Under Level Collection

[CryHaddock]

8 hours ago, theSLAYER said:

Sorry for the bump, but I needed to ask you about the previous method of trading for low level legend (or any raid boss, really).

Can you recall when was the last time it worked for you, and you got any screenshots of date + ball on the mon?
(just wanted to figure out when they changed the mechanic >< )

I believe they didn't change it and we had some wrong assumptions at the time.
(This was when trading was a new mechanic and we didn't actually test it because it would have meant investing a month into present and stardust grinding).

We were able to confirm that wild Pokemon and egg Pokemon are capped at your current trainer level (disregarding weather boost, which is a crude +5 levels applied after all other factors).

And we'd also observed that high level Pokemon had their levels reduced if you traded them to low level accounts.  So we assumed that the level they were reduced to would also correlate with your trainer level.
What I now think is that they're only rounded down as far as level 20.

If the account receiving a Pokemon is anywhere between levels 10-20:
 Level 21-40 Pokemon = rounded down to level 20
Level 1-20 Pokemon = level unchanged

Meaning that there's no interesting way to exploit this because everything can already be obtained at level 20.

I never figured out how trades behave if the receiving account is between level 21-29. I assume they just don't reduce levels at all at that point but I've not tested this.

[theSLAYER]

9 hours ago, CryHaddock said:

I believe they didn't change it and we had some wrong assumptions at the time.
(This was when trading was a new mechanic and we didn't actually test it because it would have meant investing a month into present and stardust grinding).

We were able to confirm that wild Pokemon and egg Pokemon are capped at your current trainer level (disregarding weather boost, which is a crude +5 levels applied after all other factors).

And we'd also observed that high level Pokemon had their levels reduced if you traded them to low level accounts.  So we assumed that the level they were reduced to would also correlate with your trainer level.
What I now think is that they're only rounded down as far as level 20.

If the account receiving a Pokemon is anywhere between levels 10-20:
 Level 21-40 Pokemon = rounded down to level 20
Level 1-20 Pokemon = level unchanged

Meaning that there's no interesting way to exploit this because everything can already be obtained at level 20.

I never figured out how trades behave if the receiving account is between level 21-29. I assume they just don't reduce levels at all at that point but I've not tested this.

I'm presently inclined to believe that as well: someone proposed that the level follows the wild encounter cap (trainer level + 2), and everyone else just took it as fact without checking for themselves.
[Most people don't have low level accounts anyway]

Do you happen to have seen any old video or photos that may help to prove this point? Just for verification, so that we can confirm whether to add downlevelling to PKHeX

[CryHaddock]

On 12/24/2020 at 12:51 AM, theSLAYER said:

I'm presently inclined to believe that as well: someone proposed that the level follows the wild encounter cap (trainer level + 2), and everyone else just took it as fact without checking for themselves.
[Most people don't have low level accounts anyway]

Do you happen to have seen any old video or photos that may help to prove this point? Just for verification, so that we can confirm whether to add downlevelling to PKHeX

I don't know any videos sorry, all my information on this was basically from me messing around and experimenting with one friend

[theSLAYER]

28 minutes ago, CryHaddock said:

I don't know any videos sorry, all my information on this was basically from me messing around and experimenting with one friend

But the experimentation you're referring to is post dec 2018 right?
I think that was the time I did the tests previously, cause back then this was what you said. (and my tests were inspired by what you said :3 )
 

On 12/7/2018 at 3:01 PM, CryHaddock said:

You have to get an account to exactly level 10 with either the dex entry for Articuno, or a massive amount of stardust that I'm not even sure is possible to get without exceeding level 10.  Then you'd need it to build 30-90 days of friendship with the Articuno donating account.

Prior to being traded, the Articuno will be at least level 15 - but when the new account receives it it will always downgrade this level to either 10 or 12.  (It's either your trainer level, or your trainer level plus 2 and I'm not totally sure).
You could then port your very small Articuno to Let's Go where I'm sure it will fall foul of some legitimacy checker because it's such a convoluted and unlikely thing to do

I suppose you could do this more easily with Alolan Marowak, Ninjask and Shedinja.  Outside the legendaries I think those are the only Pokemon which can't be obtained below level 10 in either Pogo or 3DS

Was there any information that led you to believe that back then? I know I've seen silph road posts that mentions a + 2 level, but I've seen no associated images to back up that claim

[CryHaddock]

6 minutes ago, theSLAYER said:

