6 Extra Trades

Joined
19 Jun 2012
Messages
8,410
Likes
10,892
AFL Club
Collingwood
#21
Yes.

Thinking Fast and Slow by Daniel Kahneman (which is not a light read).

But Fooled by Randomness by Nassim Taleb more of an influence for Supercoach (as well as his better known The Black Swan).


Looking forward now to reading Wait: the Art and Science of Delay by Frank Partnoy.

'The art of knowing how long you can afford to delay before committing is at the heart of many a great decision—whether in a corporate takeover or a marriage proposal. Exploring decisions from those made in half a second to those that take months and years, Partnoy demonstrates that procrastination is often virtuous, that the ability to wait is the path to happiness, and that our gut instincts often betray us.'

Sounds appropriate for Supercoach (especially the gut instinct part) so this year I have named my team Bringing the Wait. Just need to read it now.....while also procrastinating.
I enjoyed Kahnemans' book immensely, recommended it to Rowsus before Chriistmas, be right down his alley I think. Good luck for the year Stroppy.
 

Rowsus

Statistician
Joined
19 Mar 2012
Messages
28,615
Likes
61,944
AFL Club
Melbourne
#22
I enjoyed Kahnemans' book immensely, recommended it to Rowsus before Chriistmas, be right down his alley I think. Good luck for the year Stroppy.
Still trying to find it, too! Unfortunately it is out of print now, but I'm sure I'll track one down!
 

Rowsus

Statistician
Joined
19 Mar 2012
Messages
28,615
Likes
61,944
AFL Club
Melbourne
#23
Isn't it effectively more than 6 extra trades?

Don't need a bye week trading strategy this year, which added a couple of extra trades for me last year to try to get an advantage.

For me the extras will first go to fixing any initial mistakes (weeks 3-5).
If there have not been too many initial mistakes, then use the extra trades to use the pinch hitter theory to take a calculated risk by swapping out a high scoring player that the evidence suggests will come back to earth for a fallen premium that is too good to be so bad. Both for total points and for extra cash.

In the AFL and Supercoach, I do believe in reversion to the mean unless the player is truly special. And there are not many of them.
I'm a huge believer in what I call "corrections". I just don't understand that there are 1,000's of Coaches out there, that wait for that mid-season fallen Premium, and pounce on him at his low price (think Priddis last year). They quite rightly, as Stroppy suggested, are working on the "he's too good, to be so bad" theory, but next to none of them will work the other side of the strategy.
I wrote after round 6 last year, that Coaches who had Stanton might consider trading him out. I suggested that history, and his role at the club, would mean he would do well to finish the season on an average of 110. Given at the time he had played 6 games and scored 793 at 132.2/game, if he played every game for the rest of the season, to fall to my 110 average, he would only average 102/game for the rest of the season!
History proved me wrong, then right. He scored a massive 164 then a 142 in rounds 7 and 8, before his season fell in the toilet. With those 2 big scores in he averaged 92.8/game in the 14 games he played after round 6. In the 12 games he played after round 8 he averaged a meagre 82.8 with only 2 scores over 100!
It's a side of SuperCoach strategy that I think should be explored more. If you are willing to risk bringing in a player with a good history, who has been performing poorly and now his price looks attractive, why aren't you willing to do the reverse, and trade out a player that is playing well above his historic ability, and will surely have a correction? This correction usually comes sooner, rather than later.
I know there will be some sharp thinkers out there, that will point to Beams, and say under my theory/strategy you would have traded Beams out, too. I have a two part answer to that:
Firstly, Beams only averaged 109.9 up to the bye. Not quite the stellar jump we are looking at with Stanton. He went on to average 134.2 after the bye, and if you look at his season, up to round 18 he averaged 114.5, with a hot streak between rounds 9 and 16 where he averaged 127.9 (still not quite Stantonesque!). It was his last 5 rounds where he averaged 149.9 that put his season over the top. Too late for corrections then!
Secondly, even if you think I'm tailoring my thoughts/strategy to hold Beams, for every player that doesn't have a corrective downturn in a Stanton type position, there are at least 10 - 15 that do! You just don't get many "runaway trains" that steamroll through a whole season!
Knowing when to quit "riding the wave" is better than knowing when to jump on board that fallen Premium. There is no easy answer to that. Stanton faced WC, Rich, GWS, Melb, Syd in his run up to the bye. Given he was facing GWS and Melb in rounds 9 and 10, I was predicting his correction would "begin" against WC. I was wildly wrong! He scored 164 against WC, 142 against Richmond, but then fell apart with 72 against GWS, 96 against Melb and 83 against Syd. Using the draw is probably still the best tool, but as Stanton shows, even that can be flawed.
 

