blowing a ton of converters to get a single piece in a set is HORRIBLE waste of resources. Recycle that into inf, use that inf to BUY the one you want, bypass rngeesus. You will make A LOT more inf, get a LOT more of what you want, by strategically converting, instead of trying to convert in set for a single piece and hoping you dont blow a ton of merits. With about 125 merits and 1 mil starting (fees for auction), I can easily kit out a full level 50 in a week or two of background market/convert putzing. No additional merits! 

 

True to an extent.  OTOH, reward merits that aren't a multiple of 20 or 50 are loose change.  For those I feel quite free turning into converters at the merit vendor and trying my luck. 

[Hew]

31 minutes ago, Yomo Kimyata said:

 

Let's say you need an Eradication proc, but they are trading at ! 5mm !  There are five other members of the set, and they are each trading at 2mm.  Converters are trading at 66,666.

 

Two issues. #1, you have inf to buy converters with. #2, in your example you are converting through purples which have a vastly higher margin than a cheap eradication that with long-term bids you can get a fraction of the Buy It Nao price.

 

I can understand some people like the Buy/Convert It Nao process, but it burns resources far faster than "inefficient" converting, long-term bids, sniping, and a lot of other mechanisms the market has to get you what you want at a discount, sometimes enormously huge discount.

[Zepp]

9 hours ago, Zepp said:

My worst for in-set was 30, damn, I don't feel so bad anymore.

The set procs do seem to show up less than the others. In a six piece set I seem to have a 25% chance for the first enhancement, a 20% chance for the second or third, a 15% chance for the fourth or fifth, and a 5% chance to get the sixth. [Numbers are approximate based on 319 conversions].

 

2 hours ago, Hew said:

This is misleading.

 

I know that it is difficult to perceive probabilities as such and the human predilection towards certain fallacies. I was just asking based on numbers that suggested that weighting may have been occurring in regards to the actual results I am seeing not having the type of relatively flat distribution one would expect. I was obviously half joking with the numbers (319 is obviously a Prince reference).

In pre-in-set-conversions I get LotG + Global Recharge as the initial LotG in around 1-3 times out of 20. I know this falls within a standard deviation or two, but I can't remember the last time I had 4/20. or better. Because I do not record every singly conversion I do (that would be in the thousands, which would be larger than my headache database for the past five years), I just wanted to throw out the question, and let people joke around with it.