The Magic Number

The Magic Number is somewhat complicated in its design. It gives us the current value of a player and allows us to use dream team as a stock market game. The formula to work out a players price is this

(75% x old price) + (25% x Current Magic Number x 3 Game Rolling Average)

Example:

  • If a 100k rookie averages 100 over the first 3 rounds, with the MN 4152, then his price will change to:
  • =(75% x 100k) + (25% x 4152 x 100)
  • = 75,000k + 103,800k
  • = 178,8k

I read the other day the magic number changes so i wanted to check this out. Using a spread sheet and the price/scores of Swan, Pendlebury, Cross and Cornes I was able to track the magic number changes for 2010.

This change will have a bigger impact on our Dream Team values than we realise. 2010’s magic number started at 4394 and was the same in round 2. But at the end of round 3 had dropped somewhere between 4140 and 4160. Following its change from round 8 it had dropped to 4000ish and down to 3900ish to finish the year. (There is a slight discrepancy in the values and this is due to the prices being rounded to the nearest $100)

Why is it changing you might ask. Well its designed so that the value of the league stays constant. Adding up all the Collingwood player prices they start 2011 valued at just over $10 million.  I remember seeing real club teams priced between $8 and $11million last year so this seems plausible. So if Collingwood is the team with the most value and the Gold Coast we can assume is currently drastically less id assume the value of the whole league is somewhere near $160million. The combined value of all AFL players is set to stay constant, As players Like Mzungu, Krakhead, Richardson, Swallow, Heppel increase their values, someone else has to lose value, to keep the league at a constant $160 million.

Now that means when Dream team calculates a players new price and multiply his 3 game average by a slowly decreasing magic number he needs to be increasing his average  just to hold his value.

  • Dane Swan started at $523.5 he lost 40k in 8 rounds Averaging 2.2 below his start value.
  • Gary Ablett started  at $522.2 he gained 26k in 8 rounds Averaging 10 Above his start.
  • Boyd started  at $456.4 he lost 40k in his first 8 even tho his average was 4.5 higher
  • Paul Chapman started at $494.5 he gained 20k in 8 rounds with an average increase of 6.6
  • Goddard began at $455.7 and he lost 30k over 8 rounds. His average had increased by 1.1
  • Cornes started at $418.4 he lost only 8k in 8 rounds Averaging 7.3 more than 2009
  • Lake started at $356.9. He gained 15k in 8 rounds Averaging 7.6 Above his start value

The magic number generally loses 10% value most of which is in the first 8 rounds but with the introduction of the Gold Coast its tipped to drop by more like 13%. 4152 is the current magic number so by round 8/9 it may be very close to 3750 and 3600 by years end.

If you can find the Premiums that can increase their average and thus hold or increase their value, then the change in magic number is only going to help you and your total teams value. The more value you have on the field then the potential of your weekly score is higher.

In 2010 Swan lost 38k even tho his Average increased by 4.

That’s the theory of the Magic number, but in practice the only thing of value to us in all that is finding out when is the best time to trade out our rookies and trade in premiums.  I graphed the value of 12 of the best rookie and rookie priced players of 2010 to find a point in which they plateaued or dropped in value.

Somewhere between 8 and 12 games  these players peaked and this coincides with the drastic drop in Magic Number values at round 10. There is no perfect time to make the trade but if your rookie seems to Plateau or drop somewhere around game 8 or just after its time to bring in your bring guns.

 

37 Comments

  • So, as an application of this, get more gc players and less guns, who are likely to drop as gc players rise?
    Or am i reading too far into this…

    • After reading this, I came to the same conclusion

    • Yes, you are reading too much into it. If you start with more rookies you will definitely have more money to throw around later, but your starting scores will be too low to win the grand prize. Ok technique for a possible league win though…. unless you run out of trades.

      Best thing is to grab your guns and ignore their price fluctuations unless they are HUGE or due to injury.

  • I wonder how the BYES might affect the MN this season. Slow it down? Make smaller changes? Hmm. Too much thinking for 8:30am, but very bloody interesting! Following the MN will add another element to DT in this one off year.

    • Not sure about slowing it down, more so wreaking havoc!
      With Ade, Kan & especially GC not having their first price changes till round 4, I guess the value of your Swans, Boyds wont go down that much, but will fall quite a bit round 5 till round 9.
      But then again, in round 4 you have DT guns from Stk & Wb who wont be playing and dropping in price to balance the price rises of the rookies, and more so the case in round 6!

