Posted 18 May 2020 - 03:59 AM
@Horseman, Tarl Cabot, Bud Crue, Anomalocaris + others: Thank you all for the feedback, much appreciated.
I do have one (at least) more question, which I would like some insight into, if you can. Given the data you already have, can you explain the following discrepancy I am seeing? I went through some of my latest screens shots (from the newest backwards). The following numbers are my overall positions out of all 24 players in game (based on MS, loss / win indicated, occasional extra notes):
20 (L), 8 (W), 19 (L), 14 (W), 14 (W), 9 (W), 13 (L), 16 (W), 18 (L), 14 (W), 19 (L, 11-12), 22 (W), 2 (W, best of my own team, only one pilot on the losing side had higher MS), 11 (W), 12 (W, 12-0 stomp), 19 (L), 14 (W), 18 (L), 15 (L), 15 (L), 23 (L), 15 (W), 14 (L), 23 (L, 11-12), 16 (L), 22 (W), 20 (L), 19 (L), 17 (W), 15 (W), 13 (W), 8 (L, best of my own team, 6-12), 18 (L), 21 (L), 13 (L), 19 (L), 21 (L), 20 (W), 7 (W), 22 (L), 20 (L, 1-12), 18 (L, 9-12), 24 (L, 2-12), 23 (L, UAC 20 to my head), 20 (L), 21 (L), 22 (W), 22 (L), 18, (L, 0-12 stomp), 16 (W), 9 (W, 12-2), 15 (W, 6-8 on base capture), 24 (L, no excuses here, died 1st, really horribly bad even for me), 10 (L, 7-12), 9 (W), 8 (W, 12-2), 7 (W, 10-6 on base capture, only one foe had higher MS), 3 (W, 3rd MS, 3rd damage, alive, no kills, ERSmurf), 15 (W, very low MS despite two kills and surviving, ERSmurf), 20 (L), 17 (L, 10-12, really hard fought match for both sides, nine people got over 400 MS and only four got less than 150), 7 (win, alive, ERSmurf, full solos vs 2-man + solos, 12-3), 20 (L, apparent 2-man + solos own, full PUG foe), 10 (L, apparent full PUGs both), tied 20, W (4-6, on circle capture, apparent full PUGs both), 4 (W, apparent full PUG own vs 2-man and possible second 2-man, 12-6, 2 kills + stuff, alive, ERSmurf), 8 (W, 12-1, alive, ERSmurf), 20 (L, own side 3-man, foe apparent full PUG), 22 (L), tied 16 (W, 12-8), 21 (W), 21 (L), 14 (W, 12-4, apparent full PUG vs 2-man grupp + solos), 10 (L, 2-12, apparent full PUG own, 2-man + solos foe), 14 (W), 24 (L, 2-12, a single, soloing high skill foe pretty much took down whole team scoring 4 kills and assists on almost everyone else while doing only about 500 damage, I hit multiple foes, but none of them went down), 13 (W, 12-1, enemy had a cadet team, but that team scored their only kill), tied 13 (L, 4-6, on base capture), 22 (L, 2-12, enemy had a cadet team that killed three and damaged all others), 4 (W, alive, 12-3, apparent full PUGs), 20 (L, 2-12, enemy has two confirmed 2-mans, own team appears to be full PUG), 22 (L, 2-3 (two-three), on base capture, losing team had higher average MS), 7 (W, 12-6 despite multiple cadets in own team, none in foe), 24 (L, 2-11, non-AMS 'Mech dies to massive LRM and ATM rain), 3 (L, 2-12, 2nd highest damage, 3rd MS, enemy team has four cadets vs our one) and I think I will stop here.
There are multiple really poor games in that series and only a handful of really good ones. But my basic question is this: If I am statistically likely to be the weakest player out of all 24, how does the result level vary as much as above?
Additionally, as the sample is rather small, is the appearance that teams lose more than they win a pure anomaly due to lack of sample size?