Psr Update And Hold On Patch.
#421
Posted 11 June 2020 - 06:44 PM
#422
Posted 11 June 2020 - 06:55 PM
Smoke Jagaur MAX, on 11 June 2020 - 06:44 PM, said:
1) It did not take you years to get to tier one unless you seldom play.
2) This wouldn't be the first time they've reset PSR (and tiers).
3) Tiers are rather meaningless unless there are more than ~100 queuing at roughly the same time as you.
Edited by MovinTarget, 11 June 2020 - 06:55 PM.
#423
Posted 11 June 2020 - 06:56 PM
Willard Phule, on 11 June 2020 - 02:04 PM, said:
Here's the thing I don't understand about this whole new PSR calculation thing. Why is it still based on a whole win/loss scenario as opposed to individual performance? I mean, sure, winning matters but why isn't it just a flat rate modifier slapped on at the end, after everything else is calculated? Being the guy that gets a 500+ MS on the team that got stomped 1/12 kinda says more about your ability than the 11 other guys that struggled to break 100.
You get more matchscore for winning. So it is already factored into matchscore.
Can we move on from the wining/losing being a factor anymore than you get more matchscore for winning?
Edited by Remover of Obstacles, 11 June 2020 - 06:56 PM.
#424
Posted 11 June 2020 - 07:02 PM
Remover of Obstacles, on 11 June 2020 - 06:56 PM, said:
You get more matchscore for winning. So it is already factored into matchscore.
Can we move on from the wining/losing being a factor anymore than you get more matchscore for winning?
As long as match score isn't used for PSR or MM, sure, we can move on.
#425
Posted 11 June 2020 - 08:05 PM
Brauer, on 11 June 2020 - 04:35 AM, said:
What? I don't see how that's relevant. You complained about nascar so I told you how to break nascar.
FYI in case you have missed it I have been consistently against the group queue/solo queue mash up. While it's here I'm going to drop with a group, but I don't think it's good for the queue overall.
On the contrary Group PSR inflates the PSR of the under performers in the group. They effectively ride the wake of their best player
But I was just asking the question and you seem to agree that using the average PSR of a group bucket is an error for accurate PSR for MM and especially in solo que.
#426
Posted 11 June 2020 - 08:06 PM
Nightbird, on 11 June 2020 - 03:54 PM, said:
Another point on how easy it is to do rolling average:
Today we have a saved variable per player called PSR, and after a match we have a PSR_change. So the code for the new PSR is:
new_PSR = old_PSR + PSR_change.
To do a rolling average for MS over the last 50 games, we only need to save two variables, avgMS and N (N is matches played). After a match we have currentMS score.The code for the new avgMS is:
if N<50 then new_avgMS = old_avgMS * N/(N+1) + currentMS / (N+1)
if N>=50 then new_avgMS = old_avgMS * 49/50 + currentMS / 50
new_N = old_N+1
To do a rolling average for W/L over 50 games, likewise we need to save just two variables WLR and N. After a match we get a RESULT which is 1 for a win or 0 for a loss. Code:
if N<50 then new_WLR = {N * [old_WLR/(old_WLR + 1)] + RESULT} / {N * [1 - old_WLR/(old_WLR + 1)] - RESULT + 1}
if N>=50 then new_WLR = {49 * [old_WLR/(old_WLR + 1)] + RESULT} / {49 * [1 - old_WLR/(old_WLR + 1)] - RESULT + 1}
Takes a programmer 1 hr to do all of this. A new MM would take time but just a rolling MS or WL average took me 10 minutes to figure out.
While I understand this formula is easy, that's not why Paul said it wasn't feasible. The game server that determines PSR changes doesn't store all that much data about the player. It's not the same system that has that data. The game has to use what info it has from the match at hand. I'm not even sure the game server knows your current PSR--that would have been handled by the matchmaker before you connected to the game server.
PGI can query this data, but doing it on the live game server is the part that isn't feasible under their current architecture.
I'm extrapolating here, but that's my understand from Paul's comments on twitter.
--a programmer.
#427
Posted 11 June 2020 - 08:16 PM
Smoke Jagaur MAX, on 11 June 2020 - 06:44 PM, said:
Sorry matey the tier system is just an Exp. bar as it stands it means nothing.
