139 replies to this topic

[MischiefSC]

Xavori, on 28 November 2017 - 03:58 PM, said:

No. Right there in that paragraph you're pointing out exactly what I've been saying. IT'S NOT JUST WIN LOSS. Quality matters. That's why beating a good team is worth more than beating a bad team.

What's more, you also finally acknowledged that you're not looking at W/L ratios when looking at ELO (Yes, I can totally keep capitalizing if I was so ppppbbbbttttt!!!). You're looking at a rating based on your W/L vs the quality of your opponents and any other considerations built in to the equation (like with TrueSkill which builds in piles of them). Winning and losing is only used to determine direction of movement. Literally everything else involved in the rating is something other than winning or losing.

As for trying to pull meaning out of my stats on the leaderboard... *falls down laughing and dies*

If I'm so inclined, I can do this:
https://steamuserima...3A1DB72472978D/
(ie. KMDD almost half the other team)
or in faction do this:
https://steamuserima...A0800697B3A87E/
(ie. basically KMDD an entire wave worth of mechs +1...and doing it using nothing heavier than 65 tons just to see if it was possible)

and then I'll turn around and do something like this while trying to make a new (and silly) build work like a light gauss high speed Ghille:
https://steamuserima...3749171F819DE5/
(so painful)

until I actually did get the little tweaks figured out and then how to pilot it so it all came together:
https://steamuserima...250B80685B3AA5/
(sadly, not enough to carry the team tho, and it highlighted the problem with not being able to secure my own kills before my team steals them when using a light gauss at range *smirk*)

Oh, and if you're feeling froggy enough to want to see if you can get a light gauss Ghillie above 400 or so damage, here's the build:
https://steamuserima...6163D644890A4A/
I wish you luck. It's badly undergunned and as delicate as a rose petal in a meat grinder.

Those stats mean nothing to what we're talking about. You said the leaderboard was random - if it's random then it would be random every month for everyone.

I proved to you that it's not. You are wrong and the data shows you're wrong. When pushed to provide leaderboard data that is clearly random, you give screenshots of anecdotal matches, which means nothing at all.

Where is leaderboard data that is random.

On the matchmaker you don't seem to understand that there's a difference between your Elo score (or whatever the system is) and your win/loss. Elo is based on your win loss. It is adjusted based on you winning or losing. However the only metric it uses is win/loss. That you linked in someones personal Elo system where he gives a modifier for how MUCH the win or loss was doesn't mean anything - it's based off win/loss. Your score is not based on anything but win/loss. He may modify how much your score changes by but it's driven by only win/loss.

[Honeybadgers]

Sam Ael, on 25 November 2017 - 10:36 AM, said:

FP Leaderboard says OP played 35 matches. QP Leaderboard don't know him. Seems he's just a forum troll.

Not to piss on your cornflakes, but that does nothing to invalidate his opinion.

You're technically more of a forum troll for trying to find an argumentum ad hominem

[Xiphias]

Xavori, on 28 November 2017 - 02:27 PM, said:

MischiefSC, on 28 November 2017 - 02:12 PM, said:

That didn't work, all it did was add mediocres and bads into gameplay with good players, resulting in the bads getting farmed. Most matches in T1 are more like an 8 v 8 or 6 v 6 with farmables filling in. This is frustrating for everyone. The bads get thrown up to T1 where they settle into a sub 1.0 w/l and get destroyed match after match after match, rarely getting kills or decent damage.
The good players have to play every match in hardcore carry mode because they've got a bunch of dead weight stacked on them.

I actually agree with everything there except the last sentence cuz I don't do that.

Yeah, he said the good players do that.

Xavori, on 28 November 2017 - 10:52 AM, said:

My position is absolutely rational.

Everything I've said in this thread is based on reality and understanding of numbers.

Ahem.

Xavori, on 27 November 2017 - 04:18 AM, said:

I'm the guy with a zamboni who thinks the entire world is cartoon characters covered in chocolate sauce doing the macarena while singing out tech manuals as lyrics to old show tunes.

Indeed, all based in your own special reality.

