26 replies to this topic

[Angel of Annihilation]

There are several current threads about the Victor where people are going back and forth between claiming it is a piece of junk to it being the best mech of the game yet those posts are all opinion and we have seen many others like them in regards to other mechs as well. However there is an easy way to tell how a mech is performing....the tournament results.

The reason we can use them is because they all have the same forumla as a calculation and they take into account the best 10 scores of all the partisipants which gives us a large representative sample.

Here are the top results in order of best to worst:

Assault:

Victor = 2315
Highlander = 2258
Atlas = 2213
Stalker = 2183
Awesome = 1988

Heavy:

Jagermech = 2057
Cataphract = 2022
Catapult = 1978
Quickdraw = 1954
Dragon = 1920

Medium:

Hunchback = 2068
Centurion = 2050
Blackjack = 2037
Trebuchet = 1895
Cicada = 1892

By looking at these numbers we can see that the Victor is the top performing Assault mech by a fair margin and that the Awesome is hurting badly.

We can also tell that Heavy mechs in general are pretty balanced although the Dragon is lagging a tiny bit behind.

As far as medium, we have three performing pretty well with two lagging pretty badly.

Also we can compare the weight classes and we see that mediums and heavies are actually pretty well balanced against each other contrary to popular belief with Assaults giving a fairly significant performace advantage except for the Awesome which however is balanced against the numbers produced by medium and heavy mechs.

Lights unfortunately haven't had their day to shine so no numbers yet but I think the important thing to take out of this is that it is pretty easy to see how one mech compares to another.

Additionaly I would like to point out that aside from the poor Awesome, the differences in performance between mechs of a given weight class are around 0.5 more average kills per match. This is a very slim margin indicating that most mechs are fairly well balanced against each other.

[XX Sulla XX]

I thought we talked about keeping this a secret haha.

[Tennex]

yeah this is a good general indicator

but you also have to take into account that a person's match score each game is based a large part on luck.
and since there are many many more victor players (its just been released) its likely that the victor score just has a bigger sample size. (for example if you buy many lotto tickets, the chances you will get a ticket with a high number goes up)


but at least what this tells us is that the victor is able to stand toe to toe with some established chasiss in the weight class. even if the mech play count may have skewed the data a little.

also, that awesome score is very sad awesomes always do the worst in every assault tourney

Edited by Tennex, 22 July 2013 - 06:19 PM.

[Angel of Annihilation]

Tennex, on 22 July 2013 - 06:12 PM, said:

yeah this is a good general indicator

but you also have to take into account that a person's match score each game is based a large part on luck.
and since there are many many more victor players (its just been released) its likely that the victor score just has a bigger sample size. (for example if you buy many lotto tickets, the chances you will get a ticket with a high number goes up)


but at least what this tells us is that the victor is able to stand toe to toe with some established chasiss in the weight class. even if the mech play count may have skewed the data a little.

also, that awesome score is very sad awesomes always do the worst in every assault tourney

Its not about luck. They take the 10 best matches of everyone playing though each person might play 20-100 matches each. This takes most of the luck factor out of it. If you wanted to go even further, you to add up the top 15 scores that are displayed and average them for tighter numbers but I don't think you will find the gaps in the scorew\s tightening much if at all if you do this.

Also I did some tests just to see how many people are typically in a bracket and it is usually about 1000+ no matter what. Even if there were say 2000 people playing Victors and there were only 1000 playing a Atlas, your not going to see too much score creep with that large of sample size. Also you have to consider that the Victor, unlike the other mechs isn't an established mech. Many players in the tournament, myself included, hadn't finalized their best builds, learned all its quirks or even had all thier efficiencies unlocked for all variants they played. This would tend to offset any artifical inflation of score due to more people playing if there even was one.

Try to keep in mind that these scores are based on the BEST a mech can do and on the BEST pilots

[Nauht]

While I agree that the Victor is a great mech, I think your argument has one serious flaw - the amount of people playing.

Seeing that the Victor is the current new mech there would be alot more people playing this mech allowing for a far greater pool from which you can pull a far greater set of numbers to calculate performance.

There would be far less people playing Awesomes and Dragons than Victors and Cats.

But that's just as I see it from cursory look. Some statistic nut can correct me if I'm wrong.

[XX Sulla XX]

I have to say I did not see a huge amount of Victors in my matches. There were less than the other assaults not counting the Awsome.

[Narcissistic Martyr]

I think the most important thing we learned here today is that the blackjack is better than cataphracts. Could this extremely small data set composed of the top 10 match scores from the absolute best pilots indicate the dawn of the Great Blackjack Meta? We report. You decide.

[FupDup]

Narcissistic Martyr, on 22 July 2013 - 07:06 PM, said:

I think the most important thing we learned here today is that the blackjack is better than cataphracts. Could this extremely small data set composed of the top 10 match scores from the absolute best pilots indicate the dawn of the Great Blackjack Meta? We report. You decide.

I for one would welcome a BJ meta.

[Blue Hymn]

It's one thing to hear that Awesomes are bad... but it's another to actually see the numbers from the assault tournament.
We have at least a 200 point difference against the other Assault classes, and since energy weapons are being penalized in heat currently - along with the upcoming patch on the 30th, no wonder we're doing bad.

Awesome needs some loving. ;__;

[RussianWolf]

two problems.

1)as you say this is the average of the top ten scores per chassis. The ten at the extreme range of the bell curve. This is not how you arrive at an average for a subject.

For example, Likely my best match ever was in a PB. I was last mech standing, 6 kills and had people on both teams saying that "That PB is my hero". If it happened during tournament time it would probably be included as one of your 10 and give the Awesome a serious bump, but it was NOT an average match. Everything lined up with me being "in the zone" and everyone else doing things just right to allow it to happen.

