196 replies to this topic

[Solahma]

It's a tool that requires interpretation like any other tool in this game. Each stat could reflect a variety of things. What matters is that none of it matters past some storage of data with no context. The numbers for your own stats is the most important, because you know the context. You can rationalize your own stats. That's about it. The rest is very ambiguous, but it's what we have. Combining Group and Solo is enough to make any stat comparison meaningless, unless you have more data outside of the leader board itself.

That said, there is a specific formula and you could technically play to it and advance. Simply performing well shows results. That's the most we can ask for.

I use a spreadsheet I created to look at mech and build stats to the hundredth decimal. It's fun to me. But I realize that the numbers don't mean everything. It doesn't account for mech geometry, hardpoint locations, speed/mobility and how that is all affected based on the map and situation. It requires context to mean anything useful past specific simulated conditions that you can strive to replicate in a live match.

So what's the big deal? I like the Jarl's list because it paints a fairly clear picture while not being the full story. I can quickly see who is good and who is bad. I can make myself feel better by validating my opinions of a person's performance capabilities with some relative accuracy.

Like any data, you must know how to interpret it.

[Y E O N N E]

Teer Kerensky, on 23 March 2018 - 09:28 AM, said:

He is right too. Ingame it displays season 20 and ends April 1st.

But on webpage current season is 21.
https://mwomercs.com...le/leaderboards

Der Geisterbaer, on 23 March 2018 - 09:26 AM, said:

You might want to open the Leaderboard in the client itself and be surprised to find that it indeed says SEASON 20 (ENDS ON APRIL 1) at the top

He can be right about what it says, doesn't mean he's right about what it actually is.

Guess which one actually matters?

[RustyBolts]

The one that should matter is the in game one, which all players see and not the forum one which only some players see. But hey. PGI. lol

[Solahma]

The in-game leaderboard?
oooh boy

[Y E O N N E]

RustyBolts, on 23 March 2018 - 10:09 AM, said:

The one that should matter is the in game one, which all players see and not the forum one which only some players see. But hey. PGI. lol

No. From the perspective of collecting stats, which is what we are discussing, what matters is what it is or you are dropping an entire season's worth of data.

[RustyBolts]

Solahma, Thanks. That is the best description yet. I can see where it would be beneficial then.

Yeonne, I was referring to the in game saying 20 instead of 21. I should have been clearer. All players see that vice the forum season labels. If any labels should be correct, that is the one since all players see it. To be honest I didn't even know the forum one existed.

[Mystere]

Yeonne Greene, on 23 March 2018 - 10:13 AM, said:

No. From the perspective of collecting stats, which is what we are discussing, what matters is what it is or you are dropping an entire season's worth of data.

Hmm. I thought what we were discussing -- or were supposed to be discussing anyway -- was the OP's question, which just happens to have been asked by no other than RustyBolts?

[Y E O N N E]

Mystere, on 23 March 2018 - 10:21 AM, said:

Hmm. I thought what we were discussing -- or were supposed to be discussing anyway -- was the OP's question, which just happens to have been asked by no other than RustyBolts?

The OP's question was merely "What am I missing?"

[Y E O N N E]

RustyBolts, on 23 March 2018 - 10:18 AM, said:

Solahma, Thanks. That is the best description yet. I can see where it would be beneficial then.

Yeonne, I was referring to the in game saying 20 instead of 21. I should have been clearer. All players see that vice the forum season labels. If any labels should be correct, that is the one since all players see it. To be honest I didn't even know the forum one existed.

Like Bombast said: little details.

Almost sounds like there's an integer flub in the code for the in-game UI.

[RustyBolts]

@ Bombast, so I didn't agree with the post about WLR being a good indicator is not an opinion that is counter to what others have said? I actually thought the conversation was going well and non-confrontational until he got his undees tied in a knot because I didn't agree with a couple comments.

[Zergling]

I've done a heap of checking with the leaderboard stats, adding up all my individual battle stats over many battles and comparing to leaderboard and profile stats page totals, and they have always matched.

If there are errors, they are quite rare because I never detected one in hundreds of battles of data.

[RustyBolts]

@ Zergling, I was using the in game leaderboard which shows season 20 and comparing it to Jarls which also said season 20. I didn't even know that there was a Forum board which stated 21. So my comment about the stats not matching was based on the In game vs Jarls. Now it makes sense why the stats don't match.

But back to the original question of the post, Jarls is really just a repository of past seasons with a user made formula to try and provide some kind of ranking since the Tier system is obviously not correct?

[Zergling]

RustyBolts, on 23 March 2018 - 10:50 AM, said:

@ Zergling, I was using the in game leaderboard which shows season 20 and comparing it to Jarls which also said season 20. I didn't even know that there was a Forum board which stated 21. So my comment about the stats not matching was based on the In game vs Jarls. Now it makes sense why the stats don't match.

But back to the original question of the post, Jarls is really just a repository of past seasons with a user made formula to try and provide some kind of ranking since the Tier system is obviously not correct?

Yeah, that post wasn't directed at you; there's a few other people claiming that there's errors in the stats database which somehow invalidate everything.

