50/50 Strategy Vs. Tournment Strategy (as it Pertains to the Interacti (1 Viewer)

[Charybdis]

OK, I read through the "Cracking" Books. Both excellent reads.

I see that the high level advice is for cash games look for consistant players with high floors. And for tournamnets, look for high ceiling players (and stacks).

My question is this... How do I ascertain floors / ceilings and consistancy / volatility from the interactive values chart?

Thanks.

 

[Rock Lonemilk]

You don't.

 

[Maurile Tremblay]

My question is this... How do I ascertain floors / ceilings and consistancy / volatility from the interactive values chart?

The IVCs aren't really built to distinguish between high-variance and low-variance players.

You can find high-variance pairs of players by clicking on the "Top 20 stacks" and building your lineups around QB-REC combos.

One way to look for high-variance players is to see where the three projectors disagree with each other the most. Those guys are somewhat likely to be of the boom-or-bust variety. (We're going to pretty much agree on the guys who are really easy to project. The guys who are harder to project -- the boom-or-bust types -- we may agree or disagree with each other about.)

But the best way to distinguish between high-variance and low-variance players is to (a) become a generally pretty well-informed NFL fan, and (b) for any given player, ask yourself how comfortable you are predicting his performance within a pretty narrow range. The harder it is to put him within a narrow range, the higher his ceiling will be compared to his mean projection. The easier it is to put him in a narrow range, the higher his floor (but lower his ceiling) will be compared to his mean projection.

The interactive charts can't do this for you, however. You're on your own for that analysis.

The Daily Crusher (app for mobile) does attempt to assign players different floors and ceilings based on their volatility. You might want to play around with that if you haven't already.

 

[Assani Fisher]

Looking at the ratio between projected fantasy points and projected touches(rushes + receptions) is one thing that may help you. When deciding between 2 players of similar value, the one with more projected touches usually has a higher floor/lower ceiling. With that said, I think this is very unimportant. Much more important is understanding correlation vs reverse correlation between your players, which is something you hinted at when mentioning stacking. In top heavy GPPs you want as much correlation as possible, whereas in 50/50s you would rather have some reverse correlation and avoid correlation if possible(although the GPP part is much more important than the 50/50 part of this sentence). In HUs it barely matters at all.

 

[Assani Fisher]

Also I hate that 50/50s and HUs get grouped into "cash games" in these discussion, as they are completely different in this regard.

In HUs, you will likely be rewarded in accordance with the percentile you finished in. In other words, if your LU is in the 68th percentile(i.e. out of every 100 people, your score is better than 67 and worse than 31 of the entries) then you most likely will win 67%, lose 31%, and tie 1%(and then lose a few extra percent to rake). Because you are rewarded in accordance with the percentile you finish in, it is not necessary to worry about anything other than pure value.

In top heavy GPPs, the huge majority of the prizepool goes to the top finishers. In these contests, when you finish in the 68th percentile its the same as finishing in the bottom 1 percentile- you lose everything. As such, you want to sacrifice your average score in order to have the highest possible outlier scores on a select few occasions.

In 50/50s or double ups, all that matters is being in the top ~45 percentile and everyone who achieves this gets the exact same prize. So, using our previous example, if you finish in the 68th percentile here its no better than having the literal #1 score on the entire site. As such, you want to sacrifice the sometimes huge scores in order to most consistently finish in that top 45th percentile.

 

[17seconds]

I always thought it would be a good idea to have a variance measure in the projections, even before there was DFS. Maybe a color coding by level of standard deviation.

Of course the timeline for each player would be totally different so it would be subjective as well.

 

[Charybdis]

Bringing this back to life a little, I was reading the "Consensus Value Plays" and saw this

Danny Tuccitto - Cook has a top-ranked 60.8% chance of achieving GPP value and a second-ranked 67.3% chance of achieving cash value. (Thu 4:37 AM)

Is there a place I can get those numbers for all players?

 

[Scoresman]

Bringing this back to life a little, I was reading the "Consensus Value Plays" and saw this

Danny Tuccitto - Cook has a top-ranked 60.8% chance of achieving GPP value and a second-ranked 67.3% chance of achieving cash value. (Thu 4:37 AM)

Is there a place I can get those numbers for all players?

Yeah, that right there is the science of statistics in action, my friends. Would love to see the backup behind these numbers.

 

[KarmaPolice]

My question with the tourney entries is how "weird" do you guys get with them?

Seems like I am having troubles finding that balance of uniqueness vs. no chance what-so-ever of scoring decent point to win $. Do you just try to find that one or two guys cheap that you think will blow up or do you put in more than that?

