406 replies to this topic

[Mystere]

Prototelis, on 18 October 2019 - 04:58 PM, said:

Oh thats cute. The wiki written after the fact by not PGI.

On a domain owned by PGI^(*).




(*) Or are you calling the folks who wrote the pages liars?

Edited by Mystere, 19 October 2019 - 04:27 AM.

[John Bronco]

Kana wrote the wiki, and while he's not a liar, he's also not 100% correct.

[Sjorpha]

Mystere, on 19 October 2019 - 04:24 AM, said:

On a domain owned by PGI^(*).




(*) Or are you calling the folks who wrote the pages liars?

I hardly ever call people liars, imo deliberate lying isn't nearly as common as a lot of people think and more often it is about misunderstandings or someone being in genuine delusion about their own previous position after being proved wrong etc. The latter happens all the time and to almost everyone, your brain restructures your memories and interpretation of your own memories to fit an updated narrative and others will percieve it as dishonest while you beleive you're just clarifying yourself. There are many other examples of different processes that lead to percieved "lying" and projections of dishonesty, we are just still fairly primitive primates and it shows.

None the less the wiki contradicts the spirit of the original presentation of the psr/tier system, which was proposed as a better replacement for elo based matchmaking promising more balanced matches and better separation of skill levels.

You can dress it up after the fact, and it's quite possible that the designers of the system have convinced themselves that it was never intended to improve matchmaking (which seems slightly absurd on it's face but ok..) and are now feeling honest and confident in defending statements such as the one you quoted, but honest or not the fact remains that the system doesn't work to make matches more balanced than random matchmaking would, and that's a big problem.

Consider the following abstracted dialogue:

Community: "Elo isn't working well, matches are unbalanced. we need a new system"
PGI: "Here's a new system based on tiers and psr"
Community: "The new system isn't working, matches are even more unblanced now and everyone is in tier 1"
PGI and their apologists: "Oh, there is a misunderstanding here. We never intended the new system to improve matchmaking, why would you expect that?"
Community: "wtf????"

You see how absurd this above conversation is even if PGIs wiki is telling the truth about the intentions of the system?

We're obviously not going to get a new matchmaking system at this point though. The best we could hope for is something like a tweak to the psr algorithm removing the upwards bias to psr changes, which would cause at least the most misplaced part of the tier 1 blob to fall back down to a more realistic tier 2 and 3, but even that is probably beyond the scope of PGIs current investment in MWO.

Edited by Sjorpha, 19 October 2019 - 05:55 AM.

[Prototelis]

Mystere, on 19 October 2019 - 04:24 AM, said:

On a domain owned by PGI^(*).




(*) Or are you calling the folks who wrote the pages liars?

Do you believe everything you read in American history textbooks because they were written in America?

Edited by Prototelis, 19 October 2019 - 10:11 AM.

[Jackal Noble]

It’s cold hard fact that everything written in American textbooks is 1000% true.

I’m curious what they did to offend you so. Might as well bring the Bible into play under that context, and self help books too.

[Mystere]

Prototelis, on 19 October 2019 - 09:07 AM, said:

Do you believe everything you read in American history textbooks because it was written in America?

FYI, I rarely read *any* history books written by Americans, period^(*).

But PGI is Canadian, so there is *still* hope for them, lest they've already been irrevocably contaminated.



(*) If they can't even get their own history right, especially the horrors they have visited on the world, how can I expect then to get others' history right?

Edited by Mystere, 19 October 2019 - 10:10 AM.

[BackShot]

#FreeASH

[DANKnuggz]

Heh I'd been away for a bit when the tiers went in. I was already T3 when I came back so I had no idea t5/t4 were so quick.

Did they change this later? Because I seem to remember a friend being stuck in t4 for a bit and winning plenty of matches

[Prototelis]

That happens when you, for whatever reason, don't play well during the "seeding"

[Khobai]

Sjorpha, on 18 October 2019 - 02:28 AM, said:

If that was true the system would be ok. While I don't agree with your way of describing types of players, and your obvious disdain for people whose only crime is being good at something, it would be fine if their performance actually put them in those different separate tiers.

This is how it really is though:

Tier 2-5: New players on their way to tier 1, and a few extremely bad players.
Tier 1: everyone else, probably more than 90% of the players.

Thats the problem. The upwards bias is extreme, everyone gets to tier 1. You don't need to be good, you don't need to be competitive, you don't even need to be average. You can be in the bottom 30%, pretend the game is tabletop and role play in stock mechs, you will still be tier 1 sooner or later.

everyone is supposed to get to tier 1 eventually. working as intended.

the problem is theres no ELO system that kicks in at tier 1.

[Sarsaparilla Kid]

DANKnuggz, on 21 October 2019 - 04:28 PM, said:

Heh I'd been away for a bit when the tiers went in. I was already T3 when I came back so I had no idea t5/t4 were so quick.

Did they change this later? Because I seem to remember a friend being stuck in t4 for a bit and winning plenty of matches

Am in the same boat...after 2k matches in the Elo years, I left, then came back in T5, with 20% still to go to get to T4. Even on days with a number of good games in a row, the meter ticks ever so slowly. It seems to be a case of the prior 2k matches weighing me down so my progress doesn't go up as fast as someone just starting out, as others have demonstrated above. I wasn't around during the initial 6 months or so they used to seed existing players after moving from Elo, so if that's the case with your friend as it was for me, yeah...it's a very slow grind.