WandP

150 Games Club
Joined
13 Apr 2012
Messages
838
Likes
55
AFL Club
Geelong
#24
If you are willing to risk bringing in a player with a good history, who has been performing poorly and now his price looks attractive, why aren't you willing to do the reverse, and trade out a player that is playing well above his historic ability, and will surely have a correction? This correction usually comes sooner, rather than later.
This reminds me of a really interesting article in the age i read a couple of months back regarding humans' ingrained optimism bias. Being the tragic that i am, i instantly related it to supercoach and a number of horrendous trades i made last year (Heater to Malceski anyone?).

I think what you are suggesting goes against the grain of our natural instict for optimism whereby if we categorise a player as "a good player", we believe he'll either bounce back if he's scoring poorly, or will continue on if he's batting way above his average.

http://www.theage.com.au/world/why-humans-look-on-the-bright-side-20130101-2c3sr.html
 

WandP

150 Games Club
Joined
13 Apr 2012
Messages
838
Likes
55
AFL Club
Geelong
#25
Re-reading that article makes me want to change my team to to "The Irrational Optimists"..
 
Joined
19 Jun 2012
Messages
8,410
Likes
10,892
AFL Club
Collingwood
#26
I'm a huge believer in what I call "corrections". I just don't understand that there are 1,000's of Coaches out there, that wait for that mid-season fallen Premium, and pounce on him at his low price (think Priddis last year). They quite rightly, as Stroppy suggested, are working on the "he's too good, to be so bad" theory, but next to none of them will work the other side of the strategy.
I wrote after round 6 last year, that Coaches who had Stanton might consider trading him out. I suggested that history, and his role at the club, would mean he would do well to finish the season on an average of 110. Given at the time he had played 6 games and scored 793 at 132.2/game, if he played every game for the rest of the season, to fall to my 110 average, he would only average 102/game for the rest of the season!
History proved me wrong, then right. He scored a massive 164 then a 142 in rounds 7 and 8, before his season fell in the toilet. With those 2 big scores in he averaged 92.8/game in the 14 games he played after round 6. In the 12 games he played after round 8 he averaged a meagre 82.8 with only 2 scores over 100!
It's a side of SuperCoach strategy that I think should be explored more. If you are willing to risk bringing in a player with a good history, who has been performing poorly and now his price looks attractive, why aren't you willing to do the reverse, and trade out a player that is playing well above his historic ability, and will surely have a correction? This correction usually comes sooner, rather than later.
I know there will be some sharp thinkers out there, that will point to Beams, and say under my theory/strategy you would have traded Beams out, too. I have a two part answer to that:
Firstly, Beams only averaged 109.9 up to the bye. Not quite the stellar jump we are looking at with Stanton. He went on to average 134.2 after the bye, and if you look at his season, up to round 18 he averaged 114.5, with a hot streak between rounds 9 and 16 where he averaged 127.9 (still not quite Stantonesque!). It was his last 5 rounds where he averaged 149.9 that put his season over the top. Too late for corrections then!
Secondly, even if you think I'm tailoring my thoughts/strategy to hold Beams, for every player that doesn't have a corrective downturn in a Stanton type position, there are at least 10 - 15 that do! You just don't get many "runaway trains" that steamroll through a whole season!
Knowing when to quit "riding the wave" is better than knowing when to jump on board that fallen Premium. There is no easy answer to that. Stanton faced WC, Rich, GWS, Melb, Syd in his run up to the bye. Given he was facing GWS and Melb in rounds 9 and 10, I was predicting his correction would "begin" against WC. I was wildly wrong! He scored 164 against WC, 142 against Richmond, but then fell apart with 72 against GWS, 96 against Melb and 83 against Syd. Using the draw is probably still the best tool, but as Stanton shows, even that can be flawed.
Rowsus, your article is along the lines of Stroppys' earlier statement of his belief in 'regression to the mean',.. a player of X ability may have a couple of good ones, then a couple of bad ones, then a couple of fair ones, in no particular order, and overall regress to his mean average..it is not possible to guarantee with a fair degree of certainty that you ride a wave at the right time. You have stated in a way why to me, the pinch hitter is a one off maybe, but risk return is the overiding factor and one to be judged at the time, this takes a deal of thought. As compared to the Priddis deal, this offers/ed in my view a greater degree of certainty and was an offer too good to ignore last year. If the variables were so easy...
 