Xavori, on 28 November 2017 - 03:58 PM, said:

and then I'll turn around and do something like this while trying to make a new (and silly) build work like a light gauss high speed Ghille:
https://steamuserima...3749171F819DE5/
(so painful)

until I actually did get the little tweaks figured out and then how to pilot it so it all came together:
https://steamuserima...250B80685B3AA5/
(sadly, not enough to carry the team tho, and it highlighted the problem with not being able to secure my own kills before my team steals them when using a light gauss at range *smirk*)

Oh, and if you're feeling froggy enough to want to see if you can get a light gauss Ghillie above 400 or so damage, here's the build:
https://steamuserima...6163D644890A4A/
I wish you luck. It's badly undergunned and as delicate as a rose petal in a meat grinder.

Well I don't have any Enforcers and I'm certainly not going to buy a hero for this, but the Cicada 3M can run the same loadout so I decided I'd give it a run with that instead.

I altered your build to this:


I ditched some of the ammo for extra lasers to make it a better build. Other than that it's pretty much the same.

How did it perform?

Spoiler

First match wasn't great, I was running a bit too hot to dish out the damage I wanted so I tweaked the skill tree. I also only had two lasers in this match, but I upped it to three by dropping a ton of ammo.


That's more like it, 4 kills, 4 kmdd, 3 solo kills, and 865 damage. Not bad for my second testing match ever for this build.


Another 800+ damage game, 2 out of 3's not bad. Also, there's the win I was looking for.

Summary of results:

Spoiler

Over 6 matches I had 9 kills, 12 kmdd, 5 solo kills, and 3902 damage. That gives an average of 1.5 kill rate, 2 kmdd, 0.67 solo, 2.25 KDR, 1 WLR, 650 damage and average matchscore of 475.

That's not too far off from my usual stats. With more matches to even it out I'm sure it would end up pretty close. I took a "bad" build that I'd never run before and it took me a total of 1 match to optimize the weapon load out and skill tree for it.

Album

I also did all this in T1 instead of in T3 though I know you won't think that matters. My point is that you aren't nearly as good of a player as you think you are and your stats show it.

There's a reason no one wants to listen to you Xavori, it's because you've refused to listen at all to the multiple explanations given to you by people who actually understand statistics and you've moved the goal posts so many times I've lost count. It's full on Dunning-Kruger. You just keep repeating the same meaningless catchphrases like "Random numbers are random."

Do you believe a player can influence their WLR in a 1v1? What about a 2v2? What number of players do you draw the line at? Please explain the mathematical basis you have for choosing this number.

You aren't taken seriously because you can't back up your claims. You could prove the leader boards are random by showing a massive swing in stats for players from season to season. You won't because you can't.

You can prove that the stats are a mockery by farming your stats for a week in the next leader board season. If you succeed I'll admit I was wrong and apologize.

Also, to add to Mischief's data, my win percentages since season 2

Spoiler

64.7% 2 August
69.3% 3 Sept
61.3% 4 Oct
64.3% 5 Nov
69.7% 6 Dec
67.4% 7 Jan
64.5% 8 Feb (Leveling all the bad lights before the skill tree)
73.5% 9 March (Stat farming month, Yay!)
69.3% 10 April
67.1% 11 May
69.9% 12 June
60.6%* 13 July
63.4% 14 Aug
72.4% 15 Sept (Pure solo)
70.8% 16 Oct (Pure solo)
77.0%* 17 Nov (95% solo)

*Fewer than 100 matches

Overall, they stay withing a reasonable range. The biggest jump month to month is 9% when I went from constantly leveling bad mechs to directly farming my stats to see what I could achieve.

[arcana75]

Just to wade in, I'm a new player, and have stats for only 4 seasons:

S14: WL 33% 12 games
S15: 53% 215 games
S16: 49% 299 games
S17: 42% 102 games

If I discard S14 and S17, then just looking at S15-16, the percentage is well within the region of 50% win. For this to happen, it's not random, something is going on to make it near 50%. There's enough games there to eliminate personal and team skill, and accord most of the credit to the matchmaker.

Another data point which to me sheds more light onto my own personal skill, are KD ratio and average match score:

S15: 1.51 KD ratio, 193/128 K/D, 263 avg score
S16: 1.44, 271/188, 269

These data points suggest a more consistent behaviour on my own part, once you contextualise them to the earlier data points about my WL ratio at near 1:1 (+10% deviation) over 500 matches (FP and QP). In other words, despite the matchmaker creating a scenario of 50% WL, I am still able to achieve a +50% higher than 1:1 KD ratio deviation, with match scores that suggest average damage values of at least 500-530 points (250-265 match score) which by some yardsticks are respectable higher than average values. I also pilot all kinds of mechs, and every mech starts out at 0 skill nodes activated because each time I fully skill a mech, I tend not to continue playing that mech but rather try a new mech cuz my stable is small and I want to try different mechs, but with over 500 games, the mech itself (weight class, loadout, number of skillnodes activated, etc) starts to average out and become less meaningful a data point for meaningful consideration.