[Vxheous]

Sarsaparilla Kid, on 22 October 2019 - 02:01 PM, said:

Am in the same boat...after 2k matches in the Elo years, I left, then came back in T5, with 20% still to go to get to T4. Even on days with a number of good games in a row, the meter ticks ever so slowly. It seems to be a case of the prior 2k matches weighing me down so my progress doesn't go up as fast as someone just starting out, as others have demonstrated above. I wasn't around during the initial 6 months or so they used to seed existing players after moving from Elo, so if that's the case with your friend as it was for me, yeah...it's a very slow grind.

The was no initial 6 months to seed players. Our accounts were converted pretty much on a 1:1 from Elo based to PSR Tier based. For instance, when the Tier system patch hit, I was placed about 75% of the way through Tier 2 (I wasn't as good as I am now), while almost all the top players in the Elo days (the comp players) immediately had maxed out Tier 1 bars. If you came back and have a Tier 5 bar, then your Elo from the Elo days was really low and you deserve to be Tier 5.

The only accounts that get a "seeding" period are brand new accounts, as a way to discourage making alts and just farming bads all day in Tier 5 as some kind of ego boost.

Edited by Vxheous, 22 October 2019 - 03:47 PM.

[Sarsaparilla Kid]

Vxheous, on 22 October 2019 - 03:42 PM, said:

The was no initial 6 months to seed players. Our accounts were converted pretty much on a 1:1 from Elo based to PSR Tier based. For instance, when the Tier system patch hit, I was placed about 75% of the way through Tier 2 (I wasn't as good as I am now), while almost all the top players in the Elo days (the comp players) immediately had maxed out Tier 1 bars. If you came back and have a Tier 5 bar, then your Elo from the Elo days was really low and you deserve to be Tier 5.

That's not what customer service told me when I questioned how tiers worked when I returned. I'll have to dig up that conversation...

[Vxheous]

Sarsaparilla Kid, on 22 October 2019 - 03:48 PM, said:

That's not what customer service told me when I questioned how tiers worked when I returned. I'll have to dig up that conversation...

They were wrong then, because there was no "placement" period. Unless you made a brand new account. All existing accounts were placed into one of the 5 tiers based on previous Elo performance. It worked for the first 6 months, because players placed into their respective tiers belonged in their respective tiers, and played against opposition of their respective tiers, creating fairly even matches. The Tier system became a problem after 6 months to a year as the positive upward shift gradually "leveled" players out of their proper Tier.

Edited by Vxheous, 22 October 2019 - 04:07 PM.

[Dimento Graven]

As I recall from what we were told, they had the stats from like years of play, ran through the various numbers of those stats and you were placed in the Tier you would have ended up in if starting from scratch (Tier 4) with the Tier system in place at the beginning of those 2 years.

[Sjorpha]

Khobai, on 21 October 2019 - 10:45 PM, said:

everyone is supposed to get to tier 1 eventually. working as intended.

the problem is theres no ELO system that kicks in at tier 1.

Pgi replaced elo with the tier system. If we had a working elo system the would be no need for tiers.

Or you could make the tier distribution a 0 sum division of the performance bell curve. Like bottom and top 5% are tier 5 and 1, next bottom and top 15% are tier 4 and 2 and middle 50% are tier 3. Something like that.

You only need one system if it works.

Edited by Sjorpha, 22 October 2019 - 05:42 PM.

[Scout Derek]

PhoenixFire55, on 16 October 2019 - 03:44 AM, said:

Huh? ... And where did I post anything there?

It was a context post to give you "insight".

[Sarsaparilla Kid]

Vxheous, on 22 October 2019 - 04:04 PM, said:

They were wrong then, because there was no "placement" period. Unless you made a brand new account. All existing accounts were placed into one of the 5 tiers based on previous Elo performance. It worked for the first 6 months, because players placed into their respective tiers belonged in their respective tiers, and played against opposition of their respective tiers, creating fairly even matches. The Tier system became a problem after 6 months to a year as the positive upward shift gradually "leveled" players out of their proper Tier.

Gotcha...that must have been where the 6 month figure stuck in my head, but you're correct, it had nothing to do with the original conversion to the tier system. I found my conversation with customer support, but all it did was point to the patch notes and left me scratching my head as to what they really used to place me.

Per the patch notes at the time:
A Tier 1 player will never play against a Tier 4 or Tier 5 player.
New players are automatically seeded into the mid-to-high range of Tier 4.
All current players will be seeded into whichever Skill Tier best aligns with the historical data already present in their account. The historical data pool used for generating your Tier placement goes back until January 2015.

The problem is that I did not have any matches played after the fall of 2014 when I took a break from the game until I returned in the fall of 2016. If they determined placement using the historical data between January 2015 and August 2015, when the tier system went into effect, what did they use to place me if I had no history in that time frame? Did they go all the way back to the end of open beta? Did they look at the first 25 matches after I returned as if I was a new player, in which I was probably derping around experimenting with the new meta? They didn't have an official policy to account for players like me in that transfer, so I guess they had to improvise...

[PhoenixFire55]

Scout Derek, on 22 October 2019 - 07:08 PM, said:

It was a context post to give you "insight".

Yeah, I figured that like a few hours afterwards.

[justcallme A S H]

Inatu Elimor, on 18 October 2019 - 09:18 AM, said:

I don' t get the point. Why don't you assist the lower tiers at risk loosing game with your high performing skills? What matters ? You could help those peeps so much. What else would be the use of you high performer getting into low tier games only to proof your so good ??

Did you read the first post?

I clearly outlined WHY I did it.

It was to debunk some of the misinformation that some users post around here because they are no idea what they are talking about.