Opinion Questions For Rowsus
- Thread starter Rowsus
- Start date
- Status
Hi Rowsus,
Just wondering what your thoughts on Witherden are?
My first reaction is that at 479k, he's somewhat overpriced compared to your usual 2nd year player. In 2017, he only played 9 games rather than 20+ so could you say that he didn't have the usual drop off a first year player would (ie: due to tiring, etc)? For comparison McGrath averaged 77 over his first 10 and finished on 71.
Just feeling that due to his price, he doesn't quite have the upside usual breakout picks do. Even if he breaks out to a 95+, he wouldn't rise that much in price!
Cheers
B18
Just wondering what your thoughts on Witherden are?
My first reaction is that at 479k, he's somewhat overpriced compared to your usual 2nd year player. In 2017, he only played 9 games rather than 20+ so could you say that he didn't have the usual drop off a first year player would (ie: due to tiring, etc)? For comparison McGrath averaged 77 over his first 10 and finished on 71.
Just feeling that due to his price, he doesn't quite have the upside usual breakout picks do. Even if he breaks out to a 95+, he wouldn't rise that much in price!
Cheers
B18
Hi B18,
you and I seem to share similar thoughts on a lot of things.
I feel exactly the same way about Witherden. Looked good while he was playing, but didn't have that chance to go off the boil, like so many first year players do, so we'll never know if he could carry it through the season. There has to be a question mark over him, and he is priced so there is basically no buffer, if things go wrong. Basically, the more likely things are to go wrong, the more buffer you want in their price.
As we all know, the Defs are looking ridiculously hard this season. You only need to look at the ownership of the Defs priced $400k+ to see this:
Laird 43%
Hibberd 20%
Yeo 18%
Hurley 13%
and they are the only ones over 10%. It shows people are guessing, going wide, or going thin in Def this year, for now. I admit myself, my initial team had a 1-0-7 structure, as it is just so hard to pin down the ones you are willing to pay top price for.
I was a lttle surprised Witherden was only in 5% of teams, as I thought he might be higher, even with people shying away from 2nd year players. I wonder if the arrival Hodge affects Witherden's role much? Will he get a little Midtime? Who knows!
His Disposal Efficiency is up there, running at 74% last season. He did also miss Round 21 (Hamstring), and his games either side weren't brilliant, so maybe that was his mini "first year fade" right there.
With all the uncertainty surrounding who will be the better Defs this season, I can understand some rolling the dice on Witherden. I do agree with you however, that there doesn't appear to be a lot of upside, and he is priced to what you'd expect that upside to be. If you don't start him, and he bangs out some higher than expected scores, I don't think I'd be chasing him. If he looks like being that fairly consistent 90-100 averaging 95 type player, we can get him whenever we want. I have had him in and out of my team, and he is currently out. I leave the door open for the possibility for him to return by Round 1, but the fewer players we are guessing on in our starting Keepers, the better.
you and I seem to share similar thoughts on a lot of things.
I feel exactly the same way about Witherden. Looked good while he was playing, but didn't have that chance to go off the boil, like so many first year players do, so we'll never know if he could carry it through the season. There has to be a question mark over him, and he is priced so there is basically no buffer, if things go wrong. Basically, the more likely things are to go wrong, the more buffer you want in their price.
As we all know, the Defs are looking ridiculously hard this season. You only need to look at the ownership of the Defs priced $400k+ to see this:
Laird 43%
Hibberd 20%
Yeo 18%
Hurley 13%
and they are the only ones over 10%. It shows people are guessing, going wide, or going thin in Def this year, for now. I admit myself, my initial team had a 1-0-7 structure, as it is just so hard to pin down the ones you are willing to pay top price for.
I was a lttle surprised Witherden was only in 5% of teams, as I thought he might be higher, even with people shying away from 2nd year players. I wonder if the arrival Hodge affects Witherden's role much? Will he get a little Midtime? Who knows!
His Disposal Efficiency is up there, running at 74% last season. He did also miss Round 21 (Hamstring), and his games either side weren't brilliant, so maybe that was his mini "first year fade" right there.