But the experimentation you're referring to is post dec 2018 right?
I think that was the time I did the tests previously, cause back then this was what you said. (and my tests were inspired by what you said :3 )
 

Was there any information that led you to believe that back then? I know I've seen silph road posts that mentions a + 2 level, but I've seen no associated images to back up that claim

The vast bulk of the work would have been in whatever the first few months of trading being available were.
We had a lot of low level alts accounts because we used them to get Pokemon at level 1 that would be extremely hard to get otherwise. So we were very aware of the level cap (where wild Pokemon can only spawn up to your trainer level, and powering up only goes to your level +2)
We did enough level 1 hunting that we could plausibly have worked out that cap all by myself but it's also probable that my friend did read info from Silph Road. I personally didn't and just worked with what my friend told me and what I observed myself.

Edited December 26, 2020 by CryHaddock

[theSLAYER]

16 minutes ago, CryHaddock said:

The vast bulk of the work would have been in whatever the first few months of trading being available were.
We had a lot of low level alts accounts because we used them to get Pokemon at level 1 that would be extremely hard to get otherwise. So we were very aware of the level cap (where wild Pokemon can only spawn up to your trainer level, and powering up only goes to your level +2)
We did enough level 1 hunting that we could plausibly have worked out that cap all by myself but it's also probable that my friend did read info from Silph Road. I personally didn't and just worked with what my friend told me and what I observed myself.

if you still have access to those low level accounts, and could show a few screenshots of a mon traded in the first month, and its CP, that would probably help to solidify the case.
I'm unsure about mons acquired a while ago, but right now in Pokémon GO mons have the "date caught" and "date traded" at the bottom of the summary.
 

I've been trying to track down any confirmations to the trade cap being the same as the level cap, and so far the earliest trade media with a low level recipient is this video:
https://youtu.be/WyNebQJZfY4?t=883 [Video uploaded Jul 13, 2018, about 2-3 weeks after trading implemented in GO; Appears to be recorded on Jul 11, 2018.]

(Screenshot of trade, in case anything ever happens to the video):

As you can see, Gligar traded to a low level account (not level 10, but being low enough from 20).

Took me a while to find the Level of Gligar (old calculations maybe?), but it's level range and HP range indicates it is at level 20.

(https://thesilphroad.com/iv-rater)

If it was +2, the level would have been 18, not 20.
 

Based on that finding, my current belief is that either the trainer level +2 never existed [for trades, if below lv 20], or Niantic changed it in less than a month

[CryHaddock]

I do still have the logins for a large number of alt accounts but I'm not sure if any would have what you're looking for. What information specifically do you want to get?
I know for sure we never recorded a 'before and after' so the only information that could possibly be around is the 'after'.

On my main I do have some very old traded things and while the date function is mostly accurate it does fail on some very early things (I noticed it on things that were hatched out of an egg in 2016 then traded far later)

[theSLAYER]

12 minutes ago, CryHaddock said:

I do still have the logins for a large number of alt accounts but I'm not sure if any would have what you're looking for. What information specifically do you want to get?
I know for sure we never recorded a 'before and after' so the only information that could possibly be around is the 'after'.

On my main I do have some very old traded things and while the date function is mostly accurate it does fail on some very early things (I noticed it on things that were hatched out of an egg in 2016 then traded far later)

Yeah a before and after would be tough. The main thing I wanna know is if the +2 bs was ever real (for trades), within the first 2-3 weeks they rolled out GO trading. So any proof of it, would be showing the CP + traded date in summary (any lower than lv 20 CP, and the mon was traded). Because as that video + CP calculation shows, by Jul 11 2018, the supposed +2 was already not a thing then (if it was ever a thing).

One thing I'm unsure about, is when they implemented this "date" function for traded stuff, and whether it is retroactively applied to stuff traded prior.

[CryHaddock]

9 minutes ago, theSLAYER said:

Yeah a before and after would be tough. The main thing I wanna know if the +2 bs was ever real, within the first 2-3 weeks they rolled out GO trading. So any proof of it, would be showing the CP + traded date in summary (any lower than lv 20 CP, and the mon was traded). Because as that video + CP calculation shows, by Jul 11 2018, the supposed +2 was already not a thing then (if it was ever a thing).

One thing I'm unsure about, is when they implemented this "date" function for traded stuff, and whether it is retroactively applied to stuff traded prior.