Just my 2 bits.

Edited by arcana75, 28 November 2017 - 10:19 PM.

[Xiphias]

arcana75, on 28 November 2017 - 10:04 PM, said:

Just to wade in, I'm a new player, and have stats for only 4 seasons:

S14: WL 33% 12 games
S15: 53% 215 games
S16: 49% 299 games
S17: 42% 102 games

If I discard S14 and S17, then just looking at S15-16, the percentage is well within the region of 50% win. For this to happen, it's not random, something is going on to make it near 50%. There's enough games there to eliminate personal and team skill, and accord most of the credit to the matchmaker.

I think your stats do a good job of illustrating how the matchmaker works pretty well at balancing things in the lower tiers. I'm assuming you hit T3 recently and that's why your WLR is lower for the most recent season. As you improve it will probably move up to around 50% again and then that will cause your PSR to get boosted until you hit the next tier, where your WLR will probably drop again as you play against harder players more regularly.

The problem with PSR is primarily at the top and the bottom. In T1 there is a huge skill gap between the top players like Proton and the okay players who just got into T1 through attrition. You can be in T1 with a WLR < 1. You can also achieve a WLR significantly greater than 1 while solo dropping. The reason for that is that the MM treats all T1s the same and if you're on the upper limit of that you can win more matches than you lose.

If tiers were more granular in how they are divided (say 10 tiers) you would expect closer matches and for players WLRs to approach close to 1 (for solo matches). The fact that I can maintain a WLR > 2 while solo dropping shows that T1 has a large skill gap. If everyone was playing at the same level WLR would be much closer to 1.

If there was no match maker at all then WLR would directly relate to how good a player was at helping win a game. With the PSR system moving up results in getting harder opponents which will work to force a player's WLR down towards 1. This works pretty well until T1, bad players move down, average players stay about the same or maybe bounce between two tiers, and good players should move up relatively quickly. After hitting T1 it's effectively back to being random matching. Even if a player has a WLR of 10 their isn't a higher tier to push them into. Since T1 players are considered equal (even if they aren't) whether you are matched with/against a good/bad T1 player is random.

Because it's random the matchmaker can't force your WLR towards 1, instead the better T1 players will end up getting better WLRs while the "bad" T1 players will have lower WLRs. A "bad" T1 player (WLR<1) might be a better player than a "good" T5 player (WLR>2), but when that T5 player gets to the higher tiers their WLR will start to fall unless they improve. Once you're in T1, solo WLR is a pretty good indicator of how good a player is.

Your WLR being at 50% means that you are playing most of your matches with players at your skill level. The fact that my WLR is 60-70% means that for most of the matches I'm in the average skill is below my level and I see that in game. The matchmaker can't force my WLR down because it only tracks my tier and the best thing it can throw at me is a T1 player. That player could be proton who would probably walk all over me or it could be a bottom of the barrel T1 player that I could stomp. Since my ability is higher than the average in T1 the probability of getting a "bad" player is higher than getting a "good" player.

E.g. The matchmaker picks a random T1 player from a pool of 100 players. Since I'm better than 70% of those players there is a 70% chance of getting a "bad" player and only a 30% chance of getting a "good" player. Over a lot of matches this will result in me winning more than I lose inflating my WLR. A player like Proton (99%) would still win more than me and would have an even higher WLR. Obviously, in practice it's more complicated than this, but the general principle is the same.

MM will try to force a player to a WLR of 1 until they get to T1. Once they get there it can't keep the good players from farming the bad ones.

[Jay Leon Hart]

Before I do my leaderboard stat dump, a little bit of my MWO stat history.

Before PSR dropped, I had a W/L & K/D each between 0.7-0/8.

Once PSR dropped, I was about 2/3-3/4 through Tier 5.

I have played maybe 10 matches in Group over all 17 Seasons, only for event rewards.

