Finding Premo – Forwards

premo_fwds

We were soooo very spoilt last year. Beams, Sidey, Danger, ROK, Fyfe to name a few of what turned out to be a “Second Midfield”.

Do we have the same embarrassment of riches in 2013? On first appearance I would say no… but let’s look a bit closer at the numbers in the forwards.

Pack a lunch, wear some sturdy shoes… it’s time to go Finding Premos’.

It was tough to choose the 15 players to study, but I went with those that I have heard talked about most and excluded any under $400k, therefore no Cloke, Sylvia etc.

I will assign a points value based on how that player ranks. 1 point for 1st, down to 15 points for 15th rank. At the end of the article, I will rank players based on those points. Obviously this is subjective on the way that I have chosen to measure them, but you will get a guide.

Once again I will kick off with consistency

Consistency

For forwards I am using a benchmark of 95 points.

So how many times did they score 95 or over…and to be greedy.. % of time they score over 95

fwdtable1

Next I will look at their Purple patch. The number of consecutive games that they score above 95

fwdtable2

Now some of these players have had multiple spurts of consecutive games over 95, where that has happened, I have taken the higher average.

Last measure of consistency is Standard Deviation… this is a fairly standard statistical measure, that in layman’s terms looks at the variance between highest and lowest scores (its more complex than that, but you get the gist).

For this the lower number is the better, however Std Dev can be deceiving as if a players floor (lowest score) is 100 and their ceiling (Highest Score) is 155 you would probably be happy with a high std dev, so I think it is important to also look at their % of time they score under 95.  I have then multiplied these 2 figures to get what I call an Acceptable Std Dev.

fwdtable3

Ceiling

Now let’s look at ceiling… When they go BANG…. how BIG do they Bang!!

Whilst I rank consistently very high, it is also good to look at scores above 110 (this is purely me being greedy!)

fwdtable4

When they score over 100, how much do they score and what do they average after they raise their

fwdtable5

Coming Home

Now let’s look at how these guys finished 2012.

Looking at the year in halves and going under the assumption that those that finished better than they started will continue that form in to this year.

fwdtable6

Starting with a BANG!

How does there 2013 draw stack up?

Are they better to start or to upgrade too?

I have looked at the teams they play prior to their bye and used their scores from 2012 to determine a rough average (this is obviously assuming they score the same, not allowing for any improvement or decline)

fwdtable7

So there we are. All of the above does not take into account, role change in team, competition for the pill and a million other things but gives you a bit of a stat breakdown and a snapshot into more than an average.

How do they Rank?

fwdtable8

My thoughts based on this.

Will Buddy ever bring consistency? He is ripping up pre season. Not having him last year when he scored that 204 is still burnt in my memory. Unbelievably it may be unique having Buddy this year.

We know that Pavlich, Thomas and Bartel have had interrupted pre seasons or may not make it for round 1. So definitely on the watchlist to upgrade but not to start with.

We also now know that Stevie will be watching from the side lines for a week. With 44 trades, start with one of the other guys and bring him in? Risky, but an option.

Riewoldt has a fantastic early draw so definitely needs to be looked at. I also like Brisbane and Adelaide’s draw. Rocky (Should be locked in everyone’s team anyway) and Zork (yep he is that good) need to be heavily considered. Tex and Wright are stars on the rise and may be worth the punt.

Robbo is a watch for me, to see how he goes under Mick, but very nice numbers when not playing that forward tagging role.

Chappy, he is tearing up the track (who isn’t) more midfield time and the aforementioned 44 trades make him very attractive.

Lewis is super consistent, will never go large, but will give you 85-95 most weeks….is that enough?

Good old Dusty.. can he explode like we know he wants to.. I think everything is pointing to yes, but just like Broughton, if you jump on board, know that it may be a bumpy ride

Lastly De Boer, not sold enough on his role yet to go there as feels like he may still get the odd run with role, and you just know it will be in the game that you need him.

 

So there you have it. Hope this helps.

 

Cheers

@RainmanDT

44 Comments

  • Ah Rob!! What a beautiful piece of work!! I’ve been hanging out for this one especially after StevieJ had his brain fade! Good to see you confirmed what I always thought that Buddy is the number 1 forward. Rooy is going to be my replacement of StevieJ until round 4 when Ill make the switch!
    Great work my friend. Ill be reading this one over and over!!!

  • Brilliant work again Rob. Few suspicions confirmed here (Buddy is a gun, SJ’s week off hurts and Pav should definitely be considered if fit). However, Bartel has been a little forgotten hasn’t he. When he received his DPP status, he was the most spoken of forward, but not a lot since.

