43 replies to this topic

[Jyrox]

I was thinking on this for two weeks, but I'll post now.

(Was going to give it another week of though)

The leaderboards, as we all know, suck...

But, and big BUT,

They do one thing rather well.

They help you know where you are best at playing.

My ranking is 19 in mediums as I write this.

I die a heck of a lot in mediums...

In fact my best two stats are in Heavy mechs.

Yet, when I took on board the idea of the leaderboard I understood one thing...

Yes, I do die a lot in medium mechs.

But my team has a better chance of winning if I am in a medium mech.

I have a better kill ratio in a heavy mech, but my team has a higher chance of losing.

At present dying and your team winning results in a tier increase.

Getting 5 killing blows and your team losing results in no tier change.

No, the system is not perfect, but it can show you, even in its broken state, where you are better at playing...

As I say, I wanted another week thinking about this, so any mistakes are because I didn't take the time due to this topic becoming rather hot right now.

I do agree a change in formula is wanted.

[Vxheous]

If you get 5 killing blows and a bunch of damage, you still go up even on a loss.

[Jyrox]

Vxheous, on 22 September 2018 - 04:14 PM, said:

If you get 5 killing blows and a bunch of damage, you still go up even on a loss.

My Vulcan with 5 killing blows and 400+ damage stayed at same level, the video is currently on page 1. Did I need to up the damage to 1000?

(Highest damage I ever did was in a Cyclops at 1350)

[Ted Wayz]

Your profile can give you the same info as the leaderboards.

[Vxheous]

Jyrox, on 22 September 2018 - 04:19 PM, said:

My Vulcan with 5 killing blows and 400+ damage stayed at same level, the video is currently on page 1. Did I need to up the damage to 1000?

(Highest damage I ever did was in a Cyclops at 1350)

Usually around 600 damage you'll get positive even on a loss

[Y E O N N E]

If you want a better big-picture view of how well you are performing relative to the majority of the active playing population, try using this leaderboard instead.

[Kubernetes]

You do realize that the default ranking (by total wins) is worthless, right? A guy who is 20-0 is ranked lower than a guy who is 21-500.

[SFC174]

Kubernetes, on 22 September 2018 - 04:28 PM, said:

You do realize that the default ranking (by total wins) is worthless, right? A guy who is 20-0 is ranked lower than a guy who is 21-500.

Yep, look at avg match score and WLR for a better idea of where you stand. But you have to ignore the guys that only have 20-30 matches as RNG with the matchmaker can spike you up or down pretty hard until you get enough games in.

The very best players tend to have high match scores, high WLR and high KDR (the latter being the least useful - I'd rather see KMDD's per match or something).

[Y E O N N E]

Kubernetes, on 22 September 2018 - 04:28 PM, said:

You do realize that the default ranking (by total wins) is worthless, right? A guy who is 20-0 is ranked lower than a guy who is 21-500.

Already went over that with him in the other thread he made. I'm not sure if he understood.

[Jyrox]

Yeonne Greene, on 22 September 2018 - 04:26 PM, said:

If you want a better big-picture view of how well you are performing relative to the majority of the active playing population, try using this leaderboard instead.

Cheers for that, but, when do seaons run?

[Y E O N N E]

Jyrox, on 22 September 2018 - 04:52 PM, said:

Cheers for that, but, when do seaons run?

Month to month, previous season's stats get posted on the 1st of every month.

[Jon McFuzzy]

On top of what others have said - a historical stats over several seasons would give you a better and more accurate perspective. Nothing is perfect though.

Reviewing your historical stats in the website (Profile > Stats > Mech stats) would give you factual data about your performance with all of your mechs. Export them to a spreadsheet and you can find which mechs you are most successful with. Again, this is not perfect data as well but better.

[Jyrox]

Yeonne Greene, on 22 September 2018 - 05:03 PM, said:

Month to month, previous season's stats get posted on the 1st of every month.

The problem with Jarl, and not official, it cannot take into account when I have played distraction/harrasser and died to boost my team...whole thing needs a re-vamp. Self-sacrifice in a medium is something I do a lot.

I should say here...

At times I pull 3-4 mechs away, and hope my team uses that time to push and obliterate...my tactic works at the cost of my own mech...hence higher ratings in official leaderboard.

So looking at it you may think I suck, but 4 vs 1 is going on while 11 vs 8 is elsewhere.

