419 replies to this topic

[Faolan65]

What were the total number of players in each class?

[RenegadeMaster]

Some reminders for people replying to this thread:

• This is a thread for feedback about the Tournament itself and not for general feedback on other people's performance or questions about the tournament. I guess that's what happens when feedback is solicited from General Discussion.
  
• This ladder "tournament" was a means to kick start ELO, stats, and such. This was not intended to be a normal tournament like MWO Community Day/Mech Madness was.

That said...

I enjoyed the tournament because:

• I could dedicate Friday night, a few hours Saturday, and most of Sunday to MWO - Over 120 matches in ~15 hours with plenty of breaks.
  
  • If I could not dedicate Friday or Sunday to it, I would not have competed and would have done group drops with my faction's batallion instead.
• Sub scores by weight class helped me see which classes I generally do better or worse with, which was just as satisfying as my overall stats.
  
• I thought my efforts were ranked fairly given the time I put in and my stats:
  
  • #147 overall, 66 wins, 58 losses, 1.57 K/D ratio (69/44)

Improvements that would make future ladders enjoyable for more people:

• If a tournament is available for more than 24 hours - After opting in, take the best or first X number of game results to obtain a pilots' score. This makes it less of a grind.
  
• Make the score number scale bigger (by a factor of 5 or more). There was never an update interval (15 mins) when my score in any one class went up by more than 5 points, even with 3 matches in one class in one period.
  
  • A smaller point scale makes it more difficult for me to determine whether I am performing worse or better in recent matches versus past matches. I'm not going to ask for crazy detailed stats or the formula for the score if I can derive more meaningful data for what is already given to me.

[yerDEAD]

Bryan Ekman, on 22 February 2013 - 01:39 PM, said:

Games played only accounts for a small percentage of your score. Just enough to reward players who are consistent over a long period.

I get that Bryan and that is fine for an ELO test, but it is not a true test of skill. I appreciate the effort of you folks setting up a tourney - the idea was fun, the leaderboards were done great, it was fun watching some of our guys on the leaderboards. It's just that it can be better. Give us a real tournament and plan it the way you would like to play in a tournament. Hey, this is what Beta is for, so I am throwing in my 2 cents.

[yerDEAD]

RenegadeMaster, on 26 February 2013 - 05:22 AM, said:

Some reminders for people replying to this thread:

• This is a thread for feedback about the Tournament itself and not for general feedback on other people's performance or questions about the tournament. I guess that's what happens when feedback is solicited from General Discussion.
  
• This ladder "tournament" was a means to kick start ELO, stats, and such. This was not intended to be a normal tournament like MWO Community Day/Mech Madness was.

That said...



I enjoyed the tournament because:

• I could dedicate Friday night, a few hours Saturday, and most of Sunday to MWO - Over 120 matches in ~15 hours with plenty of breaks.
  
  • If I could not dedicate Friday or Sunday to it, I would not have competed and would have done group drops with my faction's batallion instead.
• Sub scores by weight class helped me see which classes I generally do better or worse with, which was just as satisfying as my overall stats.
  
• I thought my efforts were ranked fairly given the time I put in and my stats:
  
  • #147 overall, 66 wins, 58 losses, 1.57 K/D ratio (69/44)

Improvements that would make future ladders enjoyable for more people:

• If a tournament is available for more than 24 hours - After opting in, take the best or first X number of game results to obtain a pilots' score. This makes it less of a grind.
  
• Make the score number scale bigger (by a factor of 5 or more). There was never an update interval (15 mins) when my score in any one class went up by more than 5 points, even with 3 matches in one class in one period.
  
  • A smaller point scale makes it more difficult for me to determine whether I am performing worse or better in recent matches versus past matches. I'm not going to ask for crazy detailed stats or the formula for the score if I can derive more meaningful data for what is already given to me.

I said this at the beginning (re. ELO Test). I think this is great and that is what Betas are for - but I wish PGI would just say that it is testing X and that is why the tournament is set that way. I started the tourney grind without completely understanding the mechanics of the tournament as pertaining to matches played. Some guys had over 400 matches in by the end of the weekend - making that 10% count for a lot. I was doing well the first day, but with 3 kids it is kind of tough to take a weekend off. If they use this format in the future, the thing to do would be to limit the number of matches to a number achievable within one day of grinding - that way some guys could drop 1 day of grinding over the whole weekend if need be.

Phoenix182, on 25 February 2013 - 10:47 PM, said:

Tourney should be entirely rate based, not a grind. Number of matches should have zero impact, other than to your average rating in various metrics (hit percentage, percent of damage in a match dealt by you, win/loss ratio, etc). Accumulating points per match just encourages poopsocking, not quality play.

It was amusing for a short while, but the standings mean next to nothing because they're additive so I quickly lost interest.

Has to have a minimum number of matches - someone could have a completely lucky high score first match, give up and still win.

Edited by yerDEAD, 26 February 2013 - 07:09 AM.