The only vanity stat IMO should be K/D or total career kills so even the most industrious (more games) can compete to obtain some notoriety.
Kill's per match or K/D I personally have no serious interest in because a lot of the fun in this game is trying silly builds for the fun of building and tinkering. Pre'mades remove the fun of discovery from the game but they are definitely the way to go when your in a competitive group.
#428
Posted 11 June 2020 - 08:16 PM
yrrot, on 11 June 2020 - 08:06 PM, said:
PGI can query this data, but doing it on the live game server is the part that isn't feasible under their current architecture.
I'm extrapolating here, but that's my understand from Paul's comments on twitter.
--a programmer.
Bless you sir.
#429
Posted 11 June 2020 - 08:34 PM
yrrot, on 11 June 2020 - 08:06 PM, said:
PGI can query this data, but doing it on the live game server is the part that isn't feasible under their current architecture.
I'm extrapolating here, but that's my understand from Paul's comments on twitter.
--a programmer.
For PSR today, the game server passes the PSR_change result to the player database server, which has the player's PSR value stored.
For avgMS, the game server passes the MS result to the player database server, which has the player's avgMS and N values stored.
For WLR, the game server passes the 1/0 end result to the player database server, which has the player's WLR and N values stored.
I'm not sure what's different. You have the player database store 2 values instead of 1 per player. The game server still only needs to hand off one value.
- also a programmer
Edited by Nightbird, 11 June 2020 - 08:37 PM.
#430
Posted 11 June 2020 - 08:43 PM
Dauntless Blint, on 11 June 2020 - 08:05 PM, said:
On the contrary Group PSR inflates the PSR of the under performers in the group. They effectively ride the wake of their best player
But I was just asking the question and you seem to agree that using the average PSR of a group bucket is an error for accurate PSR for MM and especially in solo que.
First, re: Nascar, my apologies I believe I mixed up posts when looking over the discussion.
Re: groups and PSR, I don't know the best way to consider groups within the matchmaker. IMO it's probably destined to be a mess to some extent or another. Since WLR is the best indicator of future performance I'd imagine taking the average group WLR of players in a group is the best way to assess their impact. Unfortunately, PGI seems to have taken solutions like that off the table.
By group WLR I mean their WLR when in any group,.
#431
Posted 11 June 2020 - 08:44 PM
Honestly, I found this thread via wanting to know what this "PSR" patch was about and what exactly PSR even was. If you actually wanted input, don't assume that everyone will know what you are abbreviating.
I'm ranked Tier 2. I have zero notion of how many ranks there are, whether the lower your number is the more skilled you supposedly are or visa versa.
I am not a social media guru, I don't have time to be searching through all media outlets to try to track down the various propositions that have apparently been made. You asked for community input, but what exactly do you consider the "community". I would like to assume that means all players and not just the tech savvy and social media gurus. I do take interest in patches and the potential impact they may have on gameplay.
Based on the general gist of what's been discussed here, and the information given at the beginning of the post, I think PGI did the right thing in NOT trying to tweak the systems that are already in place. Asking for community input is nice to try to appease the vocal 2-5% (well maybe it's a higher % now with player base being so low these days). I do sincerely hope that the delaying of the patch means they have realized the main issue (which I do believe they have given the way this was proposed and the wording used in the initial post).
The main issue is not that PSR and MM are not working correctly. The problem is that we are bleeding player base, and it has reached a point that those systems are trying to make the best of a situation they were not intended for. The only input I have that is "new" from my earlier post (around page 20 if anyone wants to look back at it after skipping to the end) is that if any changes are made now to these systems, PLEASE do the usual back-up copy of where things are now. Make changes to actual gameplay to encourage player base growth. Appease the vocals with a change to the systems if need be to stop the loss of current players that are actually into this side of things. These are good moves. Grats to all those talking about this being "smoke and mirrors" for recognizing it for what it is. Shame on you same people for not catching on to the reason for them and the real issue being faced here.
I'm sure PGI has historical data from when the game had a good sized player base that let the systems work. Relatively sure they probably have someone who has tracked the player base size over the years too. And a history of the gameplay changes. Time to put all those together, find the pattern, revert the changes that caused players to leave. Regardless of what the vocal minority say.
Edit: I've now had my fill of the discussion, which now has consumed around 3-4hrs of time. I need to shoot something.