The trouble with that is it would be very hard to find something that was definitely a high level on the before account. We definitely never traded a legendary to a low level account, and probably not anything from a raid either. It was mostly 'trade something that's spawning everywhere this week so it's one of the ten things in the account's dex'

 

For the practical application of 'was it technically possible to get a level 10 Articuno' I think the stardust costs of trading might have meant that was always impossible.
If you did a month's friendship you'd get 50 000 XP and shoot up in trainer level
If you didn't do the trade would cost 1 million stardust, which you surely can't obtain wthout attaining massive amounts of XP.

The only think I can think of that would have worked is the baby account catching the legendary for itself, by participating in a raid between level 5-10, and then trading. I dunno which legendary raids were possible for this sequence of events, but surely next to no one actually did it since there was little to no reason to even have alts until trading.
 

[theSLAYER]

55 minutes ago, CryHaddock said:

The trouble with that is it would be very hard to find something that was definitely a high level on the before account. We definitely never traded a legendary to a low level account, and probably not anything from a raid either. It was mostly 'trade something that's spawning everywhere this week so it's one of the ten things in the account's dex'

The initial trade don't have to be from a high level mon from a high level companion.

From what I can tell, low levels gets bumped up to lv 20 (Jun 22, 2018, appears to be pretty much from the beginning). So all we need to do is to sort by low CP + traded, and see if anything appears to be below lv 20.

I'm basically just checking if any lower level exists from trade. It appears the minimum from trade is 20 (regardless of what level they're trading you; bumped up to 20 if mon level is lower, bumped down to 20 is recipient's level is lower, level remains if trainer and mon is higher than 20)
 

So technically I'm checking two things, but they easily could roll into one.
1. Whether low level trades can even occur
2. Whether mons can downlevel below 20.

If low level trades can't even occur, then I have no reason to believe mons can downlevel below 20.

edit: Granted, your point on stardust is valid, however I want to know the limits of trade mechanics itself.
One could always get an Articuno via raid and not have to do the 1mil stardust.. I have caught legendaries on low level accounts to test, it was just way after this initial trade phase. Presently the only limiting factor to the +2 theory is the initial phase of trading.

[CryHaddock]

I have a Beedrill that was hatched from an egg. I know for a fact that it was obtained in the first week of Pokemon go being out, by my friend who briefly played the game.
Weedle was premanently removed from the egg pool really early on so there's no doubt the egg is from 2016.
It sat unhatched on her account until 2018 when I hatched it and traded it to myself.
That Beedrill now has an 'egg received' date that's actually the 2018 hatch date, and an 'obtained in a trade' date from a few days later.

[theSLAYER]

Just now, CryHaddock said:

I have a Beedrill that was hatched from an egg. I know for a fact that it was obtained in the first week of Pokemon go being out, by my friend who briefly played the game.
Weedle was premanently removed from the egg pool really early on so there's no doubt the egg is from 2016.
It sat unhatched on her account until 2018 when I hatched it and traded it to myself.
That Beedrill now has an 'egg received' date that's actually the 2018 hatch date, and an 'obtained in a trade' date from a few days later.

Is this the lv 1 beedrill you're proud about? What's the new hatch date or present trade day?

[CryHaddock]

1 minute ago, theSLAYER said:

Is this the lv 1 beedrill you're proud about? What's the new hatch date or present trade day?

Yes he is my boy, that combination of absurd rarity and total lack of appeal is chef's kiss.
According to Pogo he was 'received 21/08/2018' and 'obtained in a trade 22/08/2018'
In very bad timing I literally deleted three similar Pokemon yesterday, eg a Zubat from a 2016 egg since he is back in the pool and no longer unusual.
I did notice that each one of those also had an 'egg obtained' date exactly one day prior to the 'traded' date.
Which means that it's either the date they hatched (pretty plausible that I did the trades the following day), or that date is just totally made up from the traded date minus 1.

[theSLAYER]

13 minutes ago, CryHaddock said:

Yes he is my boy, that combination of absurd rarity and total lack of appeal is chef's kiss.
According to Pogo he was 'received 21/08/2018' and 'obtained in a trade 22/08/2018'
In very bad timing I literally deleted three similar Pokemon yesterday, eg a Zubat from a 2016 egg since he is back in the pool and no longer unusual.
I did notice that each one of those also had an 'egg obtained' date exactly one day prior to the 'traded' date.
Which means that it's either the date they hatched (pretty plausible that I did the trades the following day), or that date is just totally made up from the traded date minus 1.

Scratch that, I just tested, and proved that low level mons can be traded.
Unsure if it's a limited only to low level accounts, but huzzah.


