Stats-based Projections - Week 8
Apologies for the slight delay in this week's projections. All-day meetings meant that I actually had to do work today.
The accuracy of projections based on stats continues to be subject to failures of form, turns of fortune, clown fiestas, throws, and roster tweaks making an immediate marked improvement to performance.
You can see how my stats-based projections compared to the actual results for week 7 in my latest comment on last week's discussion: http://boards.na.leagueoflegends.com/en/c/fantasy-lcs/5dIKEIUs-stats-based-projections-week-7?comment=0001
Team Summaries - Week 7
EU - In order of 'did the most bestest'
- OG joined the 'give ROC a good whipping' club, and had a jolly good time inflating their points for the week, despite not getting a great number of points from S04.
- SPY had a better week than expected, capitalising off a somewhat unexpected poor GIA performance.
- G2 did slightly better than expected, getting H2What to give up more deaths than usual.
- VIT as a whole performed about as expected, but there was an imbalance in the sharing of points among the VIT players.
- UOL did better than expected, but still performed about average for their respective roles.
- ROC did about as well as can be expected, continuing to money-shower opponents with points galore (joint top in most deaths per game - Equal with FOX, pretenders to the 'thown' - see what I did there?)!
- S04 got some ok points from FNC, but OG's new lineup seemed to instantly put a halt to their bleeding of points to opposing teams (giving away the fewest number of deaths for a week by OG), S04 vs OG turned into a pretty low number of kills each game, despite the game length.
- FNC got some ok points from H2K, but struggled against S04 (who went back to their typical low number of deaths per game).
- GIA - Oh dear. What happened to GIA? They got smashed 0-4, that's what, just to remind us that they're not quite as good as we were starting to think they were.
- H2K weren't expected to do well this week, but they did even worse than expected. G2 kind of gave them a few points, but FNC absolutely destroyed them!
NA - In order of 'same as aboveness'
- IMT, despite drawing their first 2 games against CLG (and disappointingly scoring big in the 3rd game - [sad first-2-games face]), were irrepressible against FOX, helping FOX on their way to the most deaths they've ever given away in a week.
- TSM as a whole did about as well as expected, which is always pretty good these days. But there was an imbalance of points distribution among the players. Doublelift went off on a feeding frenzy on poor FOX.
- NRG, as a whole, had a decent set of matches against P1 and NV. But GBM turned out to be the main beneficiary against NV.
- TL subsidised their poor show against TSM by pummelling a much softer P1, and ended up doing better than expected overall.
- APX - Despite getting a low average number of kills per game for the week, most of them occurred within the first 2 games. APX capitalised off of a struggling C9 team.
- P1 did surprisingly well, getting some good points against NRG. But ended up pretty average overall.
- NV players did either about as well as can be expected, or under-performed. I was certainly hoping for more from their games against NRG.
- C9 continues their disappointing slide down the table, and in fantasy performance. Only Jensen managed to meet his expected projection, and that's because of a penta. Time to start looking for replacements? Are their stats boosted by early weeks? Week 8 will definitely tell all, but will it be too late by then?
- CLG did better than expected by dishing out an impressive game 1 performance against IMT, but struggled to get any meaningful points from thereon out.
- FOX uncharacteristically joined ROC in the 'most average deaths per game' stakes in week 7. Though IMT & TSM is a particularly tough set of matchups for pretty much any team that's not IMT or TSM.
Now, here are the bits that are supposedly going to help us form an opinion about who may/may not do well in week 8.
Kills Per Game Stats (for all games played)
http://i.imgur.com/RjdHFsS.jpg
Deaths Per Game Stats (for all games played)
http://i.imgur.com/8hPsgcE.jpg
If you feel like you can make any kind of prediction based on what has happened in previous matches this split, good luck! Here's what the 'stats' say are the projections for week 8, apparently.
Top Projections - Week 8
http://i.imgur.com/wUdK8iF.jpg
Jung Projections - Week 8
http://i.imgur.com/Gg7Sjx6.jpg
Mid Projections - Week 8
http://i.imgur.com/bff3QgB.jpg
ADC Projections - Week 8
http://i.imgur.com/we19LMV.jpg
Supp Projections - Week 8
http://i.imgur.com/soxPn5o.jpg
I would ordinarily ask you for up votes (which are always greatly appreciated), but even I'm starting to lose faith in historical stats being any good for being able to predict the future. So many upsets is, well, upsetting the credibility.