Joined
24 Aug 2012
Messages
193
Likes
90
AFL Club
Hawthorn
#28
I think when looking for those "regressions to the mean" you really do need them to have a few seasons to have formed your gut opinion on first. I think it gets tricky the younger they are.

For me Id say Scott Selwood and Tippett were other lesser ones alongside Stanton last year. Tippett for example started off well in the first month or so to average 100-odd to bring him into some low-end consideration for me due to his still low end price. I'm lucky that I have a mate whose notorious for jumping on these types of blokes early which thankfully deters me. He gets em wrong every year but hasn't quite worked it out yet, however it helps me not get too carried away when he picks em!

On the flip side I think i've been burnt by this a little, however looking back I maybe should have had more confidence in what their history suggested.

Fyfe's second year breakout, Kennedy last year, Montagna way back, and to a lesser degree Beams and Dangerfield last year as well.

They are obviously all solid fantasy scorers in hindsight but at the time I expected them all, to varying degrees, to drop off, mainly because by the time I was convinced of their validity they were either too expensive or I had missed the boat.

Fyfe did me cold as I was sure he'd drop right off seeing as though there was no real precedent for a 2nd year bloke to go bang and maintain the rage. What really got me was the fact that, as a Physio, i kept waiting for the bung shoulders he was playing with to let him down and justify my stubborness. I could never bring myself to trade him in.

Kennedy, Beams and Montagna shouldn't have repelled me as much as they did as they were coming off a much higher base around 95-100 the previous year. But the frustrating thing is that Stanton did too. Maybe the gut feel needs to be adjusted. I think most people have never rated Stanton but did the others pre-jump.

I think the gut works more in favour of avoiding these types of potential future regressors rather than bucking the trend and jumping on one successfully.
 

Bomber18

Leadership Group
Joined
11 Nov 2012
Messages
26,920
Likes
63,223
AFL Club
Essendon
#29
But the frustrating thing is that Stanton did too. Maybe the gut feel needs to be adjusted. I think most people have never rated Stanton but did the others pre-jump.

I think the gut works more in favour of avoiding these types of potential future regressors rather than bucking the trend and jumping on one successfully.
If Essendon ever become a top 4 side, Stanton will average 120 (think how Montagna averaged so well during St Kilda's great run but now is not an option)
 

Stroppy

Rising Star Winner
Joined
3 Jul 2012
Messages
237
Likes
11
AFL Club
Gold Coast
#31
I am with Rowsus.

It is easy to bring in a fallen premium since we all love buying quality at cheap prices.

So for a real POD look at doing the hard thing and swap out an over-priced premium. I guarantee most people will not.

So I am going into the season thinking about selling out of over-priced players. I might not because it is hard to do, but if you are not thinking about it you will never do it.

I would not have done it last year given the number of trades and the need for a bye strategy of some kind. But it can be done this year with the extra trades.



I know I am probably over-thinking this, but one player that might fit this profile by mid-season is the mighty Dean Cox.

If NicNat is slow out of the gate and Cox has a big workload and big points then Ol'Man River might be stuffed by the mid-point of the year. Yes, this would be very hard to do. Which is why (like Impromptu and Kreuzer) it might just work.
 
Joined
20 Jul 2012
Messages
929
Likes
257
AFL Club
Geelong
#32
When picking up a "fallen" premium you have the ability to have seen the games to determine why they are scoring below what you perceive they should score. Whether it be carrying an injury or an injury affected score to drop the price or even two poor games back to back (perhaps an unusually heavy tag)which can drop a price quickly. Therefore you believe the reason for the low scores will not continue and you pick the player on what they are capable of averaging under normal conditions

On the flip side I have been the type of person to jump off a player because I believe the are over-achieving and it cant continue. I traded out zorko last year as thought he had topped out at a certain point and wanted to use the money to my advantage. This obviously was the wrong move.

I too have taken the option of not jumping on just any "breakout" player just because they are scoring highly. I decided the risk was too high with fyfe in his second season with his shoulders and only being a second year player that he MUST drop off so did not trade him into my side. Again this proved to be a poor decision in hindsight.