With all the uncertainty surrounding who will be the better Defs this season, I can understand some rolling the dice on Witherden. I do agree with you however, that there doesn't appear to be a lot of upside, and he is priced to what you'd expect that upside to be. If you don't start him, and he bangs out some higher than expected scores, I don't think I'd be chasing him. If he looks like being that fairly consistent 90-100 averaging 95 type player, we can get him whenever we want. I have had him in and out of my team, and he is currently out. I leave the door open for the possibility for him to return by Round 1, but the fewer players we are guessing on in our starting Keepers, the better.
Hi Rowsus,
I am wondering about your thoughts on Lachie Neale this year. I am drawn to his pod status and the fact that i like him as a footballer but I can't help but feel with Fyfe potentially back to his best it will be hard to justify having Neale in at his expense. Can you shed any light on how you think Neale will shape up this year and if he his worth the punt? Or even if running both Fyfe and Neale in my team will be a worthwhile endeavour?
Cheers
I am wondering about your thoughts on Lachie Neale this year. I am drawn to his pod status and the fact that i like him as a footballer but I can't help but feel with Fyfe potentially back to his best it will be hard to justify having Neale in at his expense. Can you shed any light on how you think Neale will shape up this year and if he his worth the punt? Or even if running both Fyfe and Neale in my team will be a worthwhile endeavour?
Cheers
Thanks for the response.
Interesting point you make about the most owned DEFs so far. It'll be interesting to follow it over the JLT.
I agree with that bolded point. It's unlikely he goes elite (ie: 105+) in just his second year so we can just get him any time due to his already high price. I'll probably pass on him as a starter because of this.
Interesting point you make about the most owned DEFs so far. It'll be interesting to follow it over the JLT.
I agree with that bolded point. It's unlikely he goes elite (ie: 105+) in just his second year so we can just get him any time due to his already high price. I'll probably pass on him as a starter because of this.
Hi JLK5,
I think he's potentially a great, and very smart pick. Any player that has only missed 2 games in 4 seasons, and PIT70'ed 109.9 over the last 2 seasons, yet only appears in 1.9% of teams, is definitely worth consideration! I certainly wouldn't let having Fyfe stop you from getting Neale. We have seen many times where 2 players from the same Midfield can work out to be great picks, and even a couple of examples of 3 from the same Midfield. There hasn't been many games where the two of them have been at peak fitness, in the last 2 seasons. From Rounds 1 to 5 last season, they were both on about 114. If they can both stay fit (and in Ross' good books), I can't see why they can't be good picks in 2018.
I think he's potentially a great, and very smart pick. Any player that has only missed 2 games in 4 seasons, and PIT70'ed 109.9 over the last 2 seasons, yet only appears in 1.9% of teams, is definitely worth consideration! I certainly wouldn't let having Fyfe stop you from getting Neale. We have seen many times where 2 players from the same Midfield can work out to be great picks, and even a couple of examples of 3 from the same Midfield. There hasn't been many games where the two of them have been at peak fitness, in the last 2 seasons. From Rounds 1 to 5 last season, they were both on about 114. If they can both stay fit (and in Ross' good books), I can't see why they can't be good picks in 2018.
Hi Darkie,
I have a mental note on MN histories, but not a written history.
I've never found an explanation or formula for the change in the MN week to week, but I'm pretty sure it involves comparing the weeks Total Expected Score to the weeks Total Actual Score.
Just to define those two things, for people that may be unfamiliar with them.
Total Expected Score would be the raw price (ie not the rounded to nearest $100 price) of all 396 players (in a 9 game Round) in any given week, divided by the current MN.
Total Actual Score is easier, as most games are a total of 3,300 points +/- 3, so it is always going to be around 29,700 (once again, in a 9 game Round).
With the large majority of Rookies being priced to score around 20-25, and even the more expensive ones being priced to around 35-40, we can see that until the Rookies start to rise in price, the TAS will nearly always be greater than the TES.
Just to demonstrate this point, let's look at Round 1, game 1 in 2017:
Total value of players playing (I've used the rounded to $100 figure, as I don't have the raw): $16,944,900
MN: 5,440.6
TES = 16,944,900 / 5,440.6 = 3,114.5
TAS = 3,303
We can see that the TAS is around 6% higher than the TES.
