2018 Round Review: Round 7

Discussion in 'Blog' started by walesy, May 6, 2018.

By walesy on May 6, 2018 at 6:02 PM
  1. walesy

    walesy Administrator Staff Member

    Joined:
    Nov 30, 2015
    Messages:
    3,872
    Likes Received:
    2,564
    Another week rolls by, and again I'm gunna get done in the second-to-last game of the weekend for my cash league.

    Gunna need some heroics from Murray to snag a chance (Needs to match Hibberds 96)- If only his SC score was matching his DT!

    Anyways, how have you all gone this weekend?
     

Comments

Discussion in 'Blog' started by walesy, May 6, 2018.

    1. The_Swert
      The_Swert
      I've been doing some analysis on fallen premiums and whether or not they might find form and bounce back up.

      I've gone over the past 3 years and selected some premiums that had tough starts to the year up to round 7, well below their average from the previous year.
      Then I've seen how they have gone after round 7, do they bounce back up to the previous years average? or do they maintain their poor form?

      [​IMG]
      In 2015, Rockliff was the biggest shock, anybody who bought him in Rd1 would've cried as he only returned half of his investment in rounds 1 to 7 (69 points below his 2014 avg). His form did improve and his r8-23 average was 114 and very much in the premium zone. So those who bought him in R7 at a cheap price did really well.

      Most of the other fallen premiums however never recovered form and were not really bargains even at their cheaper r7 prices.

      [​IMG]
      In 2016, four players who had poor starts did really well after round 7. Anybody who picked up Rockliff (again!), Lewis, T McDonald or Priddis in round 7 got bargains. Rockliff did so well in the second half of the year that his 2016 avg surpassed his 2015 avg despite the poor start.

      But once again, most of the other fallen premiums did not recover.

      [​IMG]
      In 2017, both Zaharakis and Dangerfield were great round 7 investments and really picked up after that point.
      Some others like Parker, Sidebottom and JPK showed improvement but never quite reached expectations.

      These lessons show the risk of buying some of the fallen premiums in 2018.

      [​IMG]

      Here's a selection of fallen premiums at the moment and how far below their 2017 average they are.
      Using weighted averages, I have done some projections.
      The Orange column shows what they would need to average across rounds 8-23 so they deliver the same annual average as in 2017 - to meet expectations of those who bought them in r1.

      Its going to be tough for any of those players to deliver a 2018 average that matches their 2017 average.
      However, those buying now will be paying less, so their expectations aren't the same. They just need their form to improve to premium levels. History shows that 2 or 3 may do that, but trying to predict which ones is the hard part.

      I note that this analysis is mathematical only (and not exactly rigorous) and does not take into certain injuries etc that might affect scoring. However I believe the message is still clear - not many fallen premiums recover, but there are diamonds in the rough each year.
      • Like Like x 9
    2. Owen
      Owen
      @The_Swert great post, and something I consider most of the time when picking a fallen premo. Another thing that would be good to look at is standard deviation, as I would expect that JJK has a very large one. My hunch is that fallen premos with low standard deviations from previous years would struggle to catch up, or nearly catch up, to their previous years average.

