Small or Far Away, Father Dougal on Cows
So, it is still a week or two too soon for the normal cow projections, but I should somehow involve cows in Cow Talk. So, hmm. What do you call a lot of cows?
Right! And what do herds often have?
I asked for that. I was looking for “behavior.” As in herd behavior. Something that seems common at this point every season. And why is that? Because we are so excited to get new data that we get, well, excited, and give it too much weight. Three weeks of scores is nowhere near enough to change our thinking about players for which we already have decent data. Rookies are different, in that we start with little data and each of the first weeks is a non-trivial addition to what we know. But, non-rookies, most of them we know a lot about, and yet we throw it out the window. That makes rounds 3-5 the “Second Defenestration”
Defenestration? And what’s the first one?
To defenestrate means to throw out a window. Awesome word. The First Defenestration is when team are announced, and our pre-season plans go out the window. The Second Defenestration is now Rounds 3-5, when we throw common sense out the window, and over react to the small data samples.
I wrote in the preseason about the concept of True Level of Ability, called TLS from here out.
It basically means how good a player really is and what he would score with no luck factor. Since there is a luck factor, players will score super differently quarter to quarter, a lot differently match to match, and some differently over a season. Once a player has a decent scoring history, and more so each season past 25, we can have a good idea of what to expect from them based on their past performance. The farther over 30 they are, the more likely a drop from their past establish level.
So, for example, Dangerfield. He is 29, so I would not expect his to drop due to age.
He has established he can average 130+ and last season he was hurt and still scored 121.7. After his bye, when he was fit, he averaged 132.9. So, before this season, we would expect him to average 130 or a bit more at the end. He is currently at 126.3, so really that expectation has not changed. He should average about 130 from here on out. What if he was currently averaging 150? Then we should expect him to average 130 from here on out. If he was currently averaging 100, we should expect him to average about 130 from here on out. His final season average if he was scoring 150 now would go up to 132.7, and if he was at 100 it would go down to 125.9. But the scores that have moved his final season average have already happened. From here on out, expect a 130. But often we are so excited to get data, we treat the first few weeks as what a players TLA is, even if it is way off what we calculated it was preseason. Then we act on the idea their TLA is different than we thought.
Danger Case 1 (I love calling these Danger Cases, and send thanks to Patrick’s ancestors for facilitating my joke.) He is averaging 150. “Oh no, I must get him in now or I will fall behind” His price has gone up so you are paying more than the people who got him before, and if you are buying R4-5 a lot more. But, you have him. Then he goes 130 the rest of the season, and you have overpaid. Darn.
Danger Case 2 He is averaging 100. “Oh no, I must get rid of him now or I will fall behind” His price has gone down so you are taking a loss, and if you are selling R4-5 a big loss. But, you are rid of him. Then he goes 130 the rest of the season, and you have lost money and points. Darn.
Of course this year Dangerfield is under-priced due to last season’s injury affected scores. Makes him a not ideal example. Lets try something practical for this season.
I wonder what Lachie Neale’s TLA is?
Neale has played 128 games before this season and is 25 years old. An old 25, too, 25 and 10 months. Not impossible for a breakout years but no longer likely. He did go to a new team. We can be open minded. So, What is the value of “X” in the above table? Before the season I would have said 110. Maybe, 115. I have this change by 5 habit, he is about 25, so give him a bump. So 115. Could we have said 120? Not without a lot of optimism. New team, maybe a better role, no Fyfe. How about 125; a 15 point increase? Not liking it. If have been ok with 110 or 115. Yet, people are treating his current average as if it was his TLA. It isn’t.
Lets try a little experiment. Anyone who wants to bet me $100 that Lachie Neale will have an average of 148, no let’s give a bit, 145.0 or more at the end of the season, assuming he plays at least 10 matches, step right up.