Edit 2: Regarding faction play, at this point you might as well only give the option of being Clan or Inner Sphere. While this does cause some MAJOR issues with the loyalty reward system, it will not make it so that Houses and Clans that historically do not like each other (for the players who care, and if you're playing that mode, you generally do) from dropping along side their normal enemies.
Edited by Ghoja, 11 June 2020 - 09:02 PM.
#432
Posted 11 June 2020 - 08:51 PM
No matter how you weigh the components going into match score calculations, every player will have an average PSR change per game. Let's call this X.
After 100 games, your "skill" = 100X. After 1000 games, your "skill" = 1000X. After 5000 games, your "skill" = 5000X.
Is your true skill actually increasing by a factor of 10, 50, 100 over time? Or is this in reality not a measurement of skill? It's not. It's an EXP bar that is disjointed from any representation of your skill. Do you want to know how well a Match Maker - that can only see EXP bars - makes matches? Take a raw egg and hit it with a hammer. That well.
My suggestion: https://mwomercs.com...32#entry6337632
See the first link for a MM proven to work.
See the second link for proof MS is useless.
Edited by Nightbird, 11 June 2020 - 09:01 PM.
#433
Posted 11 June 2020 - 09:01 PM
Nightbird, on 11 June 2020 - 11:50 AM, said:
Why not see that as a reward? You can have a great match, but your team lost it for you, so the MM will reward you with an easier match next round.
Another way to see it is: You have 24 players to put into two teams, for the first 8 players 4 are in a group and 4 are solo
Player 1: 400 avgMS 10 WLR (group)
Player 2: 400 avgMS 10 WLR (group)
Player 3: 400 avgMS 10 WLR (group)
Player 4: 400 avgMS 10 WLR (group)
Player 5: 400 avgMS 2 WLR (solo)
Player 6: 400 avgMS 2 WLR (solo)
Player 7: 400 avgMS 2 WLR (solo)
Player 8: 400 avgMS 2 WLR (solo)
MM puts them into two teams:
Team A
Player 1: 400 avgMS 10 WLR (group)
Player 2: 400 avgMS 10 WLR (group)
Player 3: 400 avgMS 10 WLR (group)
Player 4: 400 avgMS 10 WLR (group)
Team B
Player 5: 400 avgMS 2 WLR (solo)
Player 6: 400 avgMS 2 WLR (solo)
Player 7: 400 avgMS 2 WLR (solo)
Player 8: 400 avgMS 2 WLR (solo)
At this point, do you
1. want the MS based PSR match maker that says, Team A and Team B are equal, split the remaining players equally?
2. want the WL match maker that says, Team A is stronger than Team B, put more good players from the remaining players on Team B?
I then want the MM to recognize the power of the chassis in those teams according to how their global average match score stacks up. (I'm appealing to you now) EDIT: (and an amendment: Chassis W/L + Average chassis Match Score)
Because my effective PSR in my serious builds is brackets higher than builds for fun or off meta nonpremades or even just a chassis I am learning. (If it can't be done in MWO I want it to be a consideration in a new game or reboot)
(Chassis Service Record) to augment PSR. It's still pulling from matchscore data
Because win loss is only accurate if I play the same chassis my entire career.
Win loss becomes a blunt tool once you hit your correct PSR as if it is working correctly,you should average out a win loss of 1 to 1. (Unless you are at the extremes of the bell curve of course)
Ideally group and solo should be separate, but taking chassis power into consideration you can make a tighter match.
(The off shoot on a Global Chassis average Match Score a PSR for mechs would take into account some mechs are more equal then others... a diagnostic tool for buffing and nerfing that no one could argue with. Especially as it would be public data.)
How do you do that? well you give an end of match modifier for if you did better or worse against the global average (MS) for that chassis and if you did: PSR+1 if you didn't PSR -1.
Now you have player & chassis proficiency.
(I just wanted to add that ultimately your example for now just shows the population is too low otherwise the MM wouldn't have so much disparity in the first place. If these W/L 10 guys are at the top of the bell curve no one can match them and they are the exception.)
You can now allow MM to stack high performance chassis against them to provide them with sport and a challenge.
(Page 14 off the top of my head)
To surmise: 3 metrics. W/L + Player MS + Global Average Chassis MS (+ chassis W/L)<<this is huge
Does PGI track this.