Here are Seasons 1 through 17;
S1 0.68 W/L, 0.94 K/D, 179 Avg. MS, 40% wins, 89 matches
S2 1.06 W/L, 1.21 K/D, 186 Avg. MS, 52% wins, 165 matches
S3 0.91 W/L, 0.75 K/D, 177 Avg. MS, 48% wins, 225 matches
S4 1.16 W/L, 1.41 K/D, 216 Avg. MS, 54% wins, 203 matches
S5 1.28 W/L, 1.35 K/D, 223 Avg. MS, 56% wins, 217 matches
S6 1.24 W/L, 1.24 K/D, 215 Avg. MS, 55% wins, 110 matches
S7 1.36 W/L, 1.44 K/D, 259 Avg. MS, 58% wins, 52 matches
S8 1.37 W/L, 1.80 K/D, 224 Avg. MS, 58% wins, 143 matches
S9 1.27 W/L, 1.42 K/D, 218 Avg. MS, 56% wins, 76 matches
S10 0.89 W/L, 1.26 K/D, 226 Avg. MS, 47% wins, 104 matches
S11 1.10 W/L, 1.45 K/D, 249 Avg. MS, 52% wins, 165 matches
S12 0.95 W/L, 1.56 K/D, 276 Avg. MS, 49% wins, 119 matches
S13 1.02 W/L, 1.21 K/D, 253 Avg. MS, 50% wins, 121 matches
S14 0.93 W/L, 1.29 K/D, 256 Avg. MS, 48% wins, 110 matches
S15 0.93 W/L, 1.19 K/D, 251 Avg. MS, 48% wins, 104 matches
S16 0.64 W/L, 1.30 K/D, 287 Avg. MS, 39% wins, 100 matches
S17 1.38 W/L, 1.90 K/D, 320 Avg. MS, 58% wins, 62 matches

So, a shaky start on win %, then a bit of stability in the mid-high 50s.
Then a drop at Season 10, but my Avg. MS went up (probably when I stopped pushing as much).
More stability, then a massive drop in Season 16, with another jump in Avg. MS (more selfish play).
Then jumps back to the mid-high 50s in Season 17 (where I've only been playing Clan laser vomit 'mechs, with 34 of those 62 matches in a 2HLL, 6ERML SNV-1).

Win % is pretty consistent between 48%-58%.

Considering I'm 3/4 through Tier 2 now, I have (in theory) been facing better and better opponents while improving the way I play (though apparently, not helping my team as much as I used to).

Now, I never got past AS-Level Statistics, but those numbers don't look random to me.

[arcana75]

All things considered, given a high enough number of games say 100, average match score (AMS) seems to give a better indication of player skill than kills or wins. K/D or W/L ratios seem meaningless, even more so without the actual numbers eg 2.0 W/L ratio over 3 games vs over 100 games.

If so... 2 questions:

1) Is MM taking into account AMS?

2) Should the Tiering be based on AMS? Eg Tier 1 a minimum of 400 AMS across an entire season, 300 for Tier 2, etc. Therefore Tiering doesn't happen on a per-match incremental basis, but you stay in a Tier throughout an entire Season, until the end of a Season where the AMS is checked and you get re-Tiered. Big problem though, AMS doesn't factor in number of matches. If we do want to factor in matches, then perhaps the number of matches would demarcate your Divisional status? Eg

Div 1 (top): at least 400 matches
Tier 1: 400ms
Tier 2: 300ms
etc

Div 2: at least 300 matches
Tier 1: 400ms
Tier 2: 300ms
etc

and so on. This way, at least it'll cull the so-called potatoes who are now in Tier 1 and supposedly have no right to be there. With such a new laddering system, those who play maybe a handful of games, and get continually destroyed in Tier 1, will now be pushed into a lower Division, in a lower Tier. Whereas the top competitive players, will be in Div 1 and be in Tier 1 or 2 due to consistently good performance.

Edited by arcana75, 29 November 2017 - 02:32 AM.

[Xavori]

arcana75, on 29 November 2017 - 02:20 AM, said:

All things considered, given a high enough number of games say 100, average match score (AMS) seems to give a better indication of player skill than kills or wins. K/D or W/L ratios seem meaningless, even more so without the actual numbers eg 2.0 W/L ratio over 3 games vs over 100 games.

If so... 2 questions:

1) Is MM taking into account AMS?

