mattnz wrote:Ok, POD theory.....
Lets say you have 2 possible teams. One is cookie cutter, basically take all the most owned players and assume conventional wisdom that the masses are choosing the right players generally and that is why they are popular. You have a team of 21 players that all have high ownership.
By contrast I am hoping to have lets say 10 PODs, with 11 players in common with the other team (not just choosing PODs for the sake of it, but if I have 2 equal choices, I would go with the POD). Lets assume each team performs just as well as the other and of the 10 players that are different in each team, 8 do great and 2 not so great in round 1. I don't feel a compulsion to suddenly add lots of non-PODs. Those that dont have my PODs may start getting a couple of them in, but they arent going to get all my 8 PODs in, especially while they are increasing in value over the first few rounds. At most they could only get 2 per round and they are still less likely to be selected than the equally well performing non PODs in the other team.
I am only looking to win the comp, not just get top 1000, top 100 or whatever. Having differentiation assists this goal. If it was entirely random like a lottery, you don't want to have the same numbers (players) as everyone else.
This article discusses this phenomenon, where if you chose the numbers 1,2,3,4,5,6 that there are 10,000 others that choose this same sequence every week. https://www.lottoland.co.uk/magazine/why-you-should-avoid-the-most-popular-lottery-numbers.html
If you won first prize, you would win very little money because you would be sharing first prize so many ways. Instead of 5 million pounds on your own, you get 500 pounds each.
It isnt that the chances of hitting these numbers is any more or less likely than any other sequence, but that their popularity reduces what you win.
Applying this game theory to Fantasy, selecting PODs doesn't make them any more likely to score higher, but means if I hit the jackpot and all my player numbers come up as winners, the rarity of my player pool amongst all other players makes it far more valuable.
Lets say that the chances of my first round players all killing it in round 1 and getting 1,000 points is equal in both the 10 POD scenario as in the 0 POD scenario. If I get it with the 0 POD team, everyone goes wow what a huge round, everyone scored amazingly well. There are likely 50 other teams over 1000 that same round and yeah I have done well, but there are others around me.
In the alternative scenario, my 10 PODs absolutely kill it and I get 1,000 points. My unique lottery numbers that no one else has picked all came up at once and I have 100 point lead in first place.
Same likelihood of happening, different outcome if it does happen.
Thats the POD theory.
I understand where you’re going with this, and I’m sorry for not having been clearer in my previous response. I was rushed and just lent on jargon instead of communicating effectively.
Your model is potentially correct for a single game, like the lottery (although I disagree with the attribution of probability, see below). But it needs to be a more complex model for a repeated game like NRLFF. This is because it is not sufficient to win once, but what is required is to win after 20-odd rounds.
Let’s set up some boundary points that reflect the discussion so far. These apply before games are played, and player outcomes can be observed:
1) Popular players exist
2) There exists a ‘wisdom of crowds’ mechanism that means the probability of popular player having a desirable score/price ratio is greater than an unpopular player. Popularity is therefore positively linked with desirable score/price ratios (Kikass from 2018).
3) (corrolary) There will be unpopular players who will also eventuate as having desireable score/price ratios (AFB, Holmes).
4) Path dependence exists, so that a good start by your team is required for a good finish, but is not sufficient, as:
5) Strategic responses also exist, so that advantage can be eroded by teams changing to pick up revealed value (eg Chompson, Maloney 2018)
What I was trying to put forward in the finance analogy was the need for any overall winner to be strongly similar to the wisdom-of-crowds optimal group, but with some points of variation. This argument is a result of the repeated-game situation, where you want to maximise the probability of fully utilising the WOC value (because it has information linked to probably outcomes). But you’ll also want some degree of variation.
The variation should ideally be high-impact, as the FF game takes place with many other players, all following similar strategies. Captaining a back is an example of the high-impact variation, but it fails due to having no positive next-game impact and is stastically unwise (if Teddy scores 30, 30, 30, 100 in every four games on average, then betting on any particular game being 100 returns an average of 70 compared with CS9 50 60 50 60 average of 110). A better strategy would be VK’s Fitz gamble last year, where a high-scoring forward with scoring volatility was added. But I digress.
The benefit of the high early season POD strategy is point 4. A good start matters, in cash, points and zero-trade opportunities. These are huge benefits.
The costs of the strategy are points 2 and 5: it probably won’t work, and if it does then other players can respond and narrow the gap. This gap will also likely automatically narrow due to point 2 - the non PODs are probably better. These are huge costs.
I suspect the strongest argument in support of my argument is the cookie-cutter nature of the top 20 final teams last year.
That said, I strongly suspect my wife will be running a POD team from the start, mostly because its fun!
Home