This week I have 2-3 lineups that I like for a tournament, but there are probably only 2 "out there" picks on the rosters. The rest would probably be higher % owned popular players. Just overthinking if I have to tweak it more, or just run with it.

 

[astukas]

My question with the tourney entries is how "weird" do you guys get with them?

Seems like I am having troubles finding that balance of uniqueness vs. no chance what-so-ever of scoring decent point to win $. Do you just try to find that one or two guys cheap that you think will blow up or do you put in more than that?

This week I have 2-3 lineups that I like for a tournament, but there are probably only 2 "out there" picks on the rosters. The rest would probably be higher % owned popular players. Just overthinking if I have to tweak it more, or just run with it.

Personally, I think people get way too weird with their GPP lineups. Over the course of the week on message boards like this, people will throw out absurd names and say well "Its GPP so its worth a shot".

Don't feel like you have to take guys that are out there just for the sake of taking guys that are low owned. You should be more focused on correlating your plays. And if you want to be contrarian, sometimes the easiest way to do that is by taking really good players that people are overlooking.

For example, this week there are some value RBs that everyone loves (Williams and Freeman). You are going to see a lot of lineups with one of these guys, some with both. A good contrarian move would be to pay up at both RB positions and spend less on WRs. This should give you some low owned players without forcing you to take a 3rd string WR and hoping he has the game of his life.

 

[KarmaPolice]

My question with the tourney entries is how "weird" do you guys get with them?

Seems like I am having troubles finding that balance of uniqueness vs. no chance what-so-ever of scoring decent point to win $. Do you just try to find that one or two guys cheap that you think will blow up or do you put in more than that?

This week I have 2-3 lineups that I like for a tournament, but there are probably only 2 "out there" picks on the rosters. The rest would probably be higher % owned popular players. Just overthinking if I have to tweak it more, or just run with it.

Personally, I think people get way too weird with their GPP lineups. Over the course of the week on message boards like this, people will throw out absurd names and say well "Its GPP so its worth a shot".

Don't feel like you have to take guys that are out there just for the sake of taking guys that are low owned. You should be more focused on correlating your plays. And if you want to be contrarian, sometimes the easiest way to do that is by taking really good players that people are overlooking.

For example, this week there are some value RBs that everyone loves (Williams and Freeman). You are going to see a lot of lineups with one of these guys, some with both. A good contrarian move would be to pay up at both RB positions and spend less on WRs. This should give you some low owned players without forcing you to take a 3rd string WR and hoping he has the game of his life.

Good points, and that is what I was more leaning towards I guess. If you have basically the same roster, but instead of Williams go with another $3400 or less back that hits and is hardly on any rosters that is a good thing. Or, If I think everybody is going Rodgers and the GB passing attack it is a good play to go the other way and stack Lacy and GB def. Still not starting scrubs, but going a bit against what I think the majority is doing.

 

[Maurile Tremblay]

If you have basically the same roster, but instead of Williams go with another $3400 or less back that hits and is hardly on any rosters that is a good thing.

You don't just have to hope that your $3,400 guys hits. You have to hope that your guy hits and Williams doesn't. Specifically, if Williams is 50% owned and your $3,400 running back is 5% owned, you have to hope that your guy will hit while Williams doesn't at least ten times [edit: I mean about 10%] as often as Williams hits and your guy doesn't.

I personally don't see anyone in the $3,400 range who I think satisfies that criterion. Maybe Lance Dunbar ($3,600) or Bilal Powell ($3,700) since DK is full-PPR, but I still think Williams is a huge favorite over either one of those guys to hit 17-18 fantasy points.

(You also have to take into account -- when any of those players fall short, by how much do they fall short? Even if they don't hit 18 points, 11 points is better than 2. Dunbar and Powell are way, way more likely to end up with 2, IMO. Similarly, when they do hit, how much do they hit by? 28 points is better than 18, and Williams is way, way more likely to hit 28, IMO.)

I think Karlos Williams is a very good GPP play ... because he is a very good play, period.

 

[jandyt]

If you have basically the same roster, but instead of Williams go with another $3400 or less back that hits and is hardly on any rosters that is a good thing.

You don't just have to hope that your $3,400 guys hits. You have to hope that your guy hits and Williams doesn't. Specifically, if Williams is 50% owned and your $3,400 running back is 5% owned, you have to hope that your guy will hit while Williams doesn't at least ten times as often as Williams hits and your guy doesn't.