2) Should the Tiering be based on AMS? Eg Tier 1 a minimum of 400 AMS across an entire season, 300 for Tier 2, etc. Therefore Tiering doesn't happen on a per-match incremental basis, but you stay in a Tier throughout an entire Season, until the end of a Season where the AMS is checked and you get re-Tiered. Big problem though, AMS doesn't factor in number of matches. If we do want to factor in matches, then perhaps the number of matches would demarcate your Divisional status? Eg

Div 1 (top): at least 400 matches
Tier 1: 400ms
Tier 2: 300ms
etc

Div 2: at least 300 matches
Tier 1: 400ms
Tier 2: 300ms
etc

and so on. This way, at least it'll cull the so-called potatoes who are now in Tier 1 and supposedly have no right to be there. With such a new laddering system, those who play maybe a handful of games, and get continually destroyed in Tier 1, will now be pushed into a lower Division, in a lower Tier. Whereas the top competitive players, will be in Div 1 and be in Tier 1 or 2 due to consistently good performance.

You cannot just pull a number out of thin air, 100, and think it's going to produce meaningful results just because it seems like a lot.

AMS would be better than W/L ratio (because in an idea setting, W/L ratio ~1 no matter how desperately Misch and Xephias and others want to believe otherwise), but it's still faulty in that the current AMS too heavily relies on damage output and almost completely ignores winning and mission objectives. It's why when match score events come up, a lot of players, including myself, start doing "bad" thing like streak boating to farm damage even tho streak boating is pretty weak for anything but damage farming.

Here's the irony of this discussion, I'm 100% on board with the idea that winning and losing could mean something. I just absolutely reject the idea that in the current environment they do because the matchmaker, PSR, and therefore leaderboards, are all garbage.

Xephias,

You cheated on your light gauss build. You put too many lasers on it. Pretty sure I could run a light with nothing but 3xERML's and put up your numbers

Oh, and I'm well aware of the limitations of my own knowledge, so not DunnyK. Math, however, is right in my wheelhouse. It was kinda a big deal back at West Point which while a military academy, is also an engineering school.

Misch,

I find it fascinating that in one moment, you can acknowledge that ELO (I'm totally going to keep doing this now) and TrueSkill would be better than our matchmaker, but then insist that W/L on our current leaderboard means something. I know you want to believe that it shows that you are a good pilot. Maybe you are. BUT YOU CANNOT USE ANYTHING ON OUR CURRENT LEADERBOARD AS EVIDENCE OF THAT.

As long as the matchmaker could be giving you potatoes to mash, you can't say anything about any of the numbers on the leaderboard. If you want meaningful stats, you have to play quality matches almost exclusively.

What you are arguing in favor of is would be like a polling company trying to find out if the US government should cut food stamps. They use the ME formula, industry accepted standards for writing the question, and sample size, but then only poll at homeless shelters. Sure, it looks all math-y and right and stuff, but the results would be garbage.

MWO produces random matches. Random matches give you random results. It doesn't matter if you think that leaderboard stats look consistent because you cannot say with certainty why that is. In fact, the stat y'all keep going to, W/L ratio, is the most obvious stat to look at how broken the MWO matchmaker is. That number should approach 1 over time if you are playing quality matches. It's why no rating system gives a crap about W/L ratio because yes, if you play bad players all the time you'll win more than lose.

Of course, MWO's team interdependency, high number of variables, and random matchmaking means you cannot even be sure you're playing against bad players more often than good. You might just be getting lucky matches with otherwise good players playing bad mechs or making their rare mistakes. You might only be logging on when the good players aren't. You might only play in heavies at times when a lot of other players are playing lights. And so on. And because there are so many perfectly valid possibilities for W/L ratios above 1 in a random environment, you cannot say with any certainty that, "Nope! It means I'm a good player!"

Maybe instead of desperately clinging to "W/L ratio means something!" you should try joining me in pointing out that the leaderboard is garbage, the matchmaker is garbage, and see if we can't convince PGI to give us a better system that produces better matches.

Cuz frankly, 12-3 stomps get old really quick.

[The Basilisk]

FupDup, on 24 November 2017 - 09:44 PM, said:

"Scouting" mode being almost entirely ruled by SRM-brawling medium mechs is just a natural byproduct of the mode's design (i.e. 4v4 players in fast mechs). Like the rest of CW, it really shouldn't be taken seriously.

Actually...no it is a byproduct of how SRMs work and in addition to smal weapons convergence aka "the boatin issue" an other showcase of how oversimplyfied mechanics can ruin a game.