      Also Rockliff hey? I wonder if when he comes back fit and firing he will be in the midfield, because if so, :)
    3. RPritch
      RPritch
      As I'm $500 off being able to go English and Finlayson to Spargo and M.Crouch, I'm leaning towards doing the same trades, but getting D.Smith in rather than M.Crouch, knowing that this will definitely allow me to do a double trade next week of Naughton and Fritsch to Simpson and a rookie (probably L.Keefe). Any general thoughts appreciated. Thought about getting McLean in rather than D.Smith, but I'm not sure how he'll continue on from his unexpected start. He would not allow me to be able to afford any upgrade target next week better than a Shane Savage
    4. The_Swert
      The_Swert
      Yeh I don't get it. Why does Rockliff always seem to have terrible starts to the year and amazing second halves?
      Actually last year he reversed it. :confused:
    5. TheTassieHawk
      TheTassieHawk
      Not sure how "unexpected" McLean's start has been, he averaged 90 plus over his final 11 games in his 3rd season so 90-100 average for 2018 seemed gettable if he continued with a SC friendly role.
      • Like Like x 1
    6. RPritch
      RPritch
      Fair call TTH, I understand where you are coming from there. It's probably more so from a viewer of the AFL game being played perspective that it is unexpected to me. As someone who has always followed the Dogs closely (as a 2nd favourite due to my brother supporting them) he does not seem very noticeable to me on game day. I'm not saying I thought he was a poor player last year, or has been poor this year, he has been very good, but I look at his avg, see 110, and think 'shit, he isn't quite that good'. That is rather irrelevant to playing SC though. My main concern is how he will score in losses when they play better teams, however he seems to be doing rather well in general so far
      • Like Like x 1
    7. Jason
      Jason
      I'm expecting English to be rested this week.
      • Like Like x 2
    8. Len
      Len
      Royal Wedding holiday?
    9. ike2112_01
      ike2112_01
      Straight off the SC game - I'm a subscriber - it says 128. It was 119 this week and 115 the week before.

      As it happens I figured out that I could drop Bonner and Petracca, switch Finlayson to defence, and bring in both Sicily and Macrae. I think that's a pretty handy situation as then depending on who is lining up for their 3rd game next week I can downgrade a defender or forward using Sicily.

      Attached Files:

    10. Jason
      Jason
      @walesy this one's for you to explain why TS is right and the SC website is wrong!
      • Like Like x 2
    11. ike2112_01
      ike2112_01
      It looks like his scoring is fairly consistent, which is good. But also his breakeven only projects to being low for another week then it rises I think to the mid-70s, then to roughly his avg (110 ish), so he will probably only rise about $30k roughly. I think it's worth waiting until the byes for him, that way we'll see if he maintains and looks a genuine top 6 forward.
    12. Jason
      Jason
      Not unexpected to me, particularly when this is my info!
    13. ike2112_01
      ike2112_01
      Damn... maybe I ought to bring in Spargo actually instead of going for Macrae just yet. His job security better than Ronke's I presume but is it good enough?

      Dogs have the Lions this week though, fair chance to score big again.
    14. port_leschenault
      port_leschenault
      At the other end of the spectrum, Billings had a 98 average over his last 12 games too.

      Good post @The_Swert
    15. anthak
      anthak
      Hey mate, you should check the breakevens on the TS stats page because they’re usually more accurate than the SC site.
      The SC site doesn’t consider that there’s an overall pool of value that stays constant every week, which makes their predictions less accurate than on TS.
      I think that’s why it is.

      http://tooserious.net/forum/stats.p...&up_code=&tsLeagueTeamId=#BVzljkAvgKAxrB1Q.97
    16. anthak
      anthak
      I’ve just found the article Walesy posted about this previously:

      http://tooserious.net/forum/threads...coach-gold-is-wrong.5796/#oREmdA7dr4GeEsEP.97
      • Like Like x 3
    17. Owen
      Owen
      I thought I'd have a look to see how well the previous weeks break evens matched up on TS:

      Witts moved $100 whilst scoring 114, TS had his BE at 99
      Hawkins moved $300 whilst scoring 108, TS had his BE at 94
      Motlop moved $500 whilst scoring 64, TS had his BE at 70
      J Thomas moved -$200 whilst scoring 113, TS has his BE at 101
      Goddard moved -$400 whilst scoring 89, TS had his BE at 74

      Seems like as there breakeven is higher, the predicted BE to the actual BE is higher. Is it possible to see previous weeks breakevens on the heraldsun subscriber page to see how they compare?
      • Like Like x 1
    18. anthak
      anthak
      Good idea. See how it’s going now. It can never be 100% accurate because you don’t know for sure how all of the players will score in advance. And there will be some weeks where the score distribution will suit one method over another. But will be interesting to see how the SC site has gone too.

Share This Page