I also believe the total of every players value is a zero sum game ie. the total price rises in any given week = total price falls.
I think what happens is, the MN gets adjusted each week by the difference between the TES and the TAS.
As the Rookies values increase and the majority of other players values fall, and the MN falls, the TES and TAS get closer together, until in about Round 8 the adjustments are very small, and sometimes the adjustment even goes up.
In the years that GWS and GC started, we saw large movements in the MN, due to so many Rookies playing.
In more recent years, the movement hasn't been as pronounced as those years.
So to answer your question:
Do I believe the MN will be higher later in the year, as it started higher this season?
It really comes down to how many Rookies play, particularly early on, and how successful they are. My guess is, yes it will be slightly higher at the end of this season, than it was last season.
I have a mental note on MN histories, but not a written history.
I've never found an explanation or formula for the change in the MN week to week, but I'm pretty sure it involves comparing the weeks Total Expected Score to the weeks Total Actual Score.
Just to define those two things, for people that may be unfamiliar with them.
Total Expected Score would be the raw price (ie not the rounded to nearest $100 price) of all 396 players (in a 9 game Round) in any given week, divided by the current MN.
Total Actual Score is easier, as most games are a total of 3,300 points +/- 3, so it is always going to be around 29,700 (once again, in a 9 game Round).
With the large majority of Rookies being priced to score around 20-25, and even the more expensive ones being priced to around 35-40, we can see that until the Rookies start to rise in price, the TAS will nearly always be greater than the TES.
Just to demonstrate this point, let's look at Round 1, game 1 in 2017:
Total value of players playing (I've used the rounded to $100 figure, as I don't have the raw): $16,944,900
MN: 5,440.6
TES = 16,944,900 / 5,440.6 = 3,114.5
TAS = 3,303
We can see that the TAS is around 6% higher than the TES.
I also believe the total of every players value is a zero sum game ie. the total price rises in any given week = total price falls.
I think what happens is, the MN gets adjusted each week by the difference between the TES and the TAS.
As the Rookies values increase and the majority of other players values fall, and the MN falls, the TES and TAS get closer together, until in about Round 8 the adjustments are very small, and sometimes the adjustment even goes up.
In the years that GWS and GC started, we saw large movements in the MN, due to so many Rookies playing.
In more recent years, the movement hasn't been as pronounced as those years.
So to answer your question:
Do I believe the MN will be higher later in the year, as it started higher this season?
It really comes down to how many Rookies play, particularly early on, and how successful they are. My guess is, yes it will be slightly higher at the end of this season, than it was last season.
Hi Rowsus,
I believe the MN actually stays the same all year, but we change it each round because it makes it easier to calculate individual player pricing.
Each week every players price is calculated as expected but using the same MN each week, (what it is at the start of the year), then the new prices of all players are totaled to give a new total player price value. This total value needs to stay the same each week, so some sort of inflation or deflation percentage is applied to everyone's price to make the total the same each week.
ie is Total player is 100M and at the end of the round total is 102M then every players price is multiplied by 0.98039
But because we can't easily calculate the every players price its easier to just apply that unknown percentage to the magic number.
This is why the MN will drop quicker at the start of the year.
Which is essential they same sort of thing you have described above.
I believe the MN actually stays the same all year, but we change it each round because it makes it easier to calculate individual player pricing.
Each week every players price is calculated as expected but using the same MN each week, (what it is at the start of the year), then the new prices of all players are totaled to give a new total player price value. This total value needs to stay the same each week, so some sort of inflation or deflation percentage is applied to everyone's price to make the total the same each week.
ie is Total player is 100M and at the end of the round total is 102M then every players price is multiplied by 0.98039
But because we can't easily calculate the every players price its easier to just apply that unknown percentage to the magic number.
This is why the MN will drop quicker at the start of the year.
Which is essential they same sort of thing you have described above.
- Joined
- 6 Jul 2012
- Messages
- 4,890
- Likes
- 11,150
- AFL Club
- West Coast
Gents - this may be another perspective of MN if I have assumed the question is solely about MN and not some adjustment that was mentioned.
The magic number WILL come down through the year, fastest in the early part of the year as rookies etc play games and build up scoring above their implied score. For MN to go up, it requires rookies not to play and scoring from existing players to be below last years average (unlikely).