2)the pilot. You may be a great Victor pilot, I may be a great Atlas pilot. Saying one is better doesn't really prove anything. Finding ten games (in each mech) where people did better in one over the other, doesn't either.

[XX Sulla XX]

On game of a life time will not move your tournament score into the top ten best I can tell. The top 10 players have 10 of your game of a life time games. And its down to smaller differences between them. They are the 10 best out of the 1000 or so playing that mech in the tournament.

Same thing with pilot skills they are the best pilots for each mech. The best out of a large number of players. For each mech they are getting the most out of each mech. So I think it does give a good idea how good each mech is. But I would suggest there is one thing throughing it off. The Victor encouraged people who were great pilots but had never been an assault pilot to buy there first assault. Since the ELO is for each weight they would have been starting out with a new ELO ranking in assaults. So you would have a great say former heavy pilot with an new assault ELO. They would outclass those they were playing against. Although to be fare to the Victor I had 2000 or so matches in Atlases and moving to the Victor gave me a bump since it is easier to pilot than an Atlas.

[tayhimself]

How is a blackjack better than a cataphract? And my poor quickdraws are barely better than the rightfully much maligned dragons.

[Shadey99]

I should point out that as the OP says (implicitly not explicitly) some of these numbers are old (only the assault numbers are new).

I think we need some new tournaments... Because if the highest score was 1892 for a cicada and I can score ~1400 not trying in a spider, then I certainly could get higher in my Cicada builds. I wish they would give us a breakdown for the other classes from the last event (even lights). I'm sure they have those numbers since each class had a category.

[hammerreborn]

More like the top victor was the best at stomping out all the other victors running amok over the weekend tourney.

[Tremendous Upside]

Performance in tournaments means absolutely nothing in terms of whether a mech is viable - or better than another. High match scores are the product of falling into the perfect situation - and not choking when that situation presents itself. Nothing more and nothing less. Either you've managed to keep from drawing attention to yourself and cleaned up everyone else's messes for them or you've fallen into a game vs idiots that file into your line of sight single-file. That's how high match scores happen... It's not the mech. It's not the loadout - and not even the pilot that matters half the time. It's like a perfect storm. The more you play, the better your chances of landing a big game... Unless you're really really bad of course.

[Iron Frost]

or maybe it's just that better players are bored with old mechs and gravitate towards freshness?

[pulupulu]

I agree with OP mostly.

One issue I got with OP is how the scores are done in this game. Like scouting generally get less point etc, so mechs like cicada would be hurt.

Another one being that the best player probably have gave up on awesome already... it is old AND less effective, absolutely no reason to play it. I expect awesome score to continue dropping each time this score list come up.

I am a little surprised that victor is so much higher than awesome though. I know victor is better than awesome being that I got both, but those number do put it into better pespective.

Banky, on 24 July 2013 - 08:13 AM, said:

Performance in tournaments means absolutely nothing in terms of whether a mech is viable - or better than another. High match scores are the product of falling into the perfect situation - and not choking when that situation presents itself. Nothing more and nothing less. Either you've managed to keep from drawing attention to yourself and cleaned up everyone else's messes for them or you've fallen into a game vs idiots that file into your line of sight single-file. That's how high match scores happen... It's not the mech. It's not the loadout - and not even the pilot that matters half the time. It's like a perfect storm. The more you play, the better your chances of landing a big game... Unless you're really really bad of course.

I agree with this mostly as well. However what I take away from this list is that victor can force and take advantageof a good situation much better than say... an awesome...

Edited by pulupulu, 24 July 2013 - 08:32 AM.

[RussianWolf]

XX Sulla XX, on 24 July 2013 - 06:58 AM, said:

On game of a life time will not move your tournament score into the top ten best I can tell. The top 10 players have 10 of your game of a life time games. And its down to smaller differences between them. They are the 10 best out of the 1000 or so playing that mech in the tournament.

Same thing with pilot skills they are the best pilots for each mech. The best out of a large number of players. For each mech they are getting the most out of each mech. So I think it does give a good idea how good each mech is. But I would suggest there is one thing throughing it off. The Victor encouraged people who were great pilots but had never been an assault pilot to buy there first assault. Since the ELO is for each weight they would have been starting out with a new ELO ranking in assaults. So you would have a great say former heavy pilot with an new assault ELO. They would outclass those they were playing against. Although to be fare to the Victor I had 2000 or so matches in Atlases and moving to the Victor gave me a bump since it is easier to pilot than an Atlas.

Again, the TOP TEN are not the average.You can't take the TOP TEN of anything as good comparable data. They are outliers.

The Victor also is not easier to pilot than an Atlas. May be Easier for YOU, but not for all. And the ELO of someone stating in a new weight class is the middle. They aren't starting off playing the dregs, but other new players and those that really are average.

I don't participate in the tournaments as they are really a joke. Individual scores in a team setup..... Just like ELO, the concept is broken. If you want ELO and tournaments to mean something then they have to fit. If they want Individual scores, then free for all tournaments. If they want team tournaments, then team scores (Only score for the same 4 people dropping count). My ELO is dependent on the 7 other people I get dropped with. That's not a real way to compute my skill level of an individual.

[Tremendous Upside]

What I take from it is that the mechs at the top of the lists tend to be the ones you see most often. Why are AWS' at the bottom? Count how many you see tonight... Then compare that to the 5-6 Victors you see each game. Bigger sample size = greater chance of being at the top of the leaderboard.

[General Taskeen]

I guarantee the Awesome would be at Atlas level if the 3D Model was skinnier, and that portion of the CT was shaved down flat, and over all adjustment of hit-boxes.