    Food for thought.

    • Also, that is some sad reading for fans of Dusty.

      • As with all of these, you take from it what you will. Despite last years numbers, Dusty is still locked into my side. His upside is too huge to ignore

  • Fantastic as always Rob. Now considering a few changes in the forward line. As much as it pains me to say it – Dusty may not make my starting line up.

  • Great write-up. I guess you can’t count him because he wasn’t listed as a forward last year, but it would be interesting to see where Cox ranks in this list. I currently have him in my forwards, but I’m very seriously considering swapping him for Buddy.

    • In fact, Franklin in forwards for Cox, Stanley out for Duffield in D, and Cox in for Jacobs in the rucks. Only 247 more changes to make between now and round 1!

    • I omitted the DPP ruck fwds as they are a different kettle of fish. Ruck article will be up next Thursday.

  • Great stuff Rob. I think Wright would rank slightly higher on that list as he only moved into the middle in the second half of last year (as his improved average suggests). Brilliant again mate.

  • Wow mate! You keep blowing everyone away with these articles.

    I will be on Dusty obviously. With all the research that I have done over the off season, I think he is going to go bang! He was very impressive again in last weeks intra club match. The thing that keeps amazing me is that he had a bad year last year and still averaged mid 80’s. What will he average when he has a good season?

    I hope everyone keeps jumping off the Buddy train haha although I reckon that might change after his first game of the NAB.

    Thank you very much mate!

  • Brilliant piece of research. This series has been amazing. Probably the best, if not THE best piece of research and analysis I’ve seen. But I love it still gives you room to hunt for some hidden gems if you know where to look.

    GOLD

  • Sheer Brilliance once again Rainman. It’s interesting to see Bartel come second. This also reaffirms my Pavlich and Rockliff selections. I think now I’m going off Dusty, he’s cheap, but I guess you get what you pay for. Absolutely awesome rainman, and I can’t wait for the next installment. Everybody, pay attention, this is how you write an article.

    • Pavlich is still recovering from off season back surgery. Has to be a risk to start the season with.

      • I know Rids, I’m watching him. I have other plans, but I just don’t see too many consistent, reliable scorers in the forward lines. Bartel and Daisy are also both injured aren’t they? Forwards is where I’m having the most trouble. Don’t worry Rids, I’m on top of it.

        • Bartel played last week in the NAB. He has had an interrupted preseason but will be fine for round 1.

          I am a little confused by what you have said but am sure you have it covered lol

  • What can I say Rainman – another brilliant article.
    I’ve now decided to rename my first born after you which she probably won’t be happy with but how else can I honor pure genius?
    I’m happy where my 5 premo fwd starters sit on your table and I can’t wait to see them all on the park this year.

  • Thanks guys

  • Great write up Rainman. My forward line is proving to be a majorly contentious bone so far!

    Too few genuine cash cow options, too many injury/suspension concerns in the premos and a lot of risk in the mid pricers.

    This has helped confirm a few of my suspicions and validate a couple of selections I’d been pondering – time will tell if that’s a good thing!

    Looking at starting with 4 premos, 2 mid pricers and 2 rooks – unless the NAB Cup can unearth a hidden jewell among the newbies.

  • Thoughts on this? Ok so there was an article yesterday about Bird. He is 90k cheaper than Dusty and has a much better draw. My thoughts are, start Bird as my 4th mid, 80k in pocket, upgrade for Stevie or even Chapman.

  • no love for roughead?? i currently have him in my rucks and cox in the fwd line

    • Could I alert you to this comment above:
      Rainman February 21, 2013 9:04 am
      I omitted the DPP ruck fwds as they are a different kettle of fish. Ruck article will be up next Thursday.

  • I believe there are two key issues that should be addressed before anyone takes the results from this series of articles too seriously.

    1. How predictive are the various statistical measures used here? Eg, if a player had a high standard deviation or “purple patch”, say, in 2011, were they more likely to have a high standard deviation or “purple patch” in 2012? It would be interesting to calculate the correlation of each stat from year to year and test whether they are statistically significant. This would need to be done separately for each of the four positions.

    2. If they are statistically significant, how useful is this information? Is there any advantage in picking a less variable player who is expected to average less than a more variable player? I think the answer to this is less clear and it depends on whether you are aiming for an overall rank, or a League win, and whether you intend to keep a player all season or pick them up during the year.