Hawk is correct the total player value stays the same. At the start of the year, the algorithim will calculate the magic number based on last years players average and rookies based on average points implied on draft posiition, injured players discount etc. It then calculates the magic number that will lead to total player value.
This is usually circa $5,500 per point. As rookies score early and of courses, existing players score different to last years average, this is recalculated each round. Estbalished players whilst large in size, probably cancel a lot out in terms of up and down movements.
Usually by round 7-8 the MN is close to bottom, usually just over $5,000. It is why when calculating what a rookie will get to, you should use $5,000 as it takes 7-8 rounds (at consistent scoring) for a rookie to get to 90-95% of fair vaulue. It then takes a lot more weeks of the rolling average to extract the last bit of value, without a spike score.
Player value is determined by (75% x current price) + (25% x rolling 3 round average x MN).
If you compare this to dreamteam/fantasy, they decided to make the rookies more expensive a few years ago. Effectively they increased the implied points average based on draft position. For instance, if Herald Sun decided they want to change the bottom rookies (pre round 1) implied points to 30 points not say 18 points, then a rookie price may go up from $102k to $165k at the start (rough numbers). This would reduce the MN before season start, reducing the $$ price of all other established players.
Its effect need to be bear in mind that all rookies may only make up 5-10% of points and hence 50% lift up on rookie who make up 5% of points, would only affect established player value (downward) by 2.5% at start of year (I think!).
The magic number WILL come down through the year, fastest in the early part of the year as rookies etc play games and build up scoring above their implied score. For MN to go up, it requires rookies not to play and scoring from existing players to be below last years average (unlikely).
Hawk is correct the total player value stays the same. At the start of the year, the algorithim will calculate the magic number based on last years players average and rookies based on average points implied on draft posiition, injured players discount etc. It then calculates the magic number that will lead to total player value.
This is usually circa $5,500 per point. As rookies score early and of courses, existing players score different to last years average, this is recalculated each round. Estbalished players whilst large in size, probably cancel a lot out in terms of up and down movements.
Usually by round 7-8 the MN is close to bottom, usually just over $5,000. It is why when calculating what a rookie will get to, you should use $5,000 as it takes 7-8 rounds (at consistent scoring) for a rookie to get to 90-95% of fair vaulue. It then takes a lot more weeks of the rolling average to extract the last bit of value, without a spike score.
Player value is determined by (75% x current price) + (25% x rolling 3 round average x MN).
If you compare this to dreamteam/fantasy, they decided to make the rookies more expensive a few years ago. Effectively they increased the implied points average based on draft position. For instance, if Herald Sun decided they want to change the bottom rookies (pre round 1) implied points to 30 points not say 18 points, then a rookie price may go up from $102k to $165k at the start (rough numbers). This would reduce the MN before season start, reducing the $$ price of all other established players.
Its effect need to be bear in mind that all rookies may only make up 5-10% of points and hence 50% lift up on rookie who make up 5% of points, would only affect established player value (downward) by 2.5% at start of year (I think!).
Last edited:
Interesting!
You are both proposing models which I don't use myself. I leave room for the possibility that my model is flawed.
I was just wondering, what model do you use/propose is being used for price adjustments?
I have seen this model of price adjusting before, but have found on application, that my model worked more accurately to what transpires when SuperCoach re-opens, and the price adjustments are revealed.
My model, which I am sure you have seen before, is:
New price = old price + (score - B/E)*440
B/E = 3*price/MN - most recent score this season - most recent score before that, in this season.
You are both proposing models which I don't use myself. I leave room for the possibility that my model is flawed.
I was just wondering, what model do you use/propose is being used for price adjustments?
I have seen this model of price adjusting before, but have found on application, that my model worked more accurately to what transpires when SuperCoach re-opens, and the price adjustments are revealed.
My model, which I am sure you have seen before, is:
New price = old price + (score - B/E)*440
B/E = 3*price/MN - most recent score this season - most recent score before that, in this season.