(my level 1 chansey remained level 1)
[once again, either limited to low level accounts, or a recent change or (the bumping up to 20) limited to first week of trading]

So I guess the next thing you can be sure of, is to check CP+Traded+Premier Ball.
If the mon was traded in 2018 and has premier ball, it logically has to be caught at lv 20.
That may help, since as you said, you can't reliably tell what levels other mons are.

Yeah, I think my 2016 entries have their dates messed up too.

edit: Your Lv Beedrill is from August you say? You got anything from July, since that's the date range I was curious about.

[CryHaddock]

I very much doubt I ever traded something from a raid to a baby alt. Most of them were churned through very quickly to get a rare level 1, reach level 10 quickly and trade it for whatever it got in the dex along the way.
I will have a look at my oldest accounts though, on the off chance there are some traded growlithes at exactly the trainer level/exactly level 20.

What hypothesis are you trying to test with trading the level 1 Chansey? I don't quite follow.

[theSLAYER]

14 minutes ago, CryHaddock said:

I very much doubt I ever traded something from a raid to a baby alt. Most of them were churned through very quickly to get a rare level 1, reach level 10 quickly and trade it for whatever it got in the dex along the way.
I will have a look at my oldest accounts though, on the off chance there are some traded growlithes at exactly the trainer level/exactly level 20.

What hypothesis are you trying to test with trading the level 1 Chansey? I don't quite follow.

As I previously said (but I may not have been super clear, sorry bout that), if low level Pokemon were traded and unable to retain low level status (a silphroad post from Jun 22, 2018 shows how it gets bumped up to level 20), I have no reason to think that mons could go below 20 when trading.
(This signifies that Level 20 is a hard limit in any direction)

However, based on your Beedrill (in August 2018) and the Chansey I did prior to this post (used low->low and low->high), Lv 1 mons can be traded. Thus, breaking the hard limit level 20 limitation I was testing.

While the trainer level + 2 claim doesn't seem entirely founded (at least I've yet to find any media to back up that claim), the image of the Lapras being bumped to Lv 20 seems to be real (doesn't help that the commenters seem to be sure of that claim, tho the lack of media is always concerning).
Assuming that both of these things are real, the noted time frame of both phenomenon seems to overlap.

If so, that seems to indicate that a change in trading mechanic possibly occurred within the first few weeks of trading being initiated (end of June, start of July, 2018).
 

[CryHaddock]

Oh I hadn't heard of that theory. I have been routinely trading level 1 Pokemon since 2018. That was most of the point of trading for me and my friend. The oldest one I have is a level 1 Party Hat Raichu traded on 24/06/2018

I was never under the impression that trading could make levels increase; everything I heard was about levels being reduced. 
I think just about everyone was in the same position I was in - they knew that level reduction existed after trying to trade some level 30 thing, but they didn't know what the exact rules were and were mostly guessing them.

[theSLAYER]

33 minutes ago, CryHaddock said:

Oh I hadn't heard of that theory. I have been routinely trading level 1 Pokemon since 2018. That was most of the point of trading for me and my friend. The oldest one I have is a level 1 Party Hat Raichu traded on 24/06/2018

I was never under the impression that trading could make levels increase; everything I heard was about levels being reduced. 
I think just about everyone was in the same position I was in - they knew that level reduction existed after trying to trade some level 30 thing, but they didn't know what the exact rules were and were mostly guessing them.

No worries. I've made a few edits, not sure if you've seen them.

The Lv 1 Party Hat Raichu just narrowed the time frame down further. (assuming both mechanics changed at the same time, they existed till Jun 24/Jun 25 [variant date due to timezone differences])

Either the Lapras image was pre-trading official images and that the mechanic of bumping up was never implemented (tho I do doubt that; [a] an easy reverse image search should prove that and [b] commenters vouching for it), or that the change occurred super early into trading (first 3 days), or some kind of requirement was necessary (maybe trading low level mons between 2 lv >20 accounts?)

I'm under the impression that everyone just guessed, but there has to be someone who argued against it right?
Just need one person on silph road. Someone with images. Anyone. XD

[CryHaddock]

CP calculation has definitely changed once since 2018 by the way, so to work out levels from old videos you'd need the old formula.
It messed up the collections of people who had kept things at level 69, 420, 666 etc and there was much distress

 

One possibility for the Lapras oddity is the 'estimated CP after trade' calculation being flat out wrong.
This has occasionally happened the whole time trading has been live. I'm sure I remember noticing it both when trading was new and I definitely still see it now.