I personally don't see anyone in the $3,400 range who I think satisfies that criterion. Maybe Lance Dunbar ($3,600) or Bilal Powell ($3,700) since DK is full-PPR, but I still think Williams is a huge favorite over either one of those guys to hit 17-18 fantasy points.

(You also have to take into account -- when any of those players fall short, by how much do they fall short? Even if they don't hit 18 points, 11 points is better than 2. Dunbar and Powell are way, way more likely to end up with 2, IMO. Similarly, when they do hit, how much do they hit by? 28 points is better than 18, and Williams is way, way more likely to hit 28, IMO.)

I think Karlos Williams is a very good GPP play ... because he is a very good play, period.

Completely agree I think on DK your trying to get too cute if your switching Williams for another RB in his price range but there just aren't any comparable ones. Now on other sites where his price is higher there may be options. For example on FantasyScore you can get Gore for 200 more or Gordon for 100 more. Those guys have much more comparable upsides of course Williams ownership % will likely not be as high as it will be on DK.

 

[sbuck143]

Bringing this back to life a little, I was reading the "Consensus Value Plays" and saw this

Danny Tuccitto - Cook has a top-ranked 60.8% chance of achieving GPP value and a second-ranked 67.3% chance of achieving cash value. (Thu 4:37 AM)

Is there a place I can get those numbers for all players?

Yeah, that right there is the science of statistics in action, my friends. Would love to see the backup behind these numbers.

This.

I've noticed Mr Tuccitto's % to hit value numbers in the Consensus articles all season, and have been searching for a source where he provides that in aggregrate. No luck so far.

 

[KarmaPolice]

If you have basically the same roster, but instead of Williams go with another $3400 or less back that hits and is hardly on any rosters that is a good thing.

You don't just have to hope that your $3,400 guys hits. You have to hope that your guy hits and Williams doesn't. Specifically, if Williams is 50% owned and your $3,400 running back is 5% owned, you have to hope that your guy will hit while Williams doesn't at least ten times as often as Williams hits and your guy doesn't.

I personally don't see anyone in the $3,400 range who I think satisfies that criterion. Maybe Lance Dunbar ($3,600) or Bilal Powell ($3,700) since DK is full-PPR, but I still think Williams is a huge favorite over either one of those guys to hit 17-18 fantasy points.

(You also have to take into account -- when any of those players fall short, by how much do they fall short? Even if they don't hit 18 points, 11 points is better than 2. Dunbar and Powell are way, way more likely to end up with 2, IMO. Similarly, when they do hit, how much do they hit by? 28 points is better than 18, and Williams is way, way more likely to hit 28, IMO.)

I think Karlos Williams is a very good GPP play ... because he is a very good play, period.

Thanks MT, great post. I am very new to this and hadn't considered that my guy would have to hit Xtimes more if it is a really small ownership player- makes so much sense.

The other player I had considered in a gpp is Rawls if Lynch was sitting out or very limited. I was considering a Tyrod/Clay stack and loading up on stud wrs, so I was looking digging for another possible dirt cheap rb option.

 

[Maurile Tremblay]

If you have basically the same roster, but instead of Williams go with another $3400 or less back that hits and is hardly on any rosters that is a good thing.

You don't just have to hope that your $3,400 guys hits. You have to hope that your guy hits and Williams doesn't. Specifically, if Williams is 50% owned and your $3,400 running back is 5% owned, you have to hope that your guy will hit while Williams doesn't at least ten times as often as Williams hits and your guy doesn't.

I personally don't see anyone in the $3,400 range who I think satisfies that criterion. Maybe Lance Dunbar ($3,600) or Bilal Powell ($3,700) since DK is full-PPR, but I still think Williams is a huge favorite over either one of those guys to hit 17-18 fantasy points.

(You also have to take into account -- when any of those players fall short, by how much do they fall short? Even if they don't hit 18 points, 11 points is better than 2. Dunbar and Powell are way, way more likely to end up with 2, IMO. Similarly, when they do hit, how much do they hit by? 28 points is better than 18, and Williams is way, way more likely to hit 28, IMO.)

I think Karlos Williams is a very good GPP play ... because he is a very good play, period.

Thanks MT, great post. I am very new to this and hadn't considered that my guy would have to hit Xtimes more if it is a really small ownership player- makes so much sense.

The other player I had considered in a gpp is Rawls if Lynch was sitting out or very limited. I was considering a Tyrod/Clay stack and loading up on stud wrs, so I was looking digging for another possible dirt cheap rb option.

I accidentally inverted things -- I meant at least 10% as often, not 10 times as often. A typo of sorts, but my point remains that I think Williams is a huge favorite to outscore any of the other RBs in his price range.