And, what was your reaction to the idea of making that bet? There are two sane ones. “Not a chance”, and “I like FD so much I will make it as an excuse to give him $100.” Because you know darn well he won’t average 145 for the season. He won’t average 135. He is unlikely to manage 130. Anyone notice Lachie being just as great as Gary Ablett Jr. previously in his career? As great as Paddy D? He is good, but after going near dead on 110 from age 22-24 he’s going to have a huge jump this year? Riiiight.
Neale is now priced to average close to 120. If you buy him now, and he goes 120 from here on out, which is a 10 point jump from his established TLA, you are getting him at cost. If he goes less, you overpaid. If he goes more, and 125 would be a great rest of season average for him, then you are getting a small bargain. IF he manages to score his average of 148 next week, he will be priced to average about 128, which is anywhere from a little high to a lot high.
Then there is the trade cost. If you are bringing him in to replace an injured guy, well, you had to trade anyways. If not, you are probably using a trade to overpay for someone.
Let’s look at what I call the “Bending over with no lubricant” case. Clayton Oliver has had a rough start to the season. Priced to average 114.7 but currently averaging 102.3. Some dropping of price has happened and could more.
So, what should the value of “Y” be? He is only 21! Preseason I would have put it at 115. If it turned out to be 120 I would not at all be surprised. I would not be shocked at 125. Having a 111.5 average at 19 and a 114.7 average at 20 is way off the curve, and he has a big upside. The only player I know of he is behind is, one guess, Gary Ablett Jr. who had a 132 average at 20. Oliver is far ahead of Dangerfield. Do you know what Oliver’s TLA is not likely to do at age 21? Drop. It is very likely to go up. But “Oh no, Oliver is under-performing and has to go.” You sell him, at a loss. Then he scores 115-125 the rest of the season. Combine that with overpaying for Neale, and you have managed to use a trade and spend extra cash to swap players who will score about the same for the rest of the season. And the one you gave up is more likely to have the higher average from here on out.
Let’s look at this another way. There are 10 players averaging over 120 right now. Do you think there will be 10 players averaging over 120 at the end of the season? Me either. There were only 11 players over 110 at the end of last season. There were 5 over 120, and only 1 over 130 when it was over in 2018. A lot of players are going to score at or below their TLA from here on out, and drop off. Who is over 120 currently?
Neale – More likely to be under 120 then over 120. He does have three high scores to give him a leg up. If you do not have him and bring him in now, and he averages 120, he will have averaged 115.6 for you. An average of 115.6 from here out is not crazy high, thus 120+ for the season is not crazy.
Cripps – Easy to see him over 120 at the end of the season. He is 24 and was at 119.4 last season.
Macrae – Easy to see him over 120 at the end of the season. He is 24 and was at 127.1 last season. Could be over 130.Lloyd – Not likely. He is 25, and 112.0 last season.
Whitfield – Probably not. He is 24, and 99.9 last season. I expect him to do better than 99.9, but not 20 points better.
Dangerfield – Easy to see happening, again.
Bontempelli – Ohh, would not be a shock. He is 23, which bodes well for a breakout, and has been held back by too much forward time in the past. I would not bet on it, nor would I bet against it.
Boak – If you think Boak will have a 120 average at the end of the season I have a bridge to sell you. No, that’s not fair, if you think he will average 110 at the end of the season I have a bridge to sell you. He is 30, and averaged 88.0 last season. If you started with him, take the money are run in a week or two.
Cunnington – No. Less insane an idea than Boak, but that’s like saying less insane than the Joker. He is 27, averaged 96.3 last season.
Sloane = No. Sorry, hate me, but no. He is 29, and never managed to average over 114.8 in the past. I won’t try to sell you a bridge if you disagree, but it would be a serious surprise. He did average 114.8 once, in 2014, which is not a lot to hang your hopes on. Now if you started with him, well done. You got a bargain. If you did not start with him, you missed the bargain. Find a current bargain.
And Odds are someone lower than 120 now will hit it, maybe more than one person. Find the person(s) they are who you want. Not the over performers.
So, well, there you go. I hope that was helpful.
Thanks for reading!
Leave a comment / Scroll to bottom