It may be that, if they committed to the trade, the Lapras would have retained its level and got a normal CP value for that level, which did not match the prediction.

For example just this week, I was about to trade a freshly caught thing. I realised I'd never checked it for random perfect IVs, so I looked at the estimated CP after trade to see if it was near the top.
The Pokemon's current CP was completely outside the range of estimated possibilities.
These were both grown up accounts with no possibility of level altering, the estimation is simply badly programmed.

Edited December 26, 2020 by CryHaddock

[theSLAYER]

6 minutes ago, CryHaddock said:

CP calculation has definitely changed once since 2018 by the way, so to work out levels from old videos you'd need the old formula.
It messed up the collections of people who had kept things at level 69, 420, 666 etc and there was much distress.

  yeah, the old silphroad IV rater I linked above uses the old formula.

Before I found that site, I was struggling to get Gligar's level using my own calculations and other CP calculations.

So given how it found Gligar's level without a struggle, I'm not too concerned with old level calculations.
 

7 minutes ago, CryHaddock said:

One possibility for the Lapras oddity is the 'estimated CP after trade' calculation being flat out wrong.
This has occasionally happened the whole time trading has been live. I'm sure I remember noticing it both when trading was new and I definitely still see it now.

It may be that, if they committed to the trade, the Lapras would have retained its level and got a normal CP value for that level, which did not match the prediction.

I would think it's an oddity, if everyone in the comment section said it was. Except they all agreed that upleveling was possible with trading (then). If we could find a wiki/page that has the old mechanics, that would be great. Or maybe an old megapost. Given that everyone agreed, I can only assume they got their information from the same source or sources.

10 minutes ago, CryHaddock said:

For example just this week, I was about to trade a freshly caught thing. I realised I'd never checked it for random perfect IVs, so I looked at the estimated CP after trade to see if it was near the top.
Instead its current CP was completely out of the range of the estimates.
These were both grown up accounts with no possibility of level altering, the estimation is simply badly programmed.

That certainly is odd. Tho I've yet to see anyone make a proper post about it, so that's just kinda sad. Need more people to report things (in proper threads/reddit channels) so we can properly keep track. We're experiencing this right now cause people didn't seem to report changes

 

[CryHaddock]

Honestly people sharing information that they absolutely have not tested with the air of someone who deeply understands the game is f a recurring theme with pokemon go culture
Because it's significantly harder to datamine than Switch Pokemon games, and Niantic essentially communicates nothing in detail, there was a real lack of information. Add in its insane popularity and you had this massive demand for information that a lot of people were filling by just... making stuff up with an air of authority.

For a hot minute everyone was taking completely seriously the idea that people in a group raid would get more shinies if only you didn't press a certain button and every raid was people yelling at each other DON'T PRESS THE BUTTON.
There was a new thing like that seemingly every month, and a vastly greater rate  than anyone actually testing or falsifying things.
Sharing the fourth infographic in as many days on how to control Eevee's evolution was a higher priority than testing if any of them worked.

Basically I find it very easy to believe that none of those commentors knew if this was normal trading behaviour and none of them ever checked


 

[theSLAYER]

22 minutes ago, CryHaddock said:

Honestly people sharing information that they absolutely have not tested with the air of someone who deeply understands the game is f a recurring theme with pokemon go culture

That is sadly something I agree with. It's also why I kinda wanna see media that could prove/disprove various of those mechanics, and also leave behind images that can help prove my points..

Speaking of which, I hope you don't mind sharing images of the Hat Raichu. It's not that I don't believe you (I do), but it'll be easier to check back in the future and go see, this information was proved with this picture. You can pixelate any private info if necessary.

22 minutes ago, CryHaddock said:

Because it's significantly harder to datamine than Switch Pokemon games, and Niantic essentially communicates nothing in detail, there was a real lack of information. Add in its insane popularity and you had this massive demand for information that a lot of people were filling by just... making stuff up with an air of authority.

Yeah and it doesn't help that Niantic (at least in the older days) can change mechanics with the drop of a hat, without giving any warning whatsoever. Remember when Deoxys and Shadows suddenly could no longer be traded? XD

22 minutes ago, CryHaddock said:

For a hot minute everyone was taking completely seriously the idea that people in a group raid would get more shinies if only you didn't press a certain button and every raid was people yelling at each other DON'T PRESS THE BUTTON.