What I am trying to say is that even if there is logic and previous data to back up a decision, that we can still regularly get it wrong
 
Joined
16 Mar 2012
Messages
3,547
Likes
611
AFL Club
Carlton
#35
If you knew a mid was going to average 120 in the first half of the season but only 100 for the second half of the season what would their priced average have to be for you to pick them?

I'd say 101.3 (100k less than a 120 ave priced player).
 
Joined
23 Mar 2012
Messages
19,855
Likes
929
AFL Club
Carlton
#36
If you knew a mid was going to average 120 in the first half of the season but only 100 for the second half of the season what would their priced average have to be for you to pick them?

I'd say 101.3 (100k less than a 120 ave priced player).
I would not want to pick such a player for anything more than $500 K at most. The problem is such a player would be well worth his spot in first half of season when value of overall team is still building but in 2nd half of season when I hope to settle into my mostly premium midfeld line-up his 100 ppg is going to be well short of the scoring I'd want to see being contributed by my M8. The trick would be to make use of such a player and upgrade to a proper midfield premium around the time of the byes and sell him off hopefully before his proce starts dropping right away. If I got such a player for $400 k then it might be worth considering but I don't see much point of entertaining having such a player if he is stealing a role off my rookies or premium midfield spots
 
Joined
16 Mar 2012
Messages
3,547
Likes
611
AFL Club
Carlton
#37
I would not want to pick such a player for anything more than $500 K at most. The problem is such a player would be well worth his spot in first half of season when value of overall team is still building but in 2nd half of season when I hope to settle into my mostly premium midfeld line-up his 100 ppg is going to be well short of the scoring I'd want to see being contributed by my M8. The trick would be to make use of such a player and upgrade to a proper midfield premium around the time of the byes and sell him off hopefully before his proce starts dropping right away. If I got such a player for $400 k then it might be worth considering but I don't see much point of entertaining having such a player if he is stealing a role off my rookies or premium midfield spots
I was meaning that you would trade him out halfway through the season. Essentially the question is how much extra $ to spend on your initial squad would you need to justify picking a player that you would trade out.
 

Rowsus

Statistician
Joined
19 Mar 2012
Messages
28,615
Likes
61,944
AFL Club
Melbourne
#38
If you knew a mid was going to average 120 in the first half of the season but only 100 for the second half of the season what would their priced average have to be for you to pick them?

I'd say 101.3 (100k less than a 120 ave priced player).
Personally, I would take that player at nearly any price. Maybe call 130 my ceiling. How many players can you guarantee 120/game from? Three? His price would still be high enough, that when you traded him out you could get whatever player you wanted outside of those 3 players, and you've got your season off to a guaranteed jump with at least one player anyway.
 
Joined
23 Mar 2012
Messages
19,855
Likes
929
AFL Club
Carlton
#39
I was meaning that you would trade him out halfway through the season. Essentially the question is how much extra $ to spend on your initial squad would you need to justify picking a player that you would trade out.
If the player was around the $400 k mark I would consider him for such an extended pinch hitter role but over $400 k I worry the proft I made on him I could have made with a rookie anyway and the money saved of over $200 k by not getting such a player and picking a mid rookie instead would mean the spare coin I could spend on a good scoring player in another position for 1st half of the season. Any mid priced players I go with I'll be asking myself the question... are they taking the place of a rookie? or a place of a keeper?
If it is a midfield keeper he needs to score 115 ppg for me otherwise I'm not going with him. If he is taking role of a rookie he needs to make me atleast $150 K pofit when I sell him to make me think there is value on spending money on him at the start. It is going to be interesting round one when I actually go over the players I select and see if they all justify getting a gig. So much can change each week until then.
 

Philzsay

Leadership Group
Joined
21 Mar 2012
Messages
10,368
Likes
14,580
AFL Club
Essendon
#40
If you knew a mid was going to average 120 in the first half of the season but only 100 for the second half of the season what would their priced average have to be for you to pick them?

I'd say 101.3 (100k less than a 120 ave priced player).
I'd pay $588,300 for him, the price of someone who averaged 110ppg last year.

The decision on if I trade him mid year would depend on how my team is going, eg the number of trades I have used, the number of weaknesses in the team I still have to fix etc.
 
Top