35 replies to this topic

[Grimlox]

I ended up with an overall PPG of 1.32 across my limited play time in 4 brackets.

My 3L was the best at 2 PPG
My Yen Lo Wang only netted me 1PPG
My Flame got me 1.33 PPG
and my Pretty Baby (without many pilot skills) netted me 1.33 as well.

Only 34 Games played, mostly during prime time PST.

As you can tell I was mostly just farming some cbills in my hero mechs since I knew I couldn't put in a lot of play time. If I can manage those PPG on my Dragon, Awesome, and Wang then I know I could've been competitive if I had enough games played. Oh and the only OP build I had was my 3L and at 2 PPG it shows. Ended up ranked 633 with 34 games played.

[Trauglodyte]

hammerreborn, on 25 February 2013 - 09:33 AM, said:

Doubtful, I have you know, respect for myself, and wouldn't get caught dead in a 3L. And I have a wife who is far more important than this game, and wonderful enough to allow me to do an almost all nighter Friday to see how high I could climb in about 10 hours (20th).

BLASPHEMY!!!

[Soulscour]

Tarantoga, on 25 February 2013 - 09:00 AM, said:

and all it says "they played a lot during weekend"... ok

You mad bro?

[BillyM]

Thread is disappointing? ...or stats are disappointing?

akke, on 25 February 2013 - 09:46 AM, said:

Just for kicks I've done averages for the top 25 in each catagory:
Catagory - Win percent - Points per game
Lights - 1.65 - 1.49
Medium - 1.26 - 1.20
Heavy - 1.31 - 1.19
Assault - 1.38 - 1.19

What does everyone take as the reason or long-term affect of lights having considerably more Points/Game?

-Is this simply raw in-your-face proof of 3L OP-ness? Shall I get my torch and pitchfork and cry-nerf for all ECM/streak-ravens?
-What about the mention of ELO taking into account weight, allowing much higher ELO lights into a game vs a heavier team?

PGI, will you provide us the variant most-used for each player in the individual bracket top-25?

BillyM (loves statistical analyses)

Edited by BillyM, 25 February 2013 - 10:41 AM.

[akke]

Here are each of the top 25's resorted by PPG:
https://docs.google....MXc&usp=sharing

[DocBach]

resorted I'm still 10th for mediums, does that mean elo works?

[akke]

I don't think it says anything about elo tbh.

DocBach, on 25 February 2013 - 10:50 AM, said:

resorted I'm still 10th for mediums, does that mean elo works?

[hammerreborn]

akke, on 25 February 2013 - 10:44 AM, said:

Here are each of the top 25's resorted by PPG:
https://docs.google....MXc&usp=sharing

You should do 3 more columns, 1 for sorted rank, 1 for actual rank, and the difference. Love seeing data though...which reminds me that I have to finish my mech diversity atlas trials.

Man...I should have done that during this tournament. Would have completely stalled my ranking (not that it mattered) but it would have shown how many people were running cheese this weekend.

Edited by hammerreborn, 25 February 2013 - 10:57 AM.

[DocBach]

akke, on 25 February 2013 - 10:53 AM, said:

I don't think it says anything about elo tbh.

i'm terrible at math and my brain is tired, i dont know what i really was trying to ask.

but either way i slice it i made it to the top 10, PPG or mystery tournament points so I'm good with it

[Lugh]

with 30+ crashes out of games I'm at .91 PPG

Taking those 30 crashes out I'm at 1.22 per game. In mostly Dragons and muromets. So you're saying I stack up pretty well then eh?

It's ok I know this as I only got to play ~8 hours over the three days and still made rank 105 in heavies.

Edited by Lugh, 25 February 2013 - 11:04 AM.

[DocBach]

who knows how they dealt with crashes, I had screen freeze that had to be hard reset like, every five games. it was pretty annoying, especially when i was doing really good in a match.

[Pihoqahiak]

Tikkamasala, on 25 February 2013 - 09:03 AM, said:

Looking at the win-loss-ratios it seems that not everyone's Elo rating has reached the correct value for the different mech weight classes. I assume it's easier to accumulate score-points in a match your team won, therefore normalizing the score per match with the win-loss-ratio (or a factor derived from it) might be a better metric.

Something that really makes any legitimate measure of statistics like that inaccurate is the fact that not everyone in the matches was necessarily participating in the tournament. That affected the matches in very random ways throughout the weekend, even more so later in the weekend as more people realized the futility of placing even in the top 25 on any of the boards if they hadn't already and weren't going to play a minimum of 14 hours a day. Many of those people likely started joining games with friends and played with differing tactics than players in the tournament. Also, Elo ratings were likely skewed due to players in the tournament focusing more in getting the most number of matches in, going out in a blaze of glory as fast as possible, and not as concerned with winning matches. This isn't to say that everyone did that, but I think it's safe to say that a large amount of players did that.

Edited by Pihoqahiak, 25 February 2013 - 02:09 PM.

[Nostram]

DocBach, on 25 February 2013 - 11:10 AM, said:

who knows how they dealt with crashes, I had screen freeze that had to be hard reset like, every five games. it was pretty annoying, especially when i was doing really good in a match.

Fortunately I only had one of those the entire weekend and it was just after I had died. Saw someone that was in the match in my next game that was kind enough to share the outcome of the match with me so I only missed a bit of good spectating for it. My side lost

[Khobai]

Just curious but could you add another column that multiplies win ratio % by points per game?

[kuangmk11]

Quote

akke, on 25 February 2013 - 10:53 AM, said:


I don't think it says anything about elo tbh.

Doesn't it? I finished 38 in assault. I am not a great player. Slightly better than average maybe. As I worked my way up I started seeing more and more names I recognized off the leaderboards, Some obviously better than me and some not so much. Near the end last night I was seeing 4-5 players from the top 25 in each match and playing the same people again and again.

Now, with elo just rolling out and the timing of this tournament why would PGI make up a separate scoring system so obviously flawed just for this tournament? Or is what we see on the leaderboards the direct projection of our elo score? That would explain the use of games played as an element in the score a little bit. New players will play new players and as they accrue games they would play more vets. If elo is suppose to be an accurate "rank" why not just use that as the scoring system? Its not, Its a way to build matches. This was just a test.

Edited by kuangmk11, 25 February 2013 - 01:42 PM.

[BillyM]

I'm really curious about the #1 and #2 on the heavy board. Their W/L ratio was so different, yet Points/Game so close that I'm curious as to their ELO score and wonder how much it had to do with the outcome...

--billyM

Edited by BillyM, 25 February 2013 - 04:19 PM.