What the heck, that was a thing? I was lucky I never went through that ><

22 minutes ago, CryHaddock said:

There was a new thing like that seemingly every month, and a vastly greater rate  than anyone actually testing or falsifying things.
Sharing the fourth infographic in as many days on how to control Eevee's evolution was a higher priority than testing if any of them worked.

Basically I find it very easy to believe that none of those commentors knew if this was normal trading behaviour and none of them ever checked

As much as the opinions of the commenters being united is something that bothers me, and that it affects whether I believe upleveling is a thing, I'm more intrigued in their source.
If they're just perpetuating an idea (that they never verified for themselves), the idea has to come from somewhere, and hopefully that source has images (or something). I've yet to find it tho.

So far, all we've really been able to confirm is:
1. Jul 13, 2018 and after, when trading high lvl mon from high->low lv acocunt, mon's lv bumped down to Lv 20.
2. Jun 24, 2018 and after, Lv 1 mons can trade and retain it's level (doesn't get upleveled to 20)

I frankly would like to see proper images of bumping down to trainer level + 2. and maybe other sources of upleveling.
My 2 hypothesis are: (1) these 2 mechanics never happened or (2) they were likely both changed before Jun 24, 2018.
[I am grouping these 2 mechanics together, because imo, it's easier to push the update to change/fix all the perceived faults at once.]
 

[CryHaddock]

I found one account so far that tentatively supports 'level reduction was always to level 20'

05/08/2018 (trade date on both Pokemon)
Krabby and Spheal, both level 20, both caught in wild.
Account was approx level 11-14 at the time.

This account is currently level 22, but I'm certain that this is entirely because of the Eevee special research I did on it recently, and that the account was still under level 15 as late as 2020.

Edit: found a similar story on a second account, though all those trades were in 2019.
Two of the traded Pokemon are exactly level 20, several others are various levels under 20, the account itself was not level 20 when these trades happened.

A lot of the time the source for these false facts were sample sizes of one. As in people would try something once, it went a certain way, and they made a completely wrong conclusion about why it happened.
Very possible someone's level happened to be 18 when they discovered the level rounding down.

People might have got it wrong by being forced to guess, since Pokemon Go doesn't exactly tell you Pokemon levels.  You've got to use a screen reader app unless you are crazy good at eyeballing that arc display. I can't remember how difficult it was in 2018 to even know your Pokemon's level accurately.

Edited December 26, 2020 by CryHaddock

[theSLAYER]

15 minutes ago, CryHaddock said:

I found one account so far that tentatively supports 'level reduction was always to level 20'

05/08/2018 (trade date on both Pokemon)
Krabby and Spheal, both level 20, both caught in wild.
Account was approx level 11-14 at the time.

This account is currently level 22, but I'm certain that this is entirely because of the Eevee special research I did on it recently, and that the account was still under level 15 as late as 2020.

As per the video I linked way earlier, at least on July 11th 2018, the downleveling below 20 wasn't a thing.
Thus, your August 2018 trades doesn't really affect it, since July came before August. (tho it does add to it :3 )

My interest is in trades between End June and Early July. That's the timeframe where downleveling below 20 may have been possible.

Quote

A lot of the time the source for these false facts were sample sizes of one. As in people would try something once, it went a certain way, and they made a completely wrong conclusion about why it happened.
Very possible someone's level happened to be 18 when they discovered the level rounding down, or else that they were simply forced to guess.  The game never exactly tells you your Pokemon's levels without using some kind of screen reader app. I can't remember how difficult it was in 2018 to even know your Pokemon's level accurately.

I don't disagree, but all this is just conjecture. At the very least, we know that July 11th and onward, downleveling past lv 20 isn't a thing.
But Jun 21st to Jul 10th 2018 is unaccounted for. Niantic could have changed the trade mechanics in the initial weeks of launch.
It's not uncommon for devs to change things up within first few weeks of a new feature launch, due to exploits, bug reports or overwhelming feedback.

As much as it was hard to judge levels then, we can judge them now. So as long as some photo or footage exists, we can properly prove or disprove these claims.
Don't get me wrong tho: since finding the July 11th video (that was uploaded on July 13th), I am in the camp that thinks people probably thought it was trainer level + 2 when it was merely down to lv 20, but without concrete proof it's hard to say for sure.
 

Still, it could help if we could find who first made such guesses/conclusions.

[CryHaddock]

Yeah I'm just giving you the data I have access to, not trying to argue a counter thesis.
It would take a real stroke of luck to find data from such a narrow window, particularly when no one spotted the effect at the time.  I hope you get lucky with it some time.