Because you could give a team of COMP guys a run for their money if you need a bunch of average potato in madcats (say 12) just to provide them with sport. It's an exaggeration of course to serve the example.
Edited by Dauntless Blint, 11 June 2020 - 10:32 PM.
#434
Posted 11 June 2020 - 09:21 PM
Nightbird, on 11 June 2020 - 08:34 PM, said:
I'm not sure what's different. You have the player database store 2 values instead of 1 per player. The game server still only needs to hand off one value.
Literally this part. This isn't a hard change, but it's now touching 1) the end of match logic, 2) the queries that run end of game, 3) the player database, 4) the matchmaker as a go between to hand off the new values you need in the game server to computer change in PSR. So now a change that was going to change float values in a table on the game server would be touching 2 game critical systems that require programmer intervention rather than a designer changing values. Plus QA scope creep.
#435
Posted 11 June 2020 - 10:01 PM
yrrot, on 11 June 2020 - 09:21 PM, said:
Literally this part. This isn't a hard change, but it's now touching 1) the end of match logic, 2) the queries that run end of game, 3) the player database, 4) the matchmaker as a go between to hand off the new values you need in the game server to computer change in PSR. So now a change that was going to change float values in a table on the game server would be touching 2 game critical systems that require programmer intervention rather than a designer changing values. Plus QA scope creep.
Gotcha. Unfortunately it's the minimum work needed for a functioning MM.
#436
Posted 11 June 2020 - 10:08 PM
1) Are a person's WLR in both the current and proposed MM the only factor that would contribute to the hidden skill?
2) When you refer to it as being a "hidden" value in current MM, is it hidden because current MM is matching based mostly off of Tier ranking. Thus Tier 1 players having a hidden value that, if the tier system was actually representative of skill and not mostly an XP bar, a lot of Tier 1 players would be in a lower tier currently?
3) If I understand the new MM proposal you have made, it would make the entire Tier system obsolete would it not.
4) An interesting point was made a few posts ago regarding high-performance chassis (I assume he's referring to how well it performs overall or using a meta build on it as well). I too bounce often between running meta builds and non-meta builds. If I'm being matched solely off of WLR, the MM can't tell what I'm running. Is you're mindset (and perhaps data you've analysed) that my WLR would basically average out my skill level and use that for MM purposes?
Anywho, keep up the good work.
- Not a programmer.
#437
Posted 11 June 2020 - 10:18 PM
#438
Posted 11 June 2020 - 10:23 PM
Dauntless Blint, on 11 June 2020 - 09:01 PM, said:
Now you have player & chassis proficiency.
(I just wanted to add that ultimately your example for now just shows the population is too low otherwise the MM wouldn't have so much disparity in the first place. If these W/L 10 guys are at the top of the bell curve no one can match them and they are the exception.)
You can now allow MM to stack high performance chassis against them to provide them with sport and a challenge.
(Page 14 off the top of my head)
To surmise: 3 metrics. W/L + Player MS + Global Average Chassis MS <<this is huge
Great idea, I also think the MM should take the chassis in account.
But I would not modify the amount of PSR gained/lost based on the chassis. I would say that MM modifies your effective Tier based on the chassis. So, if you use an unskilled and/or lemon chassis, MM should treat you as a Tier 2 or even Tier 3 player, even if you are normally Tier 1. And perhaps put you in a higher Tier if you run an OP build.
So:
- unskilled mech: effective Tier + 1
- known bad chassis: effective Tier + 1
- OP chassis: effective Tier - 1
The reason is: If you really skill a new chassis, or you are running around in a troll / fun build you are not playing at your best. Playing in a lower tier might result in more enjoyable games where you might make a difference instead of just being carried.
#439
Posted 11 June 2020 - 10:32 PM
Brauer, on 11 June 2020 - 11:23 AM, said:
Only because in the game, no one cares about "living to take part in the next mission".
There's basically no repair cost, no rebuild time, no injury, no resupply limits.
It would be a d**n sight different if you sent 12 people in one end of a 3 mission run, each mission facing 4-8 opponents, and a limited resupply capacity until they came out the other end.
17 user(s) are reading this topic
0 members, 17 guests, 0 anonymous users