[Psychobunny]

yerDEAD, on 26 February 2013 - 07:10 AM, said:

Has to have a minimum number of matches - someone could have a completely lucky high score first match, give up and still win.

must play 50/100 matches fixes that, maby even give a cap of 100.

silentD11, on 25 February 2013 - 03:20 PM, said:

Marathon game sessions like done here aren't terribly uncommon at LANs

thats marathon though, not realy a torney, unless it was a torney about how many games u can spam launch :3

RenegadeMaster, on 26 February 2013 - 05:22 AM, said:

• This ladder "tournament" was a means to kick start ELO, stats, and such. This was not intended to be a normal tournament like MWO Community Day/Mech Madness was.

i did not notice that : /
wish i had earlyer, was it mentioned anywhere ?

Edited by Psychobunny, 26 February 2013 - 07:25 AM.

[DocBach]

yerDEAD, on 26 February 2013 - 07:00 AM, said:

but it is not a true test of skill.

Points weren't all just from matches. I noticed I was gaining points way, way, way faster in my first games, and it started to even out as I got placed in games with the other leader board players.

If you didn't have skill when you got to the leaderboard, you weren't on it very long as someone somehow was placed on it with half your games. You couldn't stay on it by just playing match after match, you had to play good matches, against some of the best players in this game, or you'd get pushed down.

Edited by DocBach, 26 February 2013 - 07:49 AM.

[StandingCow]

I'd enjoy it more if the formula for future matches didn't heavily favor those that could drop non-stop. Perhaps something that would take your best 25 matches or so....

[VladoG]

The tournament went pretty well, but I think that either ELO is not working correctly or players were just queuing with buddies.

I was playing Pretty Baby 2x LRM20 2xLR most of the time (90% of games or more) and my W/L ratio is what one could be expecting from ELO - 104 Wins 100 Losses got around 2,7 K:D ratio during tournament and due to LRM's quite a lot assists.

If you check other players you can pretty much clearly see that either ELO sucks in most cases or the players were queuing with buddies at least at some point of the tournament. Maybe they are just that good to turn 33% of matches to victory, who knows?


I am not trying to do a policeman, just standard troll here but I would like to hear some statement to clarify this.

Edited by VladoG, 26 February 2013 - 08:00 AM.

[DirePhoenix]

Bryan Ekman, on 25 February 2013 - 09:45 AM, said:

Thank you to all 8026 players who participated. The event was a great success!

Woo... that means my overall 121st place is still in the top 1.5% of all players!

[DirePhoenix]

Also, I think a lot of people either missed or ignored this dev comment several pages back:

Bryan Ekman, on 23 February 2013 - 01:30 PM, said:

We did explain the rules and scoring system. They only part we left out were the ratios. And no this weekend had nothing to do with Elo. We are testing the leaderboard system though.

[Bilbo]

Thontor, on 26 February 2013 - 08:47 AM, said:

Don't bother, they'll just say he's lying.

QFT

[ciller]

VladoG, on 26 February 2013 - 07:59 AM, said:

The tournament went pretty well, but I think that either ELO is not working correctly or players were just queuing with buddies.

I was playing Pretty Baby 2x LRM20 2xLR most of the time (90% of games or more) and my W/L ratio is what one could be expecting from ELO - 104 Wins 100 Losses got around 2,7 K:D ratio during tournament and due to LRM's quite a lot assists.

If you check other players you can pretty much clearly see that either ELO sucks in most cases or the players were queuing with buddies at least at some point of the tournament. Maybe they are just that good to turn 33% of matches to victory, who knows?


I am not trying to do a policeman, just standard troll here but I would like to hear some statement to clarify this.

Sync dropping wasn't really viable. And I am sure some people here are good enough to turn a good % of would be losses into victories.

Some of my friends in the guild tried sync dropping with me a few times but I mostly just ended up against them, which was more of a headache since they would coordinate better then my team was. I'd ask them to save me for last over chat in game, mostly because it's kinda funny.

I believe that ELO was working for the most part. When the game took no time searching you knew that you were going to get a good slobber-knocker of a fight against two even teams and when it took a long time because it was having difficulty seeding you into a match for your rating, you would start worrying about how crappy of a team you are about to get and if you had any chance of herding them to victory.

Was a crazy tournament, I don't know if I will participate in another one if its scored the same way but I'm glad I did it once and managed to place 2nd in mediums.

[BlackSquirrel]

Have a casual type tourney that takes only top scores from 2 hours or less from the day. As it stands no one should have to grind out 72 hours to be competitive.

[Monky]

Only had the time/attention span to play mediums up to the 150-200 range each night, but it was fun. The TBT-5J is a beast with a fast engine and good use of JJ's.

Future tournament recommendations; create a 'matches played' bracket.

1-15 matches
16-30 matches
31-45 matches

and so on, then break it up by day. This will provide a much bigger incentive for people to participate if they can win a category. Combine that with tracking stats in order of match played and it should work out fairly well as a skill/cheese build (hee) indicator. ADDITIONALLY, provide an off/on switch for having your stats submitted for the tournament - so we can still experiment around or run something else for variety, after all it is a game first.