      So shit, logically that would mean that any gun not having a bye in the first 6 rounds will be subject to the greatest price drop!!!
      That throws a spanner in the works!!!
      If thats the case, not starting with Swan, & picking him up possibly $80K cheaper, has merit.
      But then again it also supports the Guns & Rooks strategy!
      I’m confused, but I’d still prefer to bank the points that Swanny & Boyd will be churning out!
      Not so bad now that teams are so similar- puts even more importance on trading.

      • Hey Eeepo,

        I wouldn’t stress about it too much. Dont forget it is the same every year for every premium. You cant not pick the Swan’s and Goddards just because they may go down 10% or so by years end. We are all in the same boat dont forget.

        The best bet would be just to enjoy all those rooks going up 100 – 150K and the fact that Didak and Ablett will be cheaper when you upgrade to them later!

      • I’m thinking your logic is mostly sound, expect you’ve forgotten one thing. After a premium player has a bye in the first 8-10 rounds, he’ll be more overvalued than other premiums that didn’t have the bye. As such, he should drop more in the few rounds after his bye, I would’ve thought.

        This probably makes the thought of not having players with no byes early redundant. Either way, dropping a Swan for someone else will probably be good/bad more based on their output on the field than the MN and byes.

        • So starting with Stanton. (1st bye rnd10,) and trading him to Ablett after round 9 could be interesting!
          Yes, sideways, but Stanton has ugly rnd 24 bye, so kinda killing 2 birds with 1 stone.

    • I personally think this is a crap system. What is wrong with the value of the entire competion going up. It is unfair that players drop value after increasing their averages.

  • FF Genie was very good at assisting in when to trade players in/out. It wasn’t foolproof, but helped me understand the MN and how the 3 game average works. Too bad I only discovered it 1/2 way through 2010! Hey Warnie, doesn’t assistant coach have a similar feature? Wonder if it works as good?

    Again though, alot of DT is luck. I thought Trengrove was ripe to trade at rd 9 last year. He turned out to be a keeper as your bench midfield. Got Ablett out of it though. (And it was the week he was rested too, hmm that was the end of Thursday night trading for me!). Good ol DT, always keeping you on your toes!

    • Yep, the ASSISTANT COACH feature will do what the FFGenie does… but will be ‘official’ – ie. it’s made by Virtual Sports and it’s not ‘guessing’ like the Genie or other sites.

      • Hmm interesting. Might be what would get me over the line in regards to getting the program. That and you guys writing articles on it as well, kinda like two in one. The entertaining articles, and all the stats in one spot!

      • Did you mean to type “not guessing” instead of “guessing”? I think when you look at the formula, the closeness of the sites/apps in their magic number formula won’t make a huge difference. Plus, the owners of those sites would be idiots if they didn’t just use the formula to reverse engineer the magic number themselves.

        On using this analysis to work out when to trade rookies, I think its either why over simplified or made it way to complex! The best judge of when to trade a rookie is:
        a) need for cash (if need to store it or use it for double trade) or availablitiy of cash in your back (upgrade to premium)
        b) quality of replacement player (no use trading someone to a crap player that is no value to your team. NB: Multiposition players that won’t play could have value you to your team at some point in the season)
        c) Expected scores vs Breakeven point (So if you are trying to make money from a rookie at some point they will level out or maybe even fall away. As long as you have b) covered, that could be the time to move)
        d) Is the player getting games/suspended/injury (even if they have matured in price, if there not playing and don’t look like playing, you may have to move them and cut your losses (or potiental gains))

        JMTC (Maybe I should have wrote an article on this! ;-) )

      • Disclaimer: I am the FFGenie developer.

        I would like to point out that all breakevens are ‘guesses’ to some extent. The actual breakeven for the week is dependent on the all the scores for every single player in that round. No one person, site or application can tell you exactly what the breakeven is (unless they have a flux capacitor handy ;))

        Having said that, if you’re relying on the breakeven to be that exact then you are using it wrong. Breakevens should be a general guide to whether a player has peaked in price, rather then trying to squeeze every single last dollar out of your salary cap. If FFGenie says Player X has a breakeven of -5 and Assistant Coach says he has a breakeven of -8, is that really going to make much of a difference? All you should pick up from that is that the player is certain to rise in price after the completion of the round.