- Joined
- 6 Jul 2012
- Messages
- 4,890
- Likes
- 11,150
- AFL Club
- West Coast
I have seen this model of price adjusting before, but have found on application, that my model worked more accurately to what transpires when SuperCoach re-opens, and the price adjustments are revealed.
My model, which I am sure you have seen before, is:
New price = old price + (score - B/E)*440
B/E = 3*price/MN - most recent score this season - most recent score before that, in this season.[/QUOTE]
May well be that mine is flawed Rowsus. Might be my model was used to explain it to the masses and may not be 100pct correct. I find this model is very easily to apply in a formula when projecting forward. I started in dreamteam and it may well be that is where I picked up the formula.
As I am not getting hit up for lots of questions, something that easily gets me to within the number is fine for me. Hopefully my comments ex the calc helped the discussion and not cloud it.
Main point of my comment was that the MN comes down and why it comes down, which I think everyone was in agreement on although we all seem to get there via different paths.
Keep up the great work, appreciate the explanations you have been providing.
My model, which I am sure you have seen before, is:
New price = old price + (score - B/E)*440
B/E = 3*price/MN - most recent score this season - most recent score before that, in this season.[/QUOTE]
May well be that mine is flawed Rowsus. Might be my model was used to explain it to the masses and may not be 100pct correct. I find this model is very easily to apply in a formula when projecting forward. I started in dreamteam and it may well be that is where I picked up the formula.
As I am not getting hit up for lots of questions, something that easily gets me to within the number is fine for me. Hopefully my comments ex the calc helped the discussion and not cloud it.
Main point of my comment was that the MN comes down and why it comes down, which I think everyone was in agreement on although we all seem to get there via different paths.
Keep up the great work, appreciate the explanations you have been providing.
How I think its calculated which I think will give the same as your formula just a different way of going about it, is the standard formula we all know.
(Starting Price x 0.75) + (3 game average x MN x 0.25) = unadjusted player price, but I think the MN stays the same all year and stays at what it was at the start ie $/point value.
After you have calculated the new price for all the players you total them to get a new total player value and compare this to the original total player value at the start of the year, what ever the percentage difference is you then apply this to the unadjusted player price to give you the new price.
The idea of the MN changing each week is just to make it easier to calculate players price as if you were to apply the total player price percentage difference to the MN you get the same result.
Your zero sum game idea isn't something I've thought about before but essential its the same concept, the price rises and falls need to balance and a percentage is applied.
(Starting Price x 0.75) + (3 game average x MN x 0.25) = unadjusted player price, but I think the MN stays the same all year and stays at what it was at the start ie $/point value.
After you have calculated the new price for all the players you total them to get a new total player value and compare this to the original total player value at the start of the year, what ever the percentage difference is you then apply this to the unadjusted player price to give you the new price.
The idea of the MN changing each week is just to make it easier to calculate players price as if you were to apply the total player price percentage difference to the MN you get the same result.
Your zero sum game idea isn't something I've thought about before but essential its the same concept, the price rises and falls need to balance and a percentage is applied.
Hey Rowsus, I'm warming to the idea of Lycett at R3. I've got a team I'm reasonably happy with that can fit this unusual structure in, but am I barking up the wrong statistical tree and is the extra outlay too much for the potential for more ruck security over the season?
My theory is that if Lycett can save me a trade in the rucks over the course of the season, that'll be worth the extra 150k I'm spending on him now.
My theory is that if Lycett can save me a trade in the rucks over the course of the season, that'll be worth the extra 150k I'm spending on him now.
Hey Eagling,
I know people will howl me down, particularly the "traditionalists", but I actually think the idea has potential merit.
The "traditionalists" will tell you, you can't have $277k sitting on the bench. I say, given we are looking at potentially driest season ever for Ruck coverage, that the "traditionalists" are thinking of the good old days when we nearly always had a sub $150k Ruck we could quite confidently sit at R3.
Having said the idea has merit, I need to point out, on your proposed scenario, the idea is not worth pursuing.
You have actually said, in different words:
"If I was offered an extra trade, and all I had to do was sacrifice $150k of my salary cap in Round 1, I'd do it".
Let's do the maths a slightly different way than normal:
$150k is 1.5% of your starting salary cap. You are hoping to start with something like 2,200 points/round
2,200 x 1.5% x 22 Rounds = 726 points.
