I realized when trying to write a different post that what I was writing used concepts that I have talked about in the past, and might be a bit confusing for anyone who wasn’t familiar with them. So, I’m gonna do some quick, I hope, summaries of some ideas and terms from past posts.
First, not so useful to coaches but getting mentioned anyhow is the CTAP, or Cow Talk Advisory Panel. That would be me, my Imaginary Interlocutor
Hi! He only thinks I’m imaginary, don’t tell him otherwise!
And my hamster, To Mini-Therion the little beast .666
Hi! I like pumpkin seeds! Squeak!
Sometimes I let them “help” with posts. Sometimes they just chime in, and sometimes I provide beer and I write while they are sleeping it off.
This is not one of those times.
Yes, clearly. Anyways, in no particular order some stuff, summarized:
Breakouts and Frogs
How old a play is matters a lot. Young players tend to score poorly; note the lack of 6 year olds in the AFL. Old players also tend to score poorly; note the lack of 60 year olds in the AFL. The solid safe productive range for players tends to be 23-28. Lots of stuff on players breaking out exists, but not a lot on aging, so here stuff on that. Aging in Supercoach scoring is like playing the old video game Frogger. They try to make it through the season without a big age-related drop off just as the frog tries to make it across the road without getting hit by a beer truck. When a play hits 29 they play on the tutorial level and hardly anyone loses the tutorial, although it can happen. At 30 they play level 0, at 31 they play level 1, at 32 they play level 3, and so on. At some point, if they keep playing, they are going to get hit by a beer truck and their scoring will drop off dramatically. The older they are the more likely it is.
Anything you want to add to that, with hindsight and more data?
Injuries should probably factor in somehow too. Maybe a career full of injuries is like a sticky joystick?
Not like that! Argh, ok….like, getting old with a history of injuries increases the risk of a negative beer truck interaction.
True Level of Ability
Every player has an actual ability level. How well a player scores is based on that True Level of Ability (TLA) but affected by luck and the circumstance in which they played.
Luck can easily account for a 5% variation on a player’s average each season, which means a 100 TLA player averaging 95-105 would be pretty normal. Bigger swings are certainly possible, although over 10%
Circumstances, so many circumstances…..other than age: Position, club fortune, where they play, injury, weather, opposition, bizarre coaching decisions, etc. Trying to understand these is a big part of SC strategy. More midfield time? Solo Ruck? Over that injury. Full pre-season? We talk about these at length and good guesses pay off, bad guesses hurt.
Age, well there seem to be some common age brackets:
0 -20 = Youth
21-23 = Early Career
24-26 = Middle Career
27-29 = Late Career
30-32 = Hanging On
33+ = Victory Laps
Middle and Late Career players are pretty reliable as for scoring, Early mostly reliable with best hope of increasing, and 30+, well they are playing Frogger and tend to drop even when they avoid the beer truck.
I made some other ok but not newsworthy either way projections.
Also, for some reason, I looked at how Nat Fyfe has been affected by his broken leg and projected that onto Tom Mitchell.
|Year||Tom’s Games||Projected averages||Tom’s age||Toms real averages|
Oh my god!
Oh my Pumpkin seeds!
Well…..that’s a pleasant surprise…..far better than I had expected looking at what I wrote at the time….for some reason Titch is in my team this season….
(Until the pre-season groin thing)
Swords, Shields, and the Iron Law of Pods
So, the swords and shields thing I think isn’t super useful, looking back. The idea was swords are players who help you pull ahead of your competition. Shields are players who you have because they are highly owned and have the potential to put you behind the competition if they do well. But, really, that goes against the more basic idea of “Just pick who you think are the best players, and do not worry about who else owns them.” So, not useless but not that important either.
The Iron Law of Pods, however that’s a thing. It is:
“The earlier and better a POD, the less time they remain a POD.”
If a player with low ownership starts doing really well, people are going to bring him in. The longer he does well and the better he does the more people will bring him in, until he is no longer low owned and thus no longer a POD.
The Inverse Law of Pods:
“The worse a POD performs, the longer he stays a POD.”
You pick someone with low ownership and they do badly, their ownership will end up dropping. So, they are a point of difference, but a bad one. Not so good.
The Rusty Curse of Pods:
“After you trade out an under-performing POD, he will most likely start doing well.”
Not a rule, but just the Supercoach gods rubbing it in.
Anyways, taking a player because they have a low ownership is a bad idea, because if they are a good pick, soon others will have them, and if not you have a bad pick. Take players because they are good players or good picks, regardless of their ownership.
Oh and stuff Freo Tragic says: “Basically there’s no use going a POD unless you truly think they are going to be good at SC period.” Yes, that, exactly that is what I was trying to say!
Floors & Ceilings, Basements & Roofs
Terms to simply and quickly describe how a player is priced v their range of likely averages.
Basement priced: A player who is going to average far over what they are priced to average. Or putting it another way, someone who’s worst case average is better than what they are priced to average.
Floor Priced: A player who is unlikely to average significantly lower than what they are priced to average and who has a significant chance of averaging much higher than where they are priced.
Accurately priced: A player who is likely to perform at about the same level as what they are priced to average.
Ceiling Priced: A player who is unlikely to average significantly higher than what they are priced to average and who has a significant chance of averaging much lower than where they are priced.
Roof Priced: A player who is so far over what he could reasonably do that his best case is below what he is priced to average.
I gave some examples, interesting with hindsight.
Basement priced: Jack Ziebel – Oh yeah
Floor Priced: Dunkley, Cripps, and Tom Phillips – Dunks, even with his injury issues was under 2 points off. Cripps, well, wrongety wrong, wow. Tom Phillips. Sigh, what was I thinking? Silly me….thinking how to describe players like him in the future….
Accurately priced: Dustin Martin, well, did fall off a lot but I feel he could have done better so not terrible as a guess at his level. Luke Parker was two over, so yes. Tom Stewart went up 7 so the average group on average was average?
Ceiling Priced: Jake Lloyd was exactly this. His 2020 average of 122.2 was crazy high and he was a terrible pick. Tomahawk dropped by 11.9 back to a very respectable 93.9, but the idea he would keep up, let alone beat a career high average he set at 32, was crazy town.
Roof Priced: Max Gawn lost 19.1 from his average, which given his 2020 139.9 was super high and 11.5 over his previous high score made regression to the mean a near certainty. Given he still went for 120.8 I wont say he was a bad pick, but, those were very expensive points and paying a 1538% premium for points is not a Good Thing. But, Roof priced just meant he was very likely to go down a lot and he did that.
So, apparently I can pick ceiling and roof priced player ok, and Basement ok, but Floor and average not so much. Or I should have thought harder for my examples….well….I think the accurate guys really were around what we might expect, I mean they were not clearly anything else….the floor guys, well….Phillps was silly optimism based on more midfield time…..
I guess guys like him are really average priced but with a chance of circumstance change. Which is not floor priced. So, good to know when I look at some of this year’s players.
Oh, I also did math-stuff to show the benefits of picking floor priced players. Not really a surprise that more points for the same cost is good though.
Ok, there are probably more things that I should have covered but I did not think of, so please let me know in the comments if you are think of any.
Thanks for reading!
Leave a comment / Scroll to bottom