[process]

RustyBolts, on 23 March 2018 - 10:50 AM, said:

But back to the original question of the post, Jarls is really just a repository of past seasons with a user made formula to try and provide some kind of ranking since the Tier system is obviously not correct?

More or less. The list is just an index. It's attempting to turn the mountain of player stat data that PGI publishes into an accessible ranking system. It is generally a good way to quickly observe and compare player performance, but it's not the end-all for discussion and I dislike seeing higher-ranked players use it to dismiss the opinions of lower-ranked players. It's certainly not perfect since it's dependent of the quality of the data from PGI and the subjective interpretation of such, but it's also generally valid.

The methodology for calculating rank is explained at the bottom of the list page. I'm not a statistician so I can't comment on how legitimate it is, but the written explanation seems reasonable to me:

Quote

Ranks are determined by adjusted match score for players who have been active in the last three months.

Players who have stopped playing are still tracked but are placed into "retirement". They will be placed back into ranks if they return but
their latest performance will have significant effect on their placement when returning.

How Adjusted Score Formula is calculated: ((Season Average Match Score)*(Season#^1.8))*(-(1.007^(-TotalGamesPlayed)-1))*(ClassMultiplier)
This will give full score around 500 games and each new season is weighed higher than the last. After this, the score is adjusted based on classes played.

The weight class coefficients are derived from each classes' recent global performance. The current class weights are:
Light: 1.090671842, Medium: 0.9627151535, Heavy: 0.9505204456, Assault: 0.91315593.

Progress is the progress a pilot has made in the current season compared to their average score.

Edited by process, 23 March 2018 - 11:28 AM.

[Der Geisterbaer]

[redacted]

But since I allegedly have no "proof" and you haven't observed no discrepancies between data sets ... tell me more about this:



I already said that I never set foot in Faction Play, so disregard Emerald Taiga as "inexplicable one off" if you wish to do so but tell me again about the quality of the PGI data and the subsequent quality of derived works like Jarl's list.

Edited by McValium, 24 March 2018 - 01:10 PM.
Redacted unconstructive fighting

[justcallme A S H]

Asym, on 23 March 2018 - 06:16 AM, said:

Ah, guys.....what a silly discussion. The data is nice and Jarl's does a good job using.....un-verified and wildly skewed data.
Go
I went to Jarl's just to "see for myself" what this tool produces. I am missing seasons 4 through 7, The data in season 20 is not correct. And, after retiring from corporations that have extreme reliability metrics, ah, MWO could never produce statistics that would ever make it to any of the corporate board rooms I've served in....never. It is that invalid as many in this article and a dozen other articles have laid out... A good try at something that has zero real value.....
.

The data is coming directly out of MWO. It is not unverifiable and not wild skewed.

It is 100% accurate in every cross reference check I have done.

You just need to remember it won't count under 10 games in a particular weight class because that doesn't show in the PGI data.

[Bombast]

Der Geisterbaer, on 23 March 2018 - 11:30 AM, said:

[redacted]

But since I allegedly have no "proof" and you haven't observed no discrepancies between data sets ... tell me more about this

I already said that I never set foot in Faction Play, so disregard Emerald Taiga as "inexplicable one off" if you wish to do so but tell me again about the quality of the PGI data and the subsequent quality of derived works like Jarl's list.

It's pretty well known the Profile Stats are jacked. But the leaderboard is fine. Its been proven accurate Everytime it's been tested.

Edited by McValium, 24 March 2018 - 01:12 PM.
cleaned up quote

[Zergling]

Bombast, on 23 March 2018 - 11:37 AM, said:

It's pretty well known the Profile Stats are jacked. But the leaderboard is fine. Its been proven accurate Everytime it's been tested.

Base stats and mech stats are accurate for QP at least; I've compared results for hundreds of battles and they matched perfectly.

Edited by Zergling, 23 March 2018 - 11:40 AM.

[EnochsBook]

Der Geisterbaer, on 23 March 2018 - 11:30 AM, said:

Ah, you're the one who made the "anonymous" accusation that people made claims and when you're called out on it it's the other person who is "trolling"?

But since I allegedly have no "proof" and you haven't observed no discrepancies between data sets ... tell me more about this:

Spoiler

I already said that I never set foot in Faction Play, so disregard Emerald Taiga as "inexplicable one off" if you wish to do so but tell me again about the quality of the PGI data and the subsequent quality of derived works like Jarl's list.

So the biggest incoherence is 5 extra games on Canyon Network? Do you think that's going to affect the accuracy of the Jarl's list during a season of more than 200 games much?
Also we don't know whether those extra/missing games are actually counted towards any stat.

[justcallme A S H]

GBxGhostRyder, on 23 March 2018 - 07:21 AM, said:

The data provided by PGI or any other source is false data period on player data for W/L the reason is any player above 1.2 is

#1 playing on a organized team in QP or FP to attain a high W/L record or
#2 there exploiting the game or
#3 there using hacks to enhance there gameplay.

Lol...

1. Wrong. FP doesn't count to those stats as they are not able to be scraped.
2. Wrong.
3. Totally false.

I fit into your exact 'window'... I am nothing of anything you posted.
www.twitch.tv/justcallmeash

Stop by my twitch, look at my videos [redacted]

Edited by McValium, 24 March 2018 - 01:13 PM.
redacted nonconstructive