[Asym]

I agree. In the past few weeks, a whole lot of players have been experimenting in "meta" builds for scouting: what is the optinmal tool for the work...

We've seen, and I am not an expert on MWO meta, laser vomit, long and short range missle vomit, and now, ballastic vomit in a repeating cycle... Of late, super fast teams where it's all about initiating that dropship return as far away from the LZ as possible (and, of course, that sometimes backfires.....) and running a 3 vrs 4 retrograde. If a game has patterns that are constricted by just about anything, you get "gamers" who find those boundries and exploit them....

What is sad in Scouting is that PGI missed the entire point of Scouting and now, it is just a 4x4 skirmish where brawling is the deliverable and brawling may be the point of an arcade FPS, but, in the MW histories, Scouting had a purpose and it wasn't to brawl vastly heavier medium mechs.........

Edited by Asym, 29 November 2017 - 07:22 AM.

[Xavori]

Asym, on 29 November 2017 - 07:21 AM, said:

I agree. In the past few weeks, a whole lot of players have been experimenting in "meta" builds for scouting: what is the optinmal tool for the work...

In my case, it's by grinding up new mechs that I'd saved up MC to buy as well as my Thanatos mechs (talk about seriously painful to grind), so my stats are all over the place. At least until I decided I want to be a good pilot again (it's almost like magic) and grab a Whammy or Hellbringer to go splatter people.

Quote

We've seen, and I am not an expert on MWO meta, laser vomit, long and short range missle vomit, and now, ballastic vomit in a repeating cycle... Of late, super fast teams where it's all about initiating that dropship return as far away from the LZ as possible (and, of course, that sometimes backfires.....) and running a 3 vrs 4 retrograde. If a game has patterns that are constricted by just about anything, you get "gamers" who find those boundries and exploit them....

What is sad in Scouting is that PGI missed the entire point of Scouting and now, it is just a 4x4 skirmish where brawling is the deliverable and brawling may be the point of an arcade FPS, but, in the MW histories, Scouting had a purpose and it wasn't to brawl vastly heavier medium mechs.........

Scouting would be a huge thread in of itself. If only someone hadn't threadjacked the most recent attempt to talk about it...

[Xiphias]

Xavori, on 29 November 2017 - 03:20 AM, said:

You cheated on your light gauss build. You put too many lasers on it. Pretty sure I could run a light with nothing but 3xERML's and put up your numbers

Because I built it better than you and didn't intentionally cripple myself? You could run the same thing on your mech if you chose to but you decided no to in finding your "optimal build". I had too much Gauss ammo with only 1-2 lasers. I would use up all or nearly all of my ammo with my build as it stands, that's efficient building. Also, please feel free to demonstrate your ability to average 650 damage using only 3ERMLs. I'll wait.

Quote

Oh, and I'm well aware of the limitations of my own knowledge, so not DunnyK. Math, however, is right in my wheelhouse. It was kinda a big deal back at West Point which while a military academy, is also an engineering school.

Clearly, statistics is not in your wheel house since you've demonstrated a lack of understanding. Do you think you're the only person who has a background in math and engineering? If you did go to West Point the only thing you've done by bringing it up is lower my opinion of the school.

At this point I'm pretty convinced that you're just trolling, but I'll give it one more shot.

Back to my question. At what number of players higher than 1v1 does individual performance no longer have any effect on WLR in MWO and what is your mathematical basis for choosing this number?

[Xavori]

Xiphias, on 29 November 2017 - 09:14 AM, said:

Because I built it better than you and didn't intentionally cripple myself? You could run the same thing on your mech if you chose to but you decided no to in finding your "optimal build". I had too much Gauss ammo with only 1-2 lasers. I would use up all or nearly all of my ammo with my build as it stands, that's efficient building. Also, please feel free to demonstrate your ability to average 650 damage using only 3ERMLs. I'll wait.

I wasn't trying to cripple myself. I was trying to see if light gauss was a viable weapon. It sorta is, but not on a medium mech where you have to make too many other changes to make it work, especially in game play where I'm 100% sure I was getting my teams killed taking the time to find sniper perches.

Quote

Clearly, statistics is not in your wheel house since you've demonstrated a lack of understanding. Do you think you're the only person who has a background in math and engineering? If you did go to West Point the only thing you've done by bringing it up is lower my opinion of the school.

At this point I'm pretty convinced that you're just trolling, but I'll give it one more shot.