 

[Maurile Tremblay]

Here are some of my thoughts on uniqueness in GPPs.

Uniqueness doesn't matter at all. Ownership percentage matters -- but uniqueness is the wrong word for what matters about it. (And not just in the overly pedantic sense that uniqueness, like pregnancy, is strictly binary.)

The players you should be looking for are under-owned players, and -- as Karlos Williams may demonstrate rather well -- under-ownedness and uniqueness are entirely different concepts.

Real DFS is complicated, so let's consider a simplified toy game to examine the relevant principle.

Suppose we have a 100-person contest with a tournament-style payout structure (you can make it winner-take-all for simplicity; it doesn't matter), and each lineup consists of a single player -- either the Broncos defense or the Raiders defense. Suppose the Broncos defense is a 2-1 favorite to outscore the Raiders defense, and suppose that 80% of the entrants will own the Broncos.

It is easy to see that the lineups with the Raiders will win money, on average, while the lineups with the Broncos will lose money. (Two-thirds of the time, 80% of the lineups will split the prize pool; one-third of the time, 20% of the lineups will split the prize pool. The latter group is the one that is +EV.)

The Raiders defense was more unique, and that's the one that was profitable. Does that mean that uniqueness is beneficial?

No. What made the Raiders defense profitable was that its 33% win rate was greater than its 20% ownership, and in that sense they were under-owned. But it's not at all the case that unique players are inherently under-owned while chalk players are inherently over-owned. Change the numbers from the example a bit -- make the broncos 7-1 favorites to outscore the Raiders rather than 2-1 favorites, and now it's the Broncos, at 80% ownership, that are under-owned while the Raiders, at 20%, are over-owned.

In general, uniqueness might be correlated with under-ownedness. That's an interesting empirical question that is worth looking into. But the correlation would not be perfect in any case, and there may be plenty of situations where 60%-owned guys might be under-owned because they are actually 70% likely to outscore similarly priced players at their position -- and that may well be the case with Karlos Williams this week.

See also: similar thoughts (but different in ways I'll elaborate on next week) from Steve Buzzard and John Lee.

 

[KarmaPolice]

Here are some of my thoughts on uniqueness in GPPs.

Uniqueness doesn't matter at all. Ownership percentage matters -- but uniqueness is the wrong word for what matters about it. (And not just in the overly pedantic sense that uniqueness, like pregnancy, is strictly binary.)

The players you should be looking for are under-owned players, and -- as Karlos Williams may demonstrate rather well -- under-ownedness and uniqueness are entirely different concepts.

Real DFS is complicated, so let's consider a simplified toy game to examine the relevant principle.

Suppose we have a 100-person contest with a tournament-style payout structure (you can make it winner-take-all for simplicity; it doesn't matter), and each lineup consists of a single player -- either the Broncos defense or the Raiders defense. Suppose the Broncos defense is a 2-1 favorite to outscore the Raiders defense, and suppose that 80% of the entrants will own the Broncos.

It is easy to see that the lineups with the Raiders will win money, on average, while the lineups with the Broncos will lose money. (Two-thirds of the time, 80% of the lineups will split the prize pool; one-third of the time, 20% of the lineups will split the prize pool. The latter group is the one that is +EV.)

The Raiders defense was more unique, and that's the one that was profitable. Does that mean that uniqueness is beneficial?

No. What made the Raiders defense profitable was that its 33% win rate was greater than its 20% ownership, and in that sense they were under-owned. But it's not at all the case that unique players are inherently under-owned while chalk players are inherently over-owned. Change the numbers from the example a bit -- make the broncos 7-1 favorites to outscore the Raiders rather than 2-1 favorites, and now it's the Broncos, at 80% ownership, that are under-owned while the Raiders, at 20%, are over-owned.

In general, uniqueness might be correlated with under-ownedness. That's an interesting empirical question that is worth looking into. But the correlation would not be perfect in any case, and there may be plenty of situations where 60%-owned guys might be under-owned because they are actually 70% likely to outscore similarly priced players at their position -- and that may well be the case with Karlos Williams this week.

See also: similar thoughts (but different in ways I'll elaborate on next week) from Steve Buzzard and John Lee.

Makes sense, and I have been reading the article that predicts a handful of players owned %. Besides that, is it just a blind guess for average Joe Schmoes like me as to what players are likely to be owned more than others?

 

[Quint]

Here are some of my thoughts on uniqueness in GPPs.

Uniqueness doesn't matter at all. Ownership percentage matters -- but uniqueness is the wrong word for what matters about it. (And not just in the overly pedantic sense that uniqueness, like pregnancy, is strictly binary.)