You are giving up a theoretical 726 points, to save a trade. It would want to have been one REALLY good trade to warrant that!
So why do I say the idea has some merit?
It's because I don't see it as saving a trade.
I see it as potentially saving a chain of events, that leads to potentially 3 or 4 trades.
I see it producing points, and dollars.
The question then becomes, will those points and dollars (and potential saved trades) get in the neighbourhood of the 600-750 points required, to call it a worthwhile exercise?
I see it saving trading out a player that might be someone you actually want to keep. ie, let's say you start Gawn/Goldy, and it turns out they are the 2 Rucks you needed in 2018.
When one of them misses a game, are you going to trade them out to a sub-performing commodity?
Are you then stuck with the quandary of trading them back in, to rid yourself of this sub-performing commodity?
I think one of things last season shone a light on is, we need to start thinking outside the box a little, and we also need to question some of our traditional thinking.
Tradition without solid reason is just ridiculous. Tradition for traditions sake.
I cite my Danish families Xmas dinner as an example. As in most places around the Christian world, we eat too much, to the stage of discomfort. What I fail to wrap my head around is, within this meal, we each have to eat a bowl of risengrød, which is basically rice porridge. Each year I point out to my wife, that people did this in the olden days, because they didn't have the financial means to fill the table with hams and turkeys and all the trimmings we have today. She says "I know, but it's tradition". No one particularly enjoys it. No one says "Ooooh, I can't wait for my big bowl of risengrød!". People silently suffer through it, all thinking the same thing, I wish we didn't have to do this........ but it's tradition.
Tradition without reason and understanding is just nonsense.
That applies just as much to SuperCoach, as it does Danish Xmas dinners!
I know people will howl me down, particularly the "traditionalists", but I actually think the idea has potential merit.
The "traditionalists" will tell you, you can't have $277k sitting on the bench. I say, given we are looking at potentially driest season ever for Ruck coverage, that the "traditionalists" are thinking of the good old days when we nearly always had a sub $150k Ruck we could quite confidently sit at R3.
Having said the idea has merit, I need to point out, on your proposed scenario, the idea is not worth pursuing.
You have actually said, in different words:
"If I was offered an extra trade, and all I had to do was sacrifice $150k of my salary cap in Round 1, I'd do it".
Let's do the maths a slightly different way than normal:
$150k is 1.5% of your starting salary cap. You are hoping to start with something like 2,200 points/round
2,200 x 1.5% x 22 Rounds = 726 points.
You are giving up a theoretical 726 points, to save a trade. It would want to have been one REALLY good trade to warrant that!
So why do I say the idea has some merit?
It's because I don't see it as saving a trade.
I see it as potentially saving a chain of events, that leads to potentially 3 or 4 trades.
I see it producing points, and dollars.
The question then becomes, will those points and dollars (and potential saved trades) get in the neighbourhood of the 600-750 points required, to call it a worthwhile exercise?
I see it saving trading out a player that might be someone you actually want to keep. ie, let's say you start Gawn/Goldy, and it turns out they are the 2 Rucks you needed in 2018.
When one of them misses a game, are you going to trade them out to a sub-performing commodity?
Are you then stuck with the quandary of trading them back in, to rid yourself of this sub-performing commodity?
I think one of things last season shone a light on is, we need to start thinking outside the box a little, and we also need to question some of our traditional thinking.
Tradition without solid reason is just ridiculous. Tradition for traditions sake.
I cite my Danish families Xmas dinner as an example. As in most places around the Christian world, we eat too much, to the stage of discomfort. What I fail to wrap my head around is, within this meal, we each have to eat a bowl of risengrød, which is basically rice porridge. Each year I point out to my wife, that people did this in the olden days, because they didn't have the financial means to fill the table with hams and turkeys and all the trimmings we have today. She says "I know, but it's tradition". No one particularly enjoys it. No one says "Ooooh, I can't wait for my big bowl of risengrød!". People silently suffer through it, all thinking the same thing, I wish we didn't have to do this........ but it's tradition.
Tradition without reason and understanding is just nonsense.
That applies just as much to SuperCoach, as it does Danish Xmas dinners!
- Status