You keep saying this even tho you haven't once pointed out an error in my understanding of stats. In fact, I've pointed out quite a few in other people, not the least of which is that ELO and TrueSkill ultimately aim to find a point where W/L ratio =~1. If you disagree with that statement, it's you, not me, that has a problem with math and statistics.

Quote

Back to my question. At what number of players higher than 1v1 does individual performance no longer have any effect on WLR in MWO and what is your mathematical basis for choosing this number?

Individuals always have some effect on winning or losing, even if we went all the way up to 100,000,000 v 100,000,000 scrums where the individual impact would be miniscule, it would still be there.

I've not once disagreed that individuals have an impact. What I disagree with is that you can read anything into the stats we are given and try to relate them to an individual. It's entirely possible, hell, I think it's likely that Misch is a better than average pilot. But I cannot KNOW that because I have no solid data that would confirm it. It's specifically because I do know how stats work and can be manipulated that I don't take any of the stats we have access to seriously.

I don't think pilot skill in MWO is actually quantifiable, so I don't think you can try to do any kind of stats analysis on it. And while Misch is right in that there are limits to human performance, it doesn't matter because I could create a scale of -100,000 to 100,000 and have it be every bit as valid as a scale of 0 to 1000 because the numbers I'd have to use would be arbitrary. The only difference would be precision. What's more, I'm beyond sure that the 100 matches value Misch pulled out of his posterior is wrong. It'd take a lot more than that even if we were using some form of ELO or TrueSkill to get to something we could use to compare pilots against each other (which we can do even if we can't quantify pilot skill objectively).

So short version: Individual effort matters, but you cannot know anything about a pilot based on leaderboard stats beyond that a pilot plays on good, average, or bad teams most of the time. If we used an ELO or TrueSkill system, we'd be using that ranking to compare pilots, not W/L, and we'd also have a tool available to the matchmaker to make better matches instead of the random garbage we have now.

[MischiefSC]

Xavori, on 29 November 2017 - 09:09 AM, said:

In my case, it's by grinding up new mechs that I'd saved up MC to buy as well as my Thanatos mechs (talk about seriously painful to grind), so my stats are all over the place. At least until I decided I want to be a good pilot again (it's almost like magic) and grab a Whammy or Hellbringer to go splatter people.



Scouting would be a huge thread in of itself. If only someone hadn't threadjacked the most recent attempt to talk about it...

The problem with scouting is that some people really want it to play a certain eay, others want it to play another. It's a stompy shooty robbits game, most people will want to shoot stompy robbits. Some people want wcoutinf to be a sort of non-combat thing. It can be - that's smoke diving. However the only reason that is fun at all is that the other team is trying to kill you. It's sorta pointless if you're trying to gather Intel and get out if the other team isn't trying to shoot you. However some people don't want to smoke dive, they want to shoot robbits. So they just gear for a fight.

Honestly scouting is probably the most effective game mode we have for what it's trying to do. You can just run objectives or you can fight, either works about as well. Balance is probably better than regular invasion. Just that a lot of people are not interested in mediums 4 v 4 matches. Some are but not all.

[Requiemking]

MischiefSC, on 29 November 2017 - 09:57 AM, said:

The problem with scouting is that some people really want it to play a certain eay, others want it to play another. It's a stompy shooty robbits game, most people will want to shoot stompy robbits. Some people want wcoutinf to be a sort of non-combat thing. It can be - that's smoke diving. However the only reason that is fun at all is that the other team is trying to kill you. It's sorta pointless if you're trying to gather Intel and get out if the other team isn't trying to shoot you. However some people don't want to smoke dive, they want to shoot robbits. So they just gear for a fight.

Honestly scouting is probably the most effective game mode we have for what it's trying to do. You can just run objectives or you can fight, either works about as well. Balance is probably better than regular invasion. Just that a lot of people are not interested in mediums 4 v 4 matches. Some are but not all.

Here is the problem with Scouting(as is also the issue with pretty much all the Objective-based modes). The Objective(you know, that thing we were sent here to do) pays too little. Scouting was pretty much designed to be an asymmetrical mode where the team with the combat based objectives were the defenders and the attackers had solely objective-based gameplay. The intent was that the scouts(or attackers) would race about the map, dodging, weaving, and gathering intel, all while trying to avoid combat entirely. The Defenders were to kill them, and finally Lights would have a chance to properly shine. However, PGI(being PGI), made the mistake of not altering the score system in the slightest, so even for the attackers, the Objective payed too little to be worthwhile. Thus, we arrive at our current dilemma. Scouting has devolved into 4V4 Brawl, and Lights get shafted again.