    If these issues could be explored further, it would then be of great use when Rainman is deciding which stats to include and how much weight should be put on each one.

    An interesting project for “Australia’s answer to Nate Silver”, Tbetta maybe?

    • Moving from historic stats to expected stats is a complex subject. You need to make assumptions – for example, will the next change be stochastic (i.e., a random process) or might it have a bias (output increased 10% last year, it will increase 10% this year). But I guess you appreciate that ecky – your point 1. hints strongly at that!

      I think Rainman’s articles are really useful because he tells you what he has assumed and how he has weighted that parameter. Each of us is then free to change an assumption or re-weight a parameter. That, IMO, is the thinking bit that will give an edge if the assumptions and weights prove to be “better.”

      As to correlation coefficients and their significance, my intuition would be that in the aggregate they would tell you little that could be applied to an individual player. BTW, when someone has processed the data I will be right or wrong. Until then, that is my guess / opinion.

      Hope this comment was helpful and not overly theoretical or obnoxious.

      • Ecky and Chels both raise valid points. The relationship between past statistical performance and future performance is volatile, and depending which statistic you sue van either be very linear or not at all.

        Ecky, you have mentioned similair thoughts on a few of my articles and I really welcome the thinking. Would love for you to invest the time and put something together with the points that you have mentioned above. Warnie and the boys are always looking for new and engaging content and this would hit the spot well.

  • thanks chels havent read all comments just a few!!!! roughy has to be relevant

  • Anyone want to join our league. Pretty good team in here. 906534

  • Hmmm maybe I should get Jimmy in my side instead of Roo or Zorko, then replace with Stevie J when the time is right.

  • While the stats dont flatter him Matty Wright following the 3rd year break out rule will go top 10 FWD this year bases on his numbers and comparing those numbers with other players who had a 3rd year break out :)

    Good Luck

    • a question – what do you mean by “the 3rd year break out rule”? Are you suggesting that every player’s DT (or Sc) numbers increase in their third year playing afl footy?

      I would be interested in seeing the data to support that assertion.

      If only some players have a “break out” in their 3rd year how can you be certain MW will be one of them?

      • We dont have to be certain we have 44 trades, I am taking a punt bases on numbers i have crunched and the research i have done regarding players who did have a break out year i think Matty will have a great season..

  • Redden 67,79,109
    Rocky 23,87,112
    Beams 75,87,101
    Murphy 75,74,98
    Matty Wright 65,85,???

    Good Luck

  • cheers for the write up, As this is my first season of fantasy footy I have decided to experiment with my team until I get used to the game. Therefore after reading all the articles I might fill my team with 2 premo’s in each position and then fill the remaining spots with rookies and players in the low $2ook price range.

  • Great stuff Rainman. Your research and articles always give food for thought. Chapman might not have any upside and some of the other double speak you may hear, but he can play, am pleased he is rated highly in your essay.

  • Rainman… Love your articles! Gunna get Pav in my forward line somehow!

  • There are many here I have bantered with, and I enjoy every bit of it.

    If you are interested in joining a league with me is 445645.

    Is called ‘League Teams’,if any of you rememeber watching Jack,Lou and Bobby..you qualify to join. Cheers, Keith.

  • Firstly, I like it a lot. Great bit of work.

    However I noticed a couple of things. The rankings for the “start” rank are a bit off i think. Buddy was fifth on the list but he got a 4. Obviously that’s pretty minor.

    But something else that I was thinking was how Rainman ranked the “Coming Home”stat. He ranked it with the player who improved the most in the second half of the year being the best option (lowest rank). Is that how it should be done? For example, Matthew Wright got a much better ave in the second half of the year and therefore he was given top spot (1 point). However, doesn’t this mean that he improves as the year goes on and therefore he may average less in the first half of the year, making him an upgrade target. Based on that, should Buddy get 14 points for his position or 2 points?

    I have confused myself writing this so its just out there for discussion.

  • Thanks mate

    On Buddy he ranks equal 4th for start as has same average with Rockliff, whilst Chappy and Bartel are equal 2nd.. Therefore making buddy equal 4th.
    in regards to the “Coming Home” stat. It can be looked at 2 ways, do they take a while to warm up? Or have they changed roles etc and gone bang.

    There seems to be a correlation between strong finish to the year and good start to the next year. both Beams and Cotchin’s finish to 2011 and performance in 2012 are testament to that.

    Hope all of that makes sense.

    Thanks for the feedback and the questions.

  • Not sure if anyone is still reading this but I want to know who to take out of Martin or wright? Thankyou