[cougurt]

Jyrox, on 22 September 2018 - 05:09 PM, said:

The problem with Jarl, and not official, it cannot take into account when I have played distraction/harrasser and died to boost my team...whole thing needs a re-vamp. Self-sacrifice in a medium is something I do a lot.

I should say here...

At times I pull 3-4 mechs away, and hope my team uses that time to push and obliterate...my tactic works at the cost of my own mech...hence higher ratings in official leaderboard.

So looking at it you may think I suck, but 4 vs 1 is going on while 11 vs 8 is elsewhere.

to be honest, that's kind of a bad habit to get into. distracting and harassing the enemy team is useful, but your own survival should still be a high priority.

[Y E O N N E]

Jyrox, on 22 September 2018 - 05:09 PM, said:

The problem with Jarl, and not official, it cannot take into account when I have played distraction/harrasser and died to boost my team...whole thing needs a re-vamp. Self-sacrifice in a medium is something I do a lot.

I should say here...

At times I pull 3-4 mechs away, and hope my team uses that time to push and obliterate...my tactic works at the cost of my own mech...hence higher ratings in official leaderboard.

So looking at it you may think I suck, but 4 vs 1 is going on while 11 vs 8 is elsewhere.

If what you are doing is having a positive impact, it will show in your Win/Loss ratio.

IMHO, the ranking itself isn't the greatest because it relies entirely on match score and the correlation between match score and winning.

[Jyrox]

Yeonne Greene, on 22 September 2018 - 05:23 PM, said:

If what you are doing is having a positive impact, it will show in your Win/Loss ratio.

IMHO, the ranking itself isn't the greatest because it relies entirely on match score and the correlation between match score and winning.

I would counter that under the present system if you are having a positive impact that causes your team to win, player death be damned.

Jarl does not account for that.

If however I place no importance on official leaderboard and all on Jarl, of course my play-style will, and indeed have to, change.

Funny isn't it when you wave a leaderboard in front of someones face...

Edited by Jyrox, 22 September 2018 - 05:29 PM.

[Jyrox]

This whole thread for me is turning into this....

[Ruccus]

Just play the mechs you like with the builds you like, and don't worry about the leaderboard. If you're unsatisfied with the result of a game, try to identify what you did wrong and tweak your playing style for the next game.

[Bombast]

Jyrox, on 22 September 2018 - 05:09 PM, said:

The problem with Jarl, and not official, it cannot take into account when I have played distraction/harrasser and died to boost my team...whole thing needs a re-vamp. Self-sacrifice in a medium is something I do a lot.

I should say here...

At times I pull 3-4 mechs away, and hope my team uses that time to push and obliterate...my tactic works at the cost of my own mech...hence higher ratings in official leaderboard.

So looking at it you may think I suck, but 4 vs 1 is going on while 11 vs 8 is elsewhere.

Your stats in Jarls and the PGI board are the same. The only difference is the 'Rank,' which adjusts average match score based on weight class. If your rank is lower on Jarls than in client, its because you're running predominately heavy/assault mechs, not because you're self-sacrificing in a medium.

Aside from that... I understand why you think running distractions is a good idea and a boon to your team, but I suggest you try not doing so for a bit. I think you'll find you win more often when you preserve yourself for well considered combat than you do when you try to sacrifice yourself for some grand team strategy, when most of your team couldn't start a fire with a blow torch and a gallon of kerosene.

[Jyrox]

Bombast, on 22 September 2018 - 05:52 PM, said:

Your stats in Jarls and the PGI board are the same. The only difference is the 'Rank,' which adjusts average match score based on weight class. If your rank is lower on Jarls than in client, its because you're running predominately heavy/assault mechs, not because you're self-sacrificing in a medium.

Aside from that... I understand why you think running distractions is a good idea and a boon to your team, but I suggest you try not doing so for a bit. I think you'll find you win more often when you preserve yourself for well considered combat than you do when you try to sacrifice yourself for some grand team strategy, when most of your team couldn't start a fire with a blow torch and a gallon of kerosene.

Had a bit too much beer right now, but if stats are the same, why duplicate and follow another site...ahh, Jarl is deriving stats, but showing them in a different way...

Reminds me of Churchill, lies, damn lies and statistics...

Seriously...if Jarl is the preferred way, why does MWO official not use them...any further explanations?