The players you should be looking for are under-owned players, and -- as Karlos Williams may demonstrate rather well -- under-ownedness and uniqueness are entirely different concepts.

Real DFS is complicated, so let's consider a simplified toy game to examine the relevant principle.

Suppose we have a 100-person contest with a tournament-style payout structure (you can make it winner-take-all for simplicity; it doesn't matter), and each lineup consists of a single player -- either the Broncos defense or the Raiders defense. Suppose the Broncos defense is a 2-1 favorite to outscore the Raiders defense, and suppose that 80% of the entrants will own the Broncos.

It is easy to see that the lineups with the Raiders will win money, on average, while the lineups with the Broncos will lose money. (Two-thirds of the time, 80% of the lineups will split the prize pool; one-third of the time, 20% of the lineups will split the prize pool. The latter group is the one that is +EV.)

The Raiders defense was more unique, and that's the one that was profitable. Does that mean that uniqueness is beneficial?

No. What made the Raiders defense profitable was that its 33% win rate was greater than its 20% ownership, and in that sense they were under-owned. But it's not at all the case that unique players are inherently under-owned while chalk players are inherently over-owned. Change the numbers from the example a bit -- make the broncos 7-1 favorites to outscore the Raiders rather than 2-1 favorites, and now it's the Broncos, at 80% ownership, that are under-owned while the Raiders, at 20%, are over-owned.

In general, uniqueness might be correlated with under-ownedness. That's an interesting empirical question that is worth looking into. But the correlation would not be perfect in any case, and there may be plenty of situations where 60%-owned guys might be under-owned because they are actually 70% likely to outscore similarly priced players at their position -- and that may well be the case with Karlos Williams this week.

See also: similar thoughts (but different in ways I'll elaborate on next week) from Steve Buzzard and John Lee.

Makes sense, and I have been reading the article that predicts a handful of players owned %. Besides that, is it just a blind guess for average Joe Schmoes like me as to what players are likely to be owned more than others?

not necessarily. i've seen info (not sure how it was found) of what ownership was for Thursday GPPs on FD. since DK allows lineup changes within contests already started, it might be a way to look at those FD #s and see how you might find some [potentially] low-owned guys for your DK lineups.

 

[Rock Lonemilk]

Here are some of my thoughts on uniqueness in GPPs.

Uniqueness doesn't matter at all. Ownership percentage matters -- but uniqueness is the wrong word for what matters about it. (And not just in the overly pedantic sense that uniqueness, like pregnancy, is strictly binary.)

The players you should be looking for are under-owned players, and -- as Karlos Williams may demonstrate rather well -- under-ownedness and uniqueness are entirely different concepts.

Real DFS is complicated, so let's consider a simplified toy game to examine the relevant principle.

Suppose we have a 100-person contest with a tournament-style payout structure (you can make it winner-take-all for simplicity; it doesn't matter), and each lineup consists of a single player -- either the Broncos defense or the Raiders defense. Suppose the Broncos defense is a 2-1 favorite to outscore the Raiders defense, and suppose that 80% of the entrants will own the Broncos.

It is easy to see that the lineups with the Raiders will win money, on average, while the lineups with the Broncos will lose money. (Two-thirds of the time, 80% of the lineups will split the prize pool; one-third of the time, 20% of the lineups will split the prize pool. The latter group is the one that is +EV.)

The Raiders defense was more unique, and that's the one that was profitable. Does that mean that uniqueness is beneficial?

No. What made the Raiders defense profitable was that its 33% win rate was greater than its 20% ownership, and in that sense they were under-owned. But it's not at all the case that unique players are inherently under-owned while chalk players are inherently over-owned. Change the numbers from the example a bit -- make the broncos 7-1 favorites to outscore the Raiders rather than 2-1 favorites, and now it's the Broncos, at 80% ownership, that are under-owned while the Raiders, at 20%, are over-owned.

In general, uniqueness might be correlated with under-ownedness. That's an interesting empirical question that is worth looking into. But the correlation would not be perfect in any case, and there may be plenty of situations where 60%-owned guys might be under-owned because they are actually 70% likely to outscore similarly priced players at their position -- and that may well be the case with Karlos Williams this week.

See also: similar thoughts (but different in ways I'll elaborate on next week) from Steve Buzzard and John Lee.

It's gonna take a while for me to wrap my brain around this. Thanks for sharing, MT!

 

[Ned]

MT does such a great job here.

 

[Charybdis]

So... does anyone know where to get the Danny Tuccitto's percentages to hit cash and tournament?