Game Theory and Fantasy Sports

The more I read the moreShanbon makes sense. If you want a good ranking every year play it safe. To win you need to take risks and there isn't really a difference between rd1 and Rd 19 except that taking a chance in rd1 doesn't automatically cost you a trade

I'm not sure how useful the published ownership percentages are if we're trying to apply game theory to season fantasy. What would be more useful would be the percentages of the top 1000 players - the ones you're trying to compete against.

For example, RTS will be fairly high ownership in general but is not appearing in any Round 1 teams of more serious players. But he's probably the player with the highest ceiling in the whole game - certainly among the R1 POD options. There's an argument for starting and captaining RTS from the get go if you want to give yourself the best chance of winning the comp, the assumption being that the others who've made this selection are not going to be skilled enough to compete with your decisions further down the line.

Might try a strategy like this with my wife's team...

That's the issue with it. I don't care who the 90% playing with their mates start with but you can get a general idea from looking here and on other fantasy pages who the serious guys are going with.

Yes but they should be calculated risks. Going for Fifita over Parker is a good risk. Potentially going without some big gun forwards and going for more under utilized mid rangers might not be a smart risk because despite the fact that a lot of people own Parker so he doesn't help you as much when he goes big, if he goes big then that's 70 points you don't have

rhinoceroo wrote:I'm not sure how useful the published ownership percentages are if we're trying to apply game theory to season fantasy. What would be more useful would be the percentages of the top 1000 players - the ones you're trying to compete against.

For example, RTS will be fairly high ownership in general but is not appearing in any Round 1 teams of more serious players. But he's probably the player with the highest ceiling in the whole game - certainly among the R1 POD options. There's an argument for starting and captaining RTS from the get go if you want to give yourself the best chance of winning the comp, the assumption being that the others who've made this selection are not going to be skilled enough to compete with your decisions further down the line.

Might try a strategy like this with my wife's team...

That's the one thing I wish fanhub would let you do when looking at players. And because of the whole team ID thing I assume you can't try pull the data from it into a spreadsheet. I remember my first season playing, I just picked guys who I like as players or ones who were good in real life, never in a million years would someone like Fensom have come into my plans. The casual players will go for the big names and their favourites and so it's hard to really tell what kind of serious ownership guys like RTS, Surgess, Inglis are actually on

Like Krump said you can get an idea of ownership of the serious players by reading forums like this and some decent facebook pages.

Forgoing a big gun + nuffie for 2 mid rangers isnt just game theory it is value based projections. Now there are argument for and against both those options but it really comes down to your research and value you give each player.

Just out of interest sake what would people do?

You have another gun in 17 as steady (c) choice same amount of cap is spent in each option

Player A (projected to score 60ppg and hold cash value, as has done in past years) + Player B (projected to score 23ppg makes $10,000 very little chance of losing money or making money) both players in top 17, both players in 15-20% of serious teams
or
Player 3 (projected to score 40ppg make 100k wont lose money but in your estimation 75% of hitting projection) + Player 4 (projected to avg 45 make 40k volatile player could lose money at stages but 85% will make projection avg over season) both players in top 17, P3 in 4% P4 in 8% of serious teams

So essentially you have 83 points + $10k potentially or 85 points + $60-140k.

There is a point where picking a PoD makes no damn sense..

(Note: I haven't been following the thread. Just based off that last comment.)

A lot of the top players (and people that think they're top players) try and hide their "secret" pods, so I wouldn't trust that for getting a handle on ownership percentage

Random wrote:So essentially you have 83 points + $10k potentially or 85 points + $60-140k.

There is a point where picking a PoD makes no damn sense..

(Note: I haven't been following the thread. Just based off that last comment.)

agreed. with that scenario there is an obvious choice

picking based on ownership and trying to get players who only benefit a small percentage of players works well in daily games where the scoring only effects that one game and you can start fresh on the next. when you apply it to a 26 week contest then you have to consider other factors such as price changes, trades and how you are stuck with that player for a few weeks possibly.

Honestly, I don't see in a long term competition the need to target PoD's. Sure they are nice and if you are the only one on them (At least early) it is great.

My criteria for choosing players is generally the following.

Are they a gun? (More of these guys the better)
Can they potentially become a gun? (The cheaper these guys are, the better.)
Will they make me money? (Again cheaper the better.)
Are they Tim Simona?
NPR's (Oh god please be named on a bench.)
PoD's.

However for a Daily/Weekly competition. PoDs matter, but you shouldn't sacrifice Proven Guns for it, for the most part.

How about we start talking about the first Match Day of the round.

Broncos vs Eels at Parra

The "optimal" squad would be
McCullough (maybe Peats)
McGuire
Parker
Hunt
Roberts
Radradra

The outside backs get a lot of scoring variance, so I think the Roberts/Semi ownerships won't be *that* high. I think in the backs you are better of just trying to work out/guess who is going to grab a few tries rather than worry about ownership.

But if you wanted a sub-optimal choice for victory, ditching Parker would have to be the way to go.

A lot of people will have him, and a lot of those people will have him as captain. The odds of Parker failing aren't high, but if he does fail, and you don't have him, then suddenly you are in a very good position, and if the rest of your team scores well you are a chance.