        As for the byes, my guess is unless the teams involved in the byes are very high/low scoring teams compared to the league average, it would have a negligible effect on the MN.

        • just logged in to say … love your work grimlock!

        • Touche.

          The VS breakevens will of course be a best guess, I doubt these could even be ‘live’ thoughout the round… as a general rule, the FFGenie has the BEST accuracy of any site/application! Love your work grim! You are DT royalty!

        • Also love your work.
          Been using FFGenie for years, and just wanted to say that I really hope you are releasing it again this year. I’m trialing the assistant coach thing but I dunno I still like the look/feel of the Genie better… It’s probably just the years I’ve spent staring at it but yeah I’d really miss it if it’s not around this season.

        • Yeah, great program Grimlock. Keep up the great work!

        • grimlock’s alright with me, oh yeah! FF genie is the tits, helped me out so much last year when i first discovered it.. i duno how its possible to crack the top 1000 without it really :)

  • Having said this, this analysis backs up the experimental stradegy I have suggested my wife use for her first year of DreamTeam.

  • So when are you guys organising an interview with Guy McKenna? :)

    • We’d love one. So many things to rack his brain over. Maybe a DT study trip to GC is in order?!

      • hey warnie did you get my article is sent you on unique players i sent it early this morning..

  • Thanks, great article guys, was wondering why people were more fixated on the MN this year.

    • GC (and GWS next year) will mean larger drops in the MN over the first 8-10 rounds this year than in previous years. Which, in English, means that premiums are going to have to improve their average by more than usual to maintain their value. What you do with that info is one of the many things making this year particularly interesting.

  • Thinking of starting an Ultimate Footy draft-style for DT Talk regulars who are keen for a different version of fantasy AFL. It would only work if every coach was 100% committed, otherwise there is no point. Would be a Live Draft, in my experience long drafts cause shit teams which equals lack of interest.

    So express your interest here!

  • Does it matter if your guns drop in value though? if your planning to keep them and not trade them, then their value is irrelevant.

    • That is true. But in regards to upgrading it may help for you to choose the optimum time when the guns will be at their cheapest and rookies at their peak.

      • This is true, but also you can’t expect every premium in your team to have been upgraded – just not enough trades. So for the initial players in your squad, who cares whether they drop in price? I mean, ill never trade out Swan, so couldnt care less if he dropped 50k. But then the argument is that you should pick him up after the 9 rounds… but then thats countered by the fact that you’d rather Swan in your team for those 9 rounds than just to pick him up 50k cheaper later on.

        The magic number is very important for your upgrade targets though – For example, whether Ablett recaptures his Cats form or not, his price is basically guaranteed to drop. So he will be a great pickup after round 9 as everyone already knows. Will be following very closely to find out the best value for my upgrades!

        Cheers Knightmare, love your work.

  • Won’t the magic number also be affected by the low scoring of the substitute.
    As this will effect 2 in 22 players, and these players will only be scoring half of what they usually would this would offset the introduction of the GC side. Especially when you consider the number of GC veterans who will not be increasing in price.

  • I’m a little confused about why we use there is a “25%” in the formula. For example, if the current magic number is 4000 and we take 25% of this we get 1000. Why not just say that the magic number is 1000 and then the formula looks like this:

    (75% x old price) + (Current Magic Number x 3 Game Rolling Average)

    The formula just seem unneccessarily complex. Am I missing something? Is the magic number also used for calculating something else?

    • Old price: 400k
      3 game average: 100

      (75% * old price) + (MN * average)
      (75% * 400,000) + (4,150 * 100)
      (300,000) + (415,000)

      New price = 715,000

      • Ah shit, misread what you wrote =P. By making MN = 1000 you’ve multiplied it by 25%, so that’s why it’s in the formula.

        MN also dictates starting prices, so if MN = 1000 all the players would be 75% cheaper than they currently are.

  • Gold Coast will drastically affect the Magic Number!

    Last year Premiums dropped due to magic number dropping due to rookies increasing in price (I’m looking at you Barlow). This year with Gold Coast the amount of rookies and low price players is much higher so he magic number will decrease faster and our premiums will decrease faster. Assistant coach predicts players like Swan losing nearly $80K in first 6 weeks so maybe this year, a mid price strategy picking undervalued players, then upgrading much faster than a Gun/Rookie strategy could be the way to win the car.