OK.
So I've decided to try SuperCoach this year and started to get to grips with the numbers and tactics. I've been a long time watcher of Champion Data's technique and ratings, and while there are obviously some problems, I haven't seen a better way yet to objectively value a player's influence on a game.
So anyway, since I've only got 20 trades for the season, I thought I'd really need to make use of some good 'value' buys and trade them up for more obvious premium points scorers over the first 4-8 rounds.
At least 5-10 of my picks are going to have to be used for emergencies such as injuries, poor form or reports, which leaves about 10 up-trades to be done. I think you guys call them "cash cows", where I can buy low, sell higher. So how do I find these cash cows? Because if your going to get anywhere in this game, your starting line-up is seriously important.
Well there's two types of players that qualify here:
1) The player who's never played to his potential AND
2) The player who's coming off injury.
In this post, I'd thought it best to look at data surrounding point 1) and find out when does the average player have their breakout year?
First, you have to ask what a breakout year means? Well, players don't continually improve, statistically speaking, throughout their careers at a consistent rate. Nor do players play to their potential from their first game onwards (except for some rare occasions). There is generally a significant jump in numbers at a certain point (and in some cases this can happen twice for a player). I wanted to know what that looked like and when it would most likely happen, and of course could you identify the signs before they happened.
I looked at every current player who'd played over 100 games (all 153 of them) and determined which year was their breakout year. I used my own criteria to determine it, but it was scientifically rigorous. One thing that should be pointed out was it was solely based on stats, not SC points, therefore take that as you will. So unless the player's efficiency in his breakout year was unusually woeful compared to the previous years, the data should be pretty accurate.
Anyway here's what I found, players' breakout seasons occurred when they were this age before April 1st of the season:
18: 6%
19: 12%
20: 20%
21: 19%
22: 12%
23: 16%
24: 6%
25: 6%
26: 4%
Of those players, I also looked at how many games they had played before the start of their breakout season:
0-10: 15%
11-20: 22%
21-30: 11%
31-40: 13%
41-50: 7%
51-60: 6%
61-70: 8%
71-80: 5%
81-90: 6%
91-100: 4%
101-110: 3%
111-120: 2%
The spread was a lot wider than I'd imagined. For example, I'd never really thought players could breakout past 100 games or over 25 years old. Unfortunately, I was hoping the analysis would be a little more insightful, but to get any interesting patterns, I would need to cross-reference it with a whole lot more data. Height, injuries and teams are obvious ones. For example, I think it was Collingwood that has surprising amount of late bloomers that averaged some 50 games more than most and a couple years older.
At the risk of receiving a "No Shit Sherlock" comment, your most likely candidate for a breakout season is 20 year old who's notched between 11-30 games (and is likely to finish his career over 100 games), but even then, the data says that they're only a 10% chance that this will be their most statistically improved year over their career.
Anyway, maybe I'll look into it a little further and see if I can find some more insightful patterns.
So I've decided to try SuperCoach this year and started to get to grips with the numbers and tactics. I've been a long time watcher of Champion Data's technique and ratings, and while there are obviously some problems, I haven't seen a better way yet to objectively value a player's influence on a game.
So anyway, since I've only got 20 trades for the season, I thought I'd really need to make use of some good 'value' buys and trade them up for more obvious premium points scorers over the first 4-8 rounds.
At least 5-10 of my picks are going to have to be used for emergencies such as injuries, poor form or reports, which leaves about 10 up-trades to be done. I think you guys call them "cash cows", where I can buy low, sell higher. So how do I find these cash cows? Because if your going to get anywhere in this game, your starting line-up is seriously important.
Well there's two types of players that qualify here:
1) The player who's never played to his potential AND
2) The player who's coming off injury.
In this post, I'd thought it best to look at data surrounding point 1) and find out when does the average player have their breakout year?
First, you have to ask what a breakout year means? Well, players don't continually improve, statistically speaking, throughout their careers at a consistent rate. Nor do players play to their potential from their first game onwards (except for some rare occasions). There is generally a significant jump in numbers at a certain point (and in some cases this can happen twice for a player). I wanted to know what that looked like and when it would most likely happen, and of course could you identify the signs before they happened.
I looked at every current player who'd played over 100 games (all 153 of them) and determined which year was their breakout year. I used my own criteria to determine it, but it was scientifically rigorous. One thing that should be pointed out was it was solely based on stats, not SC points, therefore take that as you will. So unless the player's efficiency in his breakout year was unusually woeful compared to the previous years, the data should be pretty accurate.
Anyway here's what I found, players' breakout seasons occurred when they were this age before April 1st of the season:
18: 6%
19: 12%
20: 20%
21: 19%
22: 12%
23: 16%
24: 6%
25: 6%
26: 4%
Of those players, I also looked at how many games they had played before the start of their breakout season:
0-10: 15%
11-20: 22%
21-30: 11%
31-40: 13%
41-50: 7%
51-60: 6%
61-70: 8%
71-80: 5%
81-90: 6%
91-100: 4%
101-110: 3%
111-120: 2%
The spread was a lot wider than I'd imagined. For example, I'd never really thought players could breakout past 100 games or over 25 years old. Unfortunately, I was hoping the analysis would be a little more insightful, but to get any interesting patterns, I would need to cross-reference it with a whole lot more data. Height, injuries and teams are obvious ones. For example, I think it was Collingwood that has surprising amount of late bloomers that averaged some 50 games more than most and a couple years older.
At the risk of receiving a "No Shit Sherlock" comment, your most likely candidate for a breakout season is 20 year old who's notched between 11-30 games (and is likely to finish his career over 100 games), but even then, the data says that they're only a 10% chance that this will be their most statistically improved year over their career.
Anyway, maybe I'll look into it a little further and see if I can find some more insightful patterns.