Just a guess but I'd expect Hunt to have huge ownership as well

Krump wrote:Just a guess but I'd expect Hunt to have huge ownership as well

Possibly, but I think a lot of people will (rightly or wrongly) think he'll carry on how he left off in the grand final

He's suspended for WCC right? So won't have any games under his belt

But given the other options available, picking Parker but not captaining him isn't the worst idea though. Parker has the consistency that he isn't going to score poorly. An average game from him is still at a pretty high level compared to the other options

Beast From The Big East wrote:But given the other options available, picking Parker but not captaining him isn't the worst idea though. Parker has the consistency that he isn't going to score poorly. An average game from him is still at a pretty high level compared to the other options

You gain nothing over the competition by picking him. Wether he scores well or not is to an extant irrelevant.

Milchcow wrote:
How about we start talking about the first Match Day of the round.

Broncos vs Eels at Parra

The "optimal" squad would be
McCullough (maybe Peats)
McGuire
Parker
Hunt
Roberts
Radradra

The outside backs get a lot of scoring variance, so I think the Roberts/Semi ownerships won't be *that* high. I think in the backs you are better of just trying to work out/guess who is going to grab a few tries rather than worry about ownership.

But if you wanted a sub-optimal choice for victory, ditching Parker would have to be the way to go.

A lot of people will have him, and a lot of those people will have him as captain. The odds of Parker failing aren't high, but if he does fail, and you don't have him, then suddenly you are in a very good position, and if the rest of your team scores well you are a chance.

The chances of Parker completely failing are too low though - so you are better off having him but not throwing the (C) on him.

If I was looking at the most optimal sub-optimal choices........I would be looking towards those variant backs and omitting someone like Semi or Hunt.
That or I would keep Semi and throw the (C) on him.

Krump wrote:
You gain nothing over the competition by picking him. Wether he scores well or not is to an extant irrelevant.

But surely given Parker's likelihood of scoring well that it is better to have him than going with a POD as the variation of scoring on the logical "best" pick isn't the same as in other positions. If Parker has an average game he still probably goes for 45-50. You go for a POD like Beau Scott and that would be realistic for a big score for him. Don't captain him and hope that you are making a POD move by captaining someone else but if a POD is really only going to score as well as an average game from the gun then it isn't worth the risk.

Wouldn't it be better to go for PODs in high variation positions such as winger, or in traditionally consistent positions with a lot of quality options like front row.

A lineup of say:
Peats
Thaiday
Parker
Milford (c)
Jennings
Semi

Due to very limited options in a one game tournament it would be hard to pass on a consistent high floor high ceiling player like Parker. You would have to find another player in his position you project to score close, I don't have the stats now but you would need to look at comparing say how many times other option players like gillett out scored him in one off games. You could then work out your chance of winning by educated guess on ownership, I would assume Parker would be in at least 80% of teams with Gillett in maybe 5%

I think this thread is good to try figure out the difference between salary cap DFS and limited option no cap match days

Shanbon wrote:Due to very limited options in a one game tournament it would be hard to pass on a consistent high floor high ceiling player like Parker. You would have to find another player in his position you project to score close, I don't have the stats now but you would need to look at comparing say how many times other option players like gillett out scored him in one off games. You could then work out your chance of winning by educated guess on ownership, I would assume Parker would be in at least 80% of teams with Gillett in maybe 5%

I think this thread is good to try figure out the difference between salary cap DFS and limited option no cap match days

The one game tournament is the key thing to consider in my opinion. American sports, particularly in the NFL DFS, you can easily apply the POD players are better as there are so many more options to choose from, whereas this is done on a one game so you have limited players. In any given NRL week you could easily find 5-10 likely players and probably another 5-10 of risky picks who you could argue have a chance of outscoring Parker from a pool of I don't know 70-80 odd 2rf eligible players. But finding guys who can outscore Parker when given a pool of maybe 10 guys is much harder to do so given his consistent performance levels

The problem with taking Parker and not captaining him - if he scores big (70+) then you need to hope your captain goes bigger otherwise you will not win. That's as low odds as just not picking Parker in the first place


Gillett outscored Parker 4 times last year
Manu Ma'u outscored him twice (although 2 of Manu's top 3 scores were weeks Parker didn't play)

So if you pick Parker and Captain him and he scores well, you still have to beat half the comp. Your chances of winning have not improved much.

If you pick Gillett and captain him, in most cases you won't do well. But if he does outscore Parker (19% chance based on last years stats) then suddenly you are going to have a jump on the rest of the pack.

In comps like this there are no prizes for second place. There is zero value in putting up a credible high score that falls only 10 points short.

Captain Parker every week and you'll do OK. Get yourself a decent score each week.

Take a strategy like Captain Gillett though, and on a week to week basis you will probably struggle,
but in the 2-3 weeks of the year that Gillett goes big then suddenly you look Ok for the money.

Round 1 last year Gillett outscored Parker by 31 points (Hunt by 38, McCullough by 17, Peats by 46) If you captained him, you would have done super well. The big downer would be you'd lose to someone who captained Radradra (14 more than Gillett)

I think if you want to go for the high risk high reward strategy picking a winger as captain is a good way to go too. because the week you pick the guy that scores a hattrick is the week you can win the prize.

If you want to just pick the big scoring players - that is fine, but its not the type of strategy being discussed in this thread.