[MischiefSC]

Requiemking, on 29 November 2017 - 11:10 AM, said:

Here is the problem with Scouting(as is also the issue with pretty much all the Objective-based modes). The Objective(you know, that thing we were sent here to do) pays too little. Scouting was pretty much designed to be an asymmetrical mode where the team with the combat based objectives were the defenders and the attackers had solely objective-based gameplay. The intent was that the scouts(or attackers) would race about the map, dodging, weaving, and gathering intel, all while trying to avoid combat entirely. The Defenders were to kill them, and finally Lights would have a chance to properly shine. However, PGI(being PGI), made the mistake of not altering the score system in the slightest, so even for the attackers, the Objective payed too little to be worthwhile. Thus, we arrive at our current dilemma. Scouting has devolved into 4V4 Brawl, and Lights get shafted again.

All that would happen if it paid more to smoke dive is even fewer people would play it.

Lights shine a lot if played well. They swing matches as much or more than any other class. However a bug hunt game mode is of 0 interest to me on either side. Hence why so many people complain about smoke diving in scouting. If it was all about smoke diving you'd rarely get a match. Most the people who play it now are just there for 4 v 4 mediums/lights skirmish. They wouldn't stay for some dive gameplay.

The problem is that you're talking about something that's a small niche of an already niche segment. I wouldn't care either way - I don't scout much noe, I would just ignore it completely at that point. The issue is that you'd just have almost nobody playing scouting.

[Xavori]

MischiefSC, on 29 November 2017 - 11:34 AM, said:

All that would happen if it paid more to smoke dive is even fewer people would play it.

Lights shine a lot if played well. They swing matches as much or more than any other class. However a bug hunt game mode is of 0 interest to me on either side. Hence why so many people complain about smoke diving in scouting. If it was all about smoke diving you'd rarely get a match. Most the people who play it now are just there for 4 v 4 mediums/lights skirmish. They wouldn't stay for some dive gameplay.

The problem is that you're talking about something that's a small niche of an already niche segment. I wouldn't care either way - I don't scout much noe, I would just ignore it completely at that point. The issue is that you'd just have almost nobody playing scouting.

If it wasn't near impossible to find an ECM/Stealth light mech, I suspect it'd be more interesting to try to defend against it. But the reality is if you're on gather on Polar Highlands and you have a fast ECM mech, you can pretty much guarantee yourself a win if that's what you want. And guaranteed wins aren't fun.

So ya, smoke diving is boring even if it pays well which means if it paid well, players would quit playing that mode.

[arcana75]

Scouting suffers from the same fundamental issue as many of the map modes that aren't Skirmish: Not enough incentive to play the objective, and everything just devolves into Skirmish. It's such a bare truth today that most players can only just joke about it.

The really simple patch is to make playing the objective so ridiculously rewarding compared to other objectives.

Scouting: If a kill gives you 10k cbills (example), then 1 intel should be 20k, and escaping is an added 50k.
Conquest: Pay 1k cbills for every 1 point of Conquest difference between the 2 scores. If the score is 750 vs 500, the winner gets a bonus 250k (250x1k) cbills.
Assault: Capturing the base rewards 500k cbills.
Incursion: Each base asset destroyed 100k cbills, each asset charged 50k.
Escort: A successful escort or escort kill 250k

Something like that. Maybe double everything I said.

[Spheroid]

If people want fun non-combat gameplay that doesn't bore one side or the other they need to implement competent A.I. A scouting mode against bots or a horde mode against planetary militia could be exactly what people desire. Fun non-competitive coop gameplay that is not cut throat. The cumulative affect of these bot matches could factor in the larger strategic war in FP and pull in more of the player base.

However this game and its dying game engine have one foot in the grave. I don't even think a full time programmer is on staff. If you know otherwise name this individual. I am not talking back end stuff, actual mechanics and gameplay code which has not been updated in years.

Edited by Spheroid, 29 November 2017 - 08:26 PM.

[Invictus XVII]

in my experience in scouting, flamers and balistics wins the race, they cant shoot if theyre hot - clans in this case are already hot by default, they might try to counter with flamers on their own but IS balistics are cool enough to take that head on - clan balistics are not.