One more thing, if you consider this kind of structure, have a maximum number of games per day (should solve itself with daily breakdown, I would say 60 matches total). This would allow people to take breaks and participate to the level they desire. We don't need to become poopsockwarrior online.

[WVAnonymous]

Palutena, on 25 February 2013 - 08:11 AM, said:

I would rather this Korean grindfest not be running this the same weekend as the founders/hero mech bonus cbills weekend since some of us were trying to grind our founders/hero mechs for cbills and got put in with more optimized builds then usual. I swapped from the founders Atlas to Ilya for cbill grinding to avoid the Atlas DDC herd.

I appreciated the c-bill boost while playing. I opted in to the tournament but didn't change my play habits other than playing all founder/hero mechs. I thought it was a good time, I could see the Elo working better with a lot of players on line, and I bought an Awesome on Sunday night with c-bills and had money left over.

Worked for me.

[Splinters]

My only suggestion for this tourney to be more viable in the future would be to take the top 30 scores and worse 30 scores of a set time period and average out the scores from there. This would help even out the odds and keep the leader board from being a pure drop-count based scoring system.

This way those that want to optimize and play 80+ hours won't have a much larger benefit than someone who is good and plays 15-20 hours over 5 days will be able to compete and still have a chance at winning the tournament.

Of course since this was an ELO-seeding event, I think they woudl have been better off offering banners for those who played 50,100,200,500,1000 win banners. People would grind like crazy but at least everyone would get something and seed ELO across the largest population possible. So far all they did was ELO-seed the grindmasters of MWO.

-S

[Chillybill]

well if I read all this right, if player 1 has an average score of X and player 2 had an average score of X and player 1 played twice as many matches as player 2 then player 1 would have twice player 2s score. With that said even a player with an average score could be in the top if he played 5 times more matches than anyone in the tourny.
so make these tournys the average of the top 20 scores, period! That way you could play day and night but all you could do is try to improve you worst score of the 20 best. Easy to explain, easy to compute. And while you at it PGI during the tourny would be a great time to keep you eye out for the bots, just saying!

[yerDEAD]

DocBach, on 26 February 2013 - 07:40 AM, said:

Points weren't all just from matches. I noticed I was gaining points way, way, way faster in my first games, and it started to even out as I got placed in games with the other leader board players.

If you didn't have skill when you got to the leaderboard, you weren't on it very long as someone somehow was placed on it with half your games. You couldn't stay on it by just playing match after match, you had to play good matches, against some of the best players in this game, or you'd get pushed down.

Hey Doc, I agree to a point, but I scored damn well in the matches I played and ranked well at the start while grinding. Still think it was weighted to much towards quantity than quality.

WVAnonymous, on 26 February 2013 - 02:08 PM, said:

I appreciated the c-bill boost while playing. I opted in to the tournament but didn't change my play habits other than playing all founder/hero mechs. I thought it was a good time, I could see the Elo working better with a lot of players on line, and I bought an Awesome on Sunday night with c-bills and had money left over.

Worked for me.

Yup - it was still fun - always room for improvement though.

[yerDEAD]

DirePhoenix, on 26 February 2013 - 08:24 AM, said:

Also, I think a lot of people either missed or ignored this dev comment several pages back:

Ah c'mon, I am sure most of us read the page outlining the tourney before opting in. It was there, butt the mechanics weren't completely clear. Make it easier to understand the ranking system then. It doesn't need to be complicated. The sheer number of people commenting on the structure is an indicator that it could be improved. This is all good critique, that's what is called for in Beta.

[ciller]

Bryan Ekman, on 23 February 2013 - 08:59 AM, said:

For the purpose of demonstration I took a snap shot of the top 3 medium players:

• ciller 230 Score 185 Games - For this player Games Played = 8.3% of total score.
  
• Za Warudo 215 Score 124 Games - For this player Games Played = 5.6% of total score.
  
• That Dee Bloke 183 Score 150 Games - For this player Games Played = 8.2% of total score.

If I go through the list, you would see a similar variance between 5-10% of total score being from games played. Za is probably doing better (scoring higher) per match by collecting more kills and assists, which pushes the need to play games down by several percent.



The difference in play time between 1st and 2nd is 61 matches, 49.1% more games played, or 360 minutes or 6 hours if you only had one mech to use. But only 15 points separate the two positions. This means 2nd place is far more efficient or a technically better player than 1st.

2nd place could continue to play at a slower rate and easily overtake the 15 point deficit without having to play nearly as much as 1st place.

Once the tournament has concluded, we plan to go through all of the data and see what worked and what did not. This is only one type of tournament, one which rewards playing the game. We plan to have more single elimination, round robin, and bracketed tournaments in the future, along with real prizes.

Awww... I'm pretty sure this is Bryan calling me out as a bad player!! I am SOOO glad I did not read this during my tournament push. So mean Bryan, so mean.

Edited by ciller, 28 February